03 November, 2009


The Attacotti are mentioned in a small number of sources as a tribe who attacked Late Roman Britain in the second half of the fourth century. Who were they, and where did they come from?


Ammianus Marcellinus

The major source for the Attacotti’s existence is the Roman historian Ammianus Marcellinus, who wrote a history of Rome in the late fourth century. In Book 27 of his history, he writes:

It will, however, be in place to say, that at that time the Picts, divided into two tribes, called Dicalydones and Verturiones, as well as the Attacotti, a warlike race of men, and the Scots, were ranging widely and causing great devastation; while the Gallic regions, wherever anyone could break in by land or by sea, were harassed by the Franks and their neighbours, the Saxons, with cruel robbery, fire, and the murder of all who were taken prisoners.
--Ammianus Marcellinus

“At that time” refers to 364 AD, so his account is roughly contemporary with the events described.

St Jerome

St Jerome was a Christian priest who lived between about 350 and about 420 AD, and who travelled to Gaul some time around 365–370 AD. In one of his writings, he mentions the Attacotti as a British tribe and describes them as cannibals:

Why should I speak of other nations when I, a youth, in Gaul beheld the Attacotti, a British tribe, eat human flesh, and when they find herds of swine, cattle, and sheep in the woods, they are accustomed to cut off the buttocks of the shepherds, and the paps of the shepherdesses, and to consider them as the only delicacies of food.
--Quoted in the Wikipedia entry

The Wikipedia entry says the Latin is capable of a less dramatic interpretation, if the word “humanis” (human flesh) is a mistake for “inhumanis” (animal flesh”, in which case the Attacotti’s dietary preferences would be “haunches of fatted animals” and “sow belly or cow’s udder”. I’m not qualified to comment on the Latin, but I have to say I find this a much more plausible scenario. Cow udder is a traditional dish, along with things like pig’s head brawn, tripe and chitterlings. Animal haunches – otherwise known as hams – need no comment.

Notitia Dignitatum

The Notitia Dignitatum (List of Offices) is an official list of late Roman administrative and military posts from about 400 AD. Some of the military units listed have names that could be variant spellings of Attacotti:

Section VII:
In Italy:
Atecotti Honoriani iuniores

In the Gauls with the illustrious master of horse in Gauls:
Atecotti Honoriani seniores
Atecotti iuniores Gallicani.
--English translation, omitting the lists of units, Latin text, including the lists of units

If these refer to the same tribe as the Attacotti of Ammianus Marcellinus and St Jerome, this suggests that the Late Roman Army had recruited some troops from the rebellious British tribe and sent them off to serve elsewhere in the Empire. Whether the service was voluntary (for the promise of pay and the chance to see the world), or compulsory as part of the price of defeat, or a bit of both, is open to question.

Where were the Attacotti from?

St Jerome, a contemporary who could have met some of the Atecotti soldiers stationed in Gaul, is clear that they were a British tribe. Ammianus, also contemporary, considers them to be distinct from both the Picts and the Scots (Irish). Since they attacked Roman Britain and since Ammianus brackets them with other tribes from outside the Empire, it’s a reasonable inference that they did not live within the Roman province of Britannia.

I can think of two plausible locations for the Attacotti:

  • One of the tribes living in what is now southern Scotland/north-east England, north of Hadrian’s Wall but outside the area associated with the Picts

  • A tribe living in the area associated with the Picts, but sufficiently culturally distinct to be considered a separate group by Roman observers.

Southern Scotland/north-east England

Ptolemy’s Geography, compiled in the second century AD, lists four tribes living in what is now southern Scotland/north-east England, roughly in the area north of Hadrian’s Wall and south of the Forth–Clyde line. This area was outside the Roman province in 360. In later sources the Picts are usually associated with the area north of the Forth–Clyde line. I suspect that the term “Picts” was applied rather vaguely, and perhaps meant different things at different times to different people, but if Ammianus applied the same regional association the tribes of southern Scotland would not have counted as Picts. The tribes listed by Ptolemy are:

The Novantae dwell on the side toward the north below the peninsula of this name
Below are the Selgovae
From these toward the east, but more northerly, are the Damnoni
Further south are the Otalini
--Ptolemy, Geography, Book 2

None of these names looks obviously related to “Attacotti”, but the name “Picts” doesn’t appear in Ptolemy’s Geography either. It is entirely possible that one (or more) of the tribes acquired a new name between the second century and the fourth, or that “Attacotti” was invented as a new umbrella term to group them together. This possible location, combined with St Jerome’s lurid description, may underlie the legend that a race of cannibals once lived in the region of Glasgow.

Culturally distinct group among the “Picts”

Since St Jerome says the Attacotti were a British tribe, I’ll take that as an indication that they came from mainland Britain, not Ireland, and consider where they might be found amongst the “Picts”, but the same line of argument could be applied equally well to an Irish tribe.

“The Picts” was clearly a sort of umbrella term for a multiplicity of different tribes. Ammianus Marcellinus recognises two subdivisions, and there may well have been many more. The Pictish origin legend refers to seven regions, and Ptolemy’s Geography lists many tribes in what is now Scotland north of the Forth and Clyde. I touched on the likelihood of multiple regional groupings among the “Picts” in my earlier post about the name, and Jonathan Jarrett has discussed the issue in more detail. Similarly, Ptolemy reports a large number of separate tribes in Ireland. One would not necessarily expect Latin writers based in the Mediterranean lands to be experts in the detailed nomenclature or comparative anthropology of hostile “barbarian” tribes from beyond the fringes of the known world. The terms “Picts” and “Scots” may have been rather vague catch-all labels, perhaps (probably?) no more precise than modern labels like “Asian”.

If the Atecotti army units in the Notitia Dignitatum were indeed recruited from the Attacotti tribe, there is the possibility that they, or records about their recruitment, were the source of Ammianus’ and St Jerome’s information about the Attacotti. In which case, “Attacotti” may have been their own name for themselves. Presumably the Roman army bureaucracy would have wanted to know what to call the new recruits, and the simplest way to find out would have been to ask them. If the Attacotti thought of themselves as a distinct tribe, and either didn’t accept or had never heard of the Roman label of Pict, they would naturally give their tribal name and the scribe would naturally write it down as best he could.

A related possibility is that the Attacotti were somehow sufficiently distinct from the Roman idea of a “Pict” for Roman observers to conclude that they must be a separate tribe. This could have been due to a difference in any cultural marker - customs, religion, language, appearance, etc. For example, if the Romans assumed all “Picts” were “painted people”, maybe the Attacotti didn’t use body paint or tattoos? Material culture certainly varied widely across the territory associated with the “Picts” (see Jonathan Jarrett’s article for some examples). I’ll focus on one: the brochs.

The broch-builders

Brochs are impressive and sophisticated drystone towers, found only in what is now Scotland. Many have a double-skinned wall with a passageway and steps in the space between the inner and outer walls, and appear to have been two- or three-storey buildings. The double-skinned wall would act as a barrier to stop rain seeping in to the dwelling areas, and would also have helped circulate heat through the structure (see explanation and a reconstruction drawing here). If the cattle lived on the ground floor in the winter they would have contributed to the central heating – you get a lot of heat off a cow – without too much in the way of smells or mess in the dwelling area. As usual, there’s a debate about the purpose of brochs – defensive castle, farmhouse or stately home? – and there’s no reason why they couldn’t have fulfilled more than one role at different times and places.

Not only are brochs confined to what is now Scotland, they are concentrated in defined areas, mainly Caithness (the north-east corner of the mainland), the Northern Isles (Orkney and Shetland), and the Western Isles (see distribution map on the Wikipedia page). This restricted distribution is consistent with (though does not prove) the possibility that brochs were mainly built and used by one or a few tribes.

As an interesting straw in the wind, it’s worth noting that Norse place names in Scotland are also heavily concentrated in the Northern and Western Isles and to a lesser extent in Caithness (Graham-Campbell and Batey 1998). I should stress that I am not suggesting a direct association between Norse place names and brochs. For a start, they are separated by a thousand years or so, as brochs are mostly considered to have been built in the century or so either side of 0 AD, and Norse place names are mostly considered to date from around the ninth to twelfth centuries. However, the one thing that never changes about history is geography, as the saying goes. The Northern Isles and Caithness are the areas most obviously open to seaborne contact with Norway. Maybe there was cultural contact between these regions long before the historical Norse (Viking, if you prefer) settlements in Scotland, leading to the development of a distinctive cultural identity among the people living in the Northern and Western Isles and Caithness, expressed in the building of brochs (and possibly also in other ways that haven’t left evidence).

Place name

The Pictish origin legend says that their land was divided between the seven sons of Cruithne. In the Pictish Chronicle their names are given as:

Fib, Fidach, Floclaid, Fortrenn, Got, Ce, Circinn
--Pictish Chronicle

In the Irish translation of Historia Brittonum their names are given as:
Moirfeisear do Cruithne claind
Roindsed Albain a seacht raind
Cait, Ce, Cireach cetach cland,
Fib, Fidach, Fotla, Foirtreand.
--Historia Brittonum, Irish

“Got” or “Cait” is the origin of the modern name Caithness. Clutching at straws, how about a connection between “Cait” and the “-cott-” element in Attacotti? This is no more than dictionary fishing on my part, and I am not qualified to say whether there is any possible basis for a connection on linguistic grounds, so it may well just be a superficial resemblance. But possibly an interesting one.

Speculative interpretation

How about the broch-builders or their successors, living in what is now Caithness and the Northern and/or Western Isles, as a candidate for a culturally distinct tribe who in the 360s AD were called the Attacotti by Ammianus Marcellinus and St Jerome?

If other aspects of their culture were as distinctive as their architecture, such a tribe may well have seemed sufficiently culturally distinct from the other tribes the Romans called “Picts” to warrant a separate name.

Contact with Norway across the North Sea may have stimulated the development of a distinctive culture in Caithness and the Northern and/or Western Isles, as happened with the Norse (Viking) influence in similar areas in later centuries.

An echo of the name Attacotti may – and I stress ‘may’ – possibly be traceable in the name of Caithness.

Needless to say, other interpretations are possible.

Ammianus Marcellinus, available online
Graham-Campbell J, Batey CE. Vikings in Scotland: an archaeological survey. Edinburgh University Press, 1998, ISBN 978-0748606412. Searchable at Google Books.
Historia Brittonum, Irish, available online
Notitia Dignitatum, available online, English translation, omitting the lists of units, Latin text, including the lists of units
Pictish Chronicle, available online
Ptolemy, Geography, Book 2, available online

Map links
Location map showing Shetland and Orkney (The Northern isles) in relation to Scotland and Norway


tenthmedieval said...

I rather like this theory (and not just because of the links to me, though thankyou for them!); I think that fitting known names to known culture groups is probably a better way to look at this time and area than calling it `pre-Pictish', at any rate. That Wikipedia entry is a bit useless with its Jerome citation isn't it? Everything linked to unreachable Google Books in snippet view and no page references. I'll try and get a reference to let me find the original Latin and see if there's any variation over the manuscripts about `humanis'.

Carla said...

Tenthmedieval - I'm glad you like the theory :-) If you do find a way into the original Latin for St Jerome (if you have the time between all your other commitments!), I'd be very interested in your opinion on 'humanis'. I daresay the Wikipedia entry writer did their best, and at least they quoted the Latin text as well as a translation.

Another possibility that I missed mentioning in the post may be 'Attacotti' as the fourth-century name for the people and territory that were called Dal Riada in the sixth century. If they were already culturally distinctive in the 4th century they may not have counted as 'Picts' to a Latin writer.

I'm uneasy about 'pre-Pictish' as a label because it implies there was a significant cultural change of some sort from 'pre-Picts' to 'Picts', whereas it seems to me that it could just be a change of nomenclature in the documentary sources.

Lady D. said...

I'd always wondered whether the Picts were composed of seperate tribes or were one people - and now you've answered that for me, thankyou. Fascinating post!

tenthmedieval said...

I don't like the Dál Riata suggestion half as much, I'm sure they would have been Scotti to most outside observers. I confess that I have wondered whether Attacotti isn't just a made-up name meaning 'opposite the Scotti', though, like `ad scotti', not that that would be grammatical but you see where I'm going with it.

Cambridge UL has the Lezius book that the Wikipedia page cites as source for its Latin, albeit in closed storage, so next time I go I'll try and find the cite that tells us where Jerome actually said this, and then it's hopefully just a matter of playing with the Corpus Christianorum a bit. If you're able to start in the right place, the work of half an hour, in an ideal world...

Carla said...

Lady D - Like most questions in this era, I don't think it's entirely settled yet! Contemporary sources so often refer to subgroups of Picts (Ammianus Marcellinus in the post; Bede refers to northern and southern Picts) that it seems clear that they weren't homogeneous. Whether they were one people with some subdivisions, or were separate tribes in some sort of confederation, or were separate tribes whom outsiders lumped under one label, or indeed any or all of these at different times, is a matter for argument.

Tenthmedieval - Good point, although if the Scotti were thought to come from Hibernia (Ireland) and the Attacotti were known to live in what's now Argyll, could that have been enough to make them not-Scotti to a Roman observer?
'Opposite the Scotti' - would that mean 'not Scotti' (i.e. different from the Scotti in some way), or would it refer to territory, i.e. living across the sea from the Scotti? If I remember rightly Tacitus refers to Wales as being opposite Spain when he's drawing a comparison between the Silures and the Spanish tribes, and I think Julius Caesar refers to parts of southern Britain as being opposite Gaul. Would that fit with your ad Scotti at all?

tenthmedieval said...

I guess it could be read either way even if my dubious etymology be accepted. After all, a Roman observer wouldn't necessarily have known how strong the ties between the groups across the sea from each other were.

Lady D, the extreme view is one put by, among others, the late Leslie Alcock: "the Picts were those who, at any one time, paid tribute, and especially military service, to a rex Pictorum". At the other end there is a body of opinion that claims that the Picts were the pre-Celtic pre-Indo-European indigenous inhabitants of Scotland shunted aside by all the successive waves of invaders. What, in that latter case, we are supposed to call all the Celts with their Celtic language who seem to have been in charge by the time anyone uses the word 'Pict', is never quite clear...

Sarah Cuthbertson said...

Thanks for gathering all this fascinating evidence and for the links. But what about Attacotti Nell? No evidence for her? Perhaps her exploists weren't the kind of thing that gets recorder or preserved.

Gabriele C. said...

That's an interesting theory. Carla. It's not impossible that the Northmen got around earlier. We tend to underestimate the naval abilities of those old people - I'm sure there's some nucleus to Brendan's travels as there is to Odysseus' adventures. Some adventurous pre-Vikings might have got over to Scotland. Never thought about it before.

Re. tribal names in Latin sources, well, after the mess Caesar - who after all had been there - made of the Celtic and Germanic tribes at the Rhine, I don't trust those guys, lol. And most of them haven't been in the lands they describe.

Considering the social structures outside the Roman Empire in the north-west at those times, several tribes connected by culture and religion, who feuded as often as they formed alliances, seem to be more plausible. There was a tendency to larger conglomerates in the 4th century (the Germanic Allamanni come to mind) but in my opinion, the earlier Novantae, Votadini, Epidii and other tribes are as much Picts as the later Dicalydones and Verturiones (both names sound very Latin, btw). The same way fe. the Cherusci, Chatti, Bructeri and Hermunduri had been Germanic tribes in the 1st and 2nd centuries.

Carla said...

Tenthmedieval - the pre-IE theory has always puzzled me with regard to the P-Celtic place names that turn up in the Pictish areas, e.g. Aberdeen. It sort of implies two language changes, one from some pre-IE language to P-Celtic, and then another from P-Celtic to Gaelic.

Sarah - Alas, no evidence for Attacotti Nell, who is entirely a product of my imagination (possibly wishful thinking on Drust's part)! The contents of this post do underlie about half a line in Exile.

Gabriele - thank you. I agree that early cultures were probably a good deal more mobile than we tend to assume, at least for specific groups (elite, religious, traders, exiles) if not for the bulk of the population. Tim Severin showed that a leather boat could go from Ireland to Scotland to the Faroes to Iceland to Canada, so I don't see why similar craft couldn't have travelled between Scotland and Scandinavia even before the Viking longboats came along.

I tend to agree that the four tribes between the Walls probably had a lot in common with the tribes further north in Ptolemy's map who later became called the 'Picts'. The model of tribes related by culture and religion who sometimes feuded and sometimes formed alliances, according to circumstance and inclination, seems quite likely to me. As you say, tribal nomenclature is problematic - the Latin sources we have aren't exactly detailed anthropological records, and one wonders how much the writers knew, or were interested in, getting the names and relationships of distant barbarian tribes right. I suppose we can be grateful that they didn't just call all of them 'barbarians' and have done with it!

Gabriele C. said...

There may also have been shifting patterns in the tribal structure. During the early Empire, the tribes seem to have been smaller, and if they worked together it was a temporary alliance held together by the personality of a leader (Calgacus, Arminius) and without such a leader and the outside pressure from the Romans, the confederacy would fall apart.

But in the 3rd century there was a tendency to form larger groups in Germany, like the Alamanni ('all people') that swallowed a number of smaller tribes. They were bold - and strong - enough to attack the Limes border several times until the Romans withdrew to the Rhine in 260. While the Alamanni concentrated in southern Germany, another larger group, the Saxons, evolved in the northern part. Their interest seems to have been the lands beyond the sea and not so much the Rhine frontier. Look where some of them went, lol.

We can't be sure that the Roman use of Alamanni covered the same as the actual German use (if the battle of 236 indeed took place at Kalefeld, that would be very far north for a fight against the Alamanni, so they either spread further in the 3rd century than in the 4th, or the Romans failed to understand that they were fighting against a different group; the sources aren't contemporary, either), but the tendency to larger tribes is there. The same pattern could have happened with the Dicalydones and Verturiones; they could have been larger tribes that in the 3rd century evolved out of the whole mess of Novantae, Cerones, Caledones, Selgovae and whatever else was around in Scotland and the borders in the first century. And at some point the Romans may have realised the change and understood that they were facing larger, more coherent groups. Maybe that's the reaseon for Septimius Severus' attempt to carry the war north of the wall again; he might have tried to nip something in the bud. The first major Alamannic incursion happened during his time.

Rick said...

Are there anything like brochs in Norway? (I infer that there aren't.) In any case the distribution does suggest a connection with some sort of maritime population.

As usual, there’s a debate about the purpose of brochs – defensive castle, farmhouse or stately home? – and there’s no reason why they couldn’t have fulfilled more than one role at different times and places.

Particularly since those three functions are anything but mutually exclusive.

Carla said...

Gabriele - interesting parallel between the 'Picts' and the Alamanni, if they both started forming confederations at the same time. Severus' invasion beyond the Wall was presumably a response to a perceived increase in the level of threat (unless it was a PR stunt?), which might have arisen from a shift in tribal political structures even if he may not have necessarily been aware of that at the time.

Rick - as far as I know, brochs are unique to northern Scotland and the isles. So they aren't a direct import from Scandinavia, but they do seem to belong to a localised culture. My suggestion is just that this localised culture might have had contact across the sea with Scandinavia and such contact might have reinforced or contributed to its distinctiveness.

badonicus said...

Hi Carla,

Have you read Philip Rance's ‘Attacotti, Déisi and Magnus Maximus: The Case for Irish Federates in Late Roman Britain’, 2001? He suggest that they were a sept of the Déisi of Demetia and known as the aichechthúatha (‘client people’), so would have been in southwestern Wales. It's worth a read.

Carla said...

Badonicus - Hello and welcome. I've come across his theory. Since Ammianus brackets the Attacotti with Picts and Scots, who were from areas outside the Roman province, I tend to favour a location outside Roman Britain for the Attacotti, however Ammianus doesn't specify their location. As so often, many interpretations are possible.

badonicus said...

There is one slight problem I see with them being from outside Britannia, although I do agree with your point about the North. Those others that raided and were defeated don’t appear to have been captured and set up as Roman military units. Why this group? There would had to have been a substantial number of them, one would think, for them to be used. This was usually done when the Romans defeated a region and took that region. They then shipped the men from there abroad so they couldn't cause anymore trouble. I don't know of them defeating a region, shipping of men and then leaving it ... but that just could be my lack of knowledge. So this should mean that the Romans had to (re)take an area north of Britannia's borders, which seems highly unlikely.

Just a thought.

Carla said...

Badonicus - it might be connected with the mysterious province of Valentia, which Ammianus says that Theodosius 'recovered' in the late 360s. The location of Valentia isn't known either, so it's not known whether it was somewhere outside Roman Britain (perhaps the area between Hadrian's Wall and the Antonine Wall, which had been part of the Empire much earlier and might have counted as being 'recovered') or somewhere within Roman Britain that had been invaded and conquered (or rebelled) and was then retaken by Theodosius. So unfortunately, even if the Attacotti military units were connected with the recovery of Valentia, as we don't know where Valentia was it doesn't much help with locating the Attacotti either. As ever, many interpretations are possible.

badonicus said...

Hi Carla - My thoughts on Valentia are too long to post here, but I wrote a blog about the very question of where it might have been. If you'd like to read it, the address is below:


Carla said...

Badonicus - thanks for the link. As ever, the evidence can be interpreted many different ways! There's a theory that Chester/Deva was originally intended as the capital of Britannia, with the enigmatic Elliptical Building serving some sort of religious or ceremonial function. It would be rather nice in a way if it ended up as the capital of a province, even if only briefly :-)