Friday, March 11, 2016

Smelting the World: Anderita

Today, we're on about Pevensey Castle. At least, once I get there. First, some brisket!


 "You know what they're calling this bond drive? The Mighty Seventh. They might've called it the "We're Flat Fucking Broke And Can't Even Afford Bullets So We're Begging For Your Pennies" bond drive, but it didn't have quite the ring. They could've called it that, though, because the last four bond drives came up so short we just printed money instead. Ask any smart boy on Wall Street, he'll tell you our dollar is next to worthless, we've borrowed so much. And nobody is lending any more. Ships aren't being built, tanks aren't being built, machine guns, bazookas, hand grenades, zip. You think this is a farce? You want to go back to your buddies? Well stuff some rocks in your pockets before you get on the plane, because that's all we got left to throw at the Japanese. And don't be surprised if your plane doesn't make it off the runway, because the fuel dumps are empty. And our good friends, the Arabs, are only taking bullion. If we don't raise $14 billion, and that's million with a "B," this war is over by the end of the month. We make a deal with the Japanese, we give whatever they want and we come home, because you've seen them fight, and they sure as shit ain't giving up. $14 billion! The last three drives didn't make that much all together."

Saudi oil production, btw. 1945 is to the left of the start of the curve. You know, where the oil production is really, really low.


Not to be too mysterious here. This is a follow-up. First, I don't want to risk misrepresenting the data on coal mining in Britain at the end of the World War, and neither do I want to sink time into understanding it right now --the winter of 1947 will come in its own time. So I am going to reproduce the original charts from last week at the bottom of this number. There is considerable additional data in them, make of what one will. 

For now, it can stand to be pointed out that British coal miners are an early and paradigmatic example of a story that I am familiar with as pure racism  --the example I remember from my childhood was Trinidadian banana pickers. "There is no point in increasing the wage of [those people], because they will just book off work and go do things I disapprove of." There's a technical term for it from economics, and I'd be interested in the intellectual history of it if I could just turn up the phrase. I am, however, a great deal more interested in it in the context of the rapidly approaching labour shortage apocalypse of the Vancouver retail trades. The reader may also see where today's musical interlude comes in.

Second, I have Stephen Pax Leonard's Language, Society and Identity in Early Iceland before me. As Lameen warned, it is not very useful on the history of the Icelandic settlement. On the contrary, Leonard's objective is to resolve the very interesting mystery of the lack of dialectical variation in Icelandic. For this historical linguistic inquiry, he needs some kind of historical context, and takes the received, saga-based version at face value. (Fortunately, it is not critical to his thesis, and I imagine that he knows that there is a revisionist interpretation of the sagas as historicallly unreliable.) 

Here is a frustrating case in which I want to task the historical linguist, not with being too much in love with the hypothesis, but as being insufficiently willing to fly free on the wings of theory! I know. I am a very inconsistent and frivolous consumer of this science. Nevertheless, it seems that the mystery requires us to see the early settlers as speakers of many languages or at least Norse dialects, in order that they will require a koine to communicate with each other. Second, he sees a later phase of dialectical levellling, in which a model of ideal Icelandic is propounded (through the law assemblies, specifically) and becomes a social norm as a means of achieving a collectively-agreed Icelandic identity. This self-fashioning of collective identity might look to the sagas; but to a large extent to emphasise that Icelanders are not British (English or Irish.). That is, Icelanders need to make it clear that they are not British, and so need a common language. 

Out with the Procrustean bed! Clearly, poor Dr. Leonard's hypothesis pertains to the Commercial Fishing Horizon. Icelanders hold property rights to the foreshores. Foreshore ownership is absolutely critical to a fishing economy, as you can't get process the fish without it. If one must know what an Icelander is, and if there is no other way to tell them, it will have to be by the way they talk. 

What does that allow me to hypothesise about the pre-Commercial Fishing Horizon settlement of Iceland? Not much, other than that it must have been a linguistically diverse community, and perhaps that some Irish (and more English) is allowable. This is not the only product of a week spent thinking around the issues, and more importantly, frankly, the first quarterly grocery inventory since our latest reinvention. What if the earliest settlement of Iceland was by charcoal burners? Where is the archaeological state of the art on charcoal burning?

First, there is an interesting convergence of the exciting new terra preta work with what is actually happening in Brazil right now. Did you know that there is a charcoal-burning, pig-iron industry in Brazil, now? Today? The whole thing might even have a bearing on where the world is right now, as the fact that it is being done by poor Third Worlders is perhaps provoking a bit of a revisiting of "bio-char." (Charcoal making is bad, proletarian, unenvironmental, unhealthy, and leads to deforestation. "Bio-char" is good, because it remediates the soil and creates a fossil-free and renewable fuel.)

It is interesting that terra preta was discovered, or at least formalised as an archaeological paradigm (bad as Kuehn, I am with that word) in response to the sub-Roman "dark earth" layer in British urban excavations. We are not likely to find this conception of dark earth in the Roman British iron fields, because they were not built on cities. That being said, terra preta has been tardy in coming back to its British origins to be fully rehabilitated and understood as evidence of a community light on its feet, making iron, conditioning soils, and, yes, barbecuing beef with the firestick. Scholars have been looking for charcoal in Roman and sub-Roman Britain, and this will be the main focus after the break. No big revelations, but there's been some interesting work, for those who care about these things.

The final issue, a little out of order, is that laboriously retyped speech from Clint Eastwood's 2006 Iwo Jima epic, Flags of Our Fathers. The speech is so utterly anachronistic that I'm almost incapable of coherent response. The addition of the "Saudi Oil Production" chart is recent, a product of my third re-reading, as the previous two had achieved full apoplectic collapse before I got to the part where America's war effort in the spring of 1945 depended on "Arab" oil, for which the "Arabs" would only take bullion. It has been brought to my attention that Mr. Eastwoord isn't necessarily the most sophisticated of political thinkers, but I'm a little discouraged that something so blazingly silly appears in a critically well-received movie. (Though the Rotton Tomato rating of 73% suggests that regular readers had their difficulties with the headline reviews.) 

The speech wouldn't matter much more than a certain address to an empty chair were it has an independent existence on the Internet. The "Movie" subheading of the "War Bons in Popular Culture" Wikipedia article consists entirely of someone's retyping of this speech from the transcript. Whoever they are, they are, this Wikipedia contributor thinks that the war was funded by war savings bonds, and that America was about to run out of money to pay for the war in the spring of 1945, and, had that happened, America would have abandoned the Pacific war. 

This probably isn't the place to talk about Pacific First and the China Lobby. It's a story that will get old long before the seventieth anniversary of the declaration of the People's Republic arrives, much less the Korean War. (March of the Volunteeers.) What matters is that I'm a little discouraged that it isn't more obvious that the mid-century War loan drives weren't to fund the war. They were intended to soak up elevated wartime wages and prevent inflation. Even the secondary goal of spreading wartime spending over into peacetime demand wasn't much emphasised in bond sale propaganda. Although a little trawl through a Google Image search of war bond posters turned up the point that the Mighty Seventh was so important precisely because the war was almost over, and inflation was expected to overheat with demobilisation. (Thank you, Jones and Lawson Machinery Company of Springfield, Vermont!) 

But we just can't leave it at that, can we? It's never possible to just print money and spend it on an existential crisis of the state (or the species) without worrying that just putting that money out there will have some awful result. You can probably guess that I've had that frustrating argument a number of times lately in connection with the fight against climate change, so it's probably as well to take a moment to push back. Whether it's worthwhile to reconstruct the fall of the Roman Empire as an extended exercise in dear money is another matter

Anderitum is the name given in the Notum Dignitarum for what later became Pevensey Castle in East Sussex. As a bearer of Great Historical Siginficance, it is one of the forts of the "Saxon Shore," supposedly built to protect Roman Britain from the incoming Anglo-Saxon-Jutes. (Immigration is bad!)

It's a strange place.

Pevensey Castle seems to have been built at the head of the shallow estuary of the Pevensey Haven River, which has since silted in, part of the general process creating the Pevensey Levels. So the castle was probably a port overlooking tidal salt marshes in Roman times.

It's also a fairly important location from a strategic point of view, which is why there's a Medieval castle there. It's just odd that the castle is comfortably swallowed up by the Roman works.

By Pevensey_Castle_aerial_view.jpg: Lieven Smitsderivative work: Hchc2009 (talk) - Pevensey_Castle_aerial_view.jpg, CC BY-SA 3.0,
I mean, sure. A citadel in one corner of an enceinte is practically prototypical of a city, but this isn't a city. The Roman fort is the walls and towers. There's nothing, or, at least, no masonry inside. Ten hearths were found by excavators, presumably indicating timber or wattle and daub structures, so the structure can be explained as a garrison, and there is reason to think that there was a regiment here in cantonment, since one shows up in documents on campaign elsewhere, and it can't always have been on the march. So, sure,  Allectus may well have put a large garrison into this enclosure. 

Thing is, we're missing scale. Here's a ground-level picture. This is a lot of room.

By MortimerCat - Own work, CC BY-SA 3.0,

I mean, what the hell, Roman dudes? A big enceinte means a bigger garrison and more expense. This four hectare structure is about a fifth of the size of the Roman fort-city of York, and the kilometer or so of walls could easily have absorbed a cohort or more in a full-scale siege. Though why, exactly, an invading army which had already made it to shore would have wasted time besieging Pevensey is another question.

Pevensey looks like it took about 280,000 man-days to build, and the foundation work, which presented special difficulties in what was then a lacrustine environment, led to both the use of oaken timbers and coin scatters that allow us to date the construction to c. 290, and specifically to the periods of Carausius and Allectus, the consecutive Roman British usurpers of 286--93 and 293--6. So it was built in haste over three years or so, specifically to incorporate the entire tidal island of Pevensey without regards to the size of the garrison needed. This kind of implies that the scope of the work was driven by the need to incorporate the land, rather than the fort being built on a site where it could be big enough for the required garrison. What's going on, here?

Allectus wants you to know that he's all about the boats. By Classical Numismatic Group, Inc., CC BY-SA 3.0,

The specific timing suggests that Pevensey is a reaction to the siege and capture of Boulogne. (In case life is too short to follow he linnks, Carausis was onen of the separatist "usurpers" of the period in which Diocletian came to power and established the Roman empire's "Dominate" phase. He originally controlled Britain and chunks of France including the Pas de Calais around Boulogne, and Normandy around Rouen. Maximian, Diocletian's co-Caesar, reduced Boulogne in a siege in 293, then attempted to invade England, and suffered some kind of disaster, perhaps weather-sinking-ships related. After a three year delay, in which coins of Allectus circulated in the Pas de Calais and at Rouen, while Diocletian and Maximian revised their scheme of govenrment, a successful invasion ensued. This was led, not by Maximian, but by a new "junior colleague" of the emergent Tetrarchy, Constantius Chlorus, although the actual dirty work was done by the praetorian prefect, Julius Asclepiodotus, taken of late as a typical example of the Tetrarchy's collaborators amongst hte traditional Roman aristocracy.

 Panegyric informs us that the invasion of 296 was carried out by two fleets. One, leaving Rouen, commanded by Asclepiodotus, while the other, departing Boulogne, was commanded by Constantius Chlorus. Asclepiodotus avoiced Allectus' fleet near the Isle of Wight, allowing him to land, presumably in Solent Water. Allectus' army was encountered and defeated on land, but a residual force remained in London, and this was defeated by Constantius. who then made a naval entry into the provincial capital. Roger Rees reads this narrative as somewhat contrived, to emphasise the Tetrarchy's victory over the element of ocean, or Ocean, presumably seen as exemplifying anarchy or chaos or corruption or some such. It isn't that any of the details provided are wrong, but we should be open to the possibility that the facts have been cherry picked, especially in regards to the (marine) piracy at the root of the insurrection.

So Pevensey meets the specific nature of the crisis of 293/6 as Allectus understood it. The breakaway regime faced an existential threat, which required him to overthrow his boss and prepare his defences. A war fleet is attested; although we can afford to be skeptical. Pevensey is known to belong to this period, arhaeologically. It is part of the preparations for the final confrontation with Diocletian's regime. (Which, in 293--6, does not look as futile as it would in hindsight. Allectus is no prophet, and looking to recent history, one could hope that Diocletian's regime was already on borrowed time.)

This logic has led to characterisations of Pevensey as a defensive work against imminent invasion. This isn't necessarily wrong. Pevensey doesn't look like an optimal site for a Roman invasion of a breakaway province. It's only a "limb" of Hastings under the Cinque Port scheme. Even a landing in east Sussex does not necessarily have to go through Pevensey, and after that, any invasion of  England via East Sussex faces the inconvenient fact that it is going to have to cross the Wealdor go around it, as William the Conqueror did, on the way to London. William did, of course, land in East Sussex, but he chose his landing spot to draw out Harold, because Harold had lots of Sussex estates. The fact that he did not land in Kent, on the doorstep of the Archbishop of Canterbury, or in Dorset, hard under Winchester and the old Wessex royal heartland, quite possibly tells us something about the complicated politics of the Conquest. (To wit, that Winchester and Canterbury welcomed William. "I, for one. . .")

I'm not going to make some unfortunate, Delbruckian mistake and argue that the facts mean Pevensey can't possibly have been a pure defensive structure intended to dominate a possible landing spot. It has a purpose, after all. In the old days, this purpose was seen as the same as that of other oversized  "Saxon Shore forts," such as Portus Adurni, probably Portchester Castle. That is, they were either for defending the coastline from pirates, or for controlling the Saxons who already lived there by the early fifth century --it depends on how you read "Saxon Shore," or they were, given their size and architecture logistical nodes. "Fortified points for transport and supply between Britain and Gaul," to quote Wikipedia paraphrasing John Cotterill. Which was all very well until archaeology established for certain that Anderita/Pevensey was built by Carausius and Allectus.

Maybe Pevensey intended as a base for a counter-invasion? Perhaps. Only, what kind of stores? I am going ot hazard a guess that the internal space implies that those supplies included store cattle. It's pure speculation, but given that the Castle dominates a large salt marsh which would have been superb grazing, not necessarily an unreasonable one.

If Pevensey's waslls are so massive, what are these putative besiegers so upset about? Why do they want the cows inside so much?

This leads to the explanation that they are pirates, and pirates like loot. . . Sussex certainly has a rich history of piracy, though, unfortunately for the hypothesis, it's more of a source than a sink. 

We have three facts to hand, here. The first is that from round about the time of the Emperor Claudius' invasion until the middle of the Third Century, there was a "Classis Britannica," that is, "fleet of Britain," which we know about because it was involved in the production of roof tiles used for building in sites in London and the Weald --in the latter case being identified with iron-producing sites.  Second, we know that Carausius was originally assigned to hunt pirates who were troubling the shores of Roman western Europe,and got into trouble over it in exactly the way that every Admiral of every county of old England used to get into trouble over it. For someone tasked with stopping piracy, he sure ended up fencing off a lot of pirate loot! Third, a little later, there's somebody called the "Count of the Saxon Shore," who commands the above-mentiond forts. Whether the forts are garrisoned with Saxon foederati, or are intended to protect the coast from Saxon pirates, or both, the sources do not say, and nothing in the later history of Atlantic piracy suggests that the question will be easily resolved. 

So, thank God, some archaelogy:

Source: eds. Justine Bayley, David Crossley and Mathew Ponting. Author: It's a secret! Though probably not intentionally.

Roman iron-making in the Weald has a distinct geographical pattern. Whereas medieval and later activity concentrated in the north, the Roman industry is on the South High Weald and has East Sussex as a very decided centre. Our mysterious author even points out that the Roman road from Lewes (the county town of East Sussex) to London is partially paved with slag. This turns out to be no big deal. Most of the pre-modern roads in East Sussex are paved with slag. There's a lot of slag, and something has to be done with it.  More interestingly, Roman ironmaking in the Weald is concentrated in time as well as space, reaching readily detectable peaks in 100--160 and 200--250. The mid-century crisis of the Roman Empire is, among other things, a crisis of iron production in the Weald. 

There's obviously a great deal more to be said and sorted out here. Roman mining efforts slacked off noticeably in the middle of the third century. This has been presented as a question of monetary history: if something shut down bullion production, why, then, coins would have been harder to come by! I don't particularly like that explanation --or, rather, I tend to see it as effect rather than cause. Iron, though, is not bullion, and it is not mined from flooding underground mines. (Perhaps the mines were flooded by . . . climate change! Dunh dunh dunh.) It is produced from shallow diggings, with ore usually less limited than fuel.

Fuel! Charcoal, that is. Given the substantial kilns used in modern charcoal burning, I asked myself what the archaeological record might reveal. It turns out that some scholars are working on the subject: Robyn Veal is working on it as a question of environmental history. Archaeologists have also noted features where they found them, and Roman iron industry historian David Sim has opinions! Mainly, that we're not going to find kilns, or even less ambitious dedicated charcoal-producing features, such as have been suggested for the Wealden iron production sites in (Bayley, Crossley and Ponting, eds) above.

Sim's Roman charcoal burners wander a landscape of about six hundred square kilometers of coppiced East Sussex oak forest. Operating not more than about six kilometers from the dispersed  smelting sites, they use simple techniques: Cut, burn, turn over,  shovel the bottom layer into  baskets, and on you go.

He also at least invites a picture of conflict between pastoral users and charcoal makers. It is clear that the forests supporting the Roman industry were coppiced. This is true of that part of the medieval Wealden landscape where ironmaking was concentrated, as well. However, medieval coppic woods were enclosed by earthwork berms to keep livestock away from the tender shoots. No Roman earthworks exist, so at least up to the mid-Third Century cessation of iron-making the animals were being kept away from the coppice woods by other means. What these might have been, we can only speculate.

If only the Romans had built kilns! They did not, however. Fortunately, they did build salterns, so we have another access point to the past. As Alexandre Grandazzi now reminds us, the Romans knew the business of the up-and-down salt trade very well.* Between the hills and the Pontine Marshes, the routes by which Umbrian and Sabine cattle could reach the salt marshes at the mouth of the Tiber were canalised through Rome, which exploited that situation to take control of the salt marshes with a fort at Ostia. Medieval and early modern Pevensey was an important salt-making site. The Roman practice would have been to build a fort there, as much to levy a salt tax in kind as for any other reason.

No need to look out to sea, then, after all. This could be a business for East Sussex. Pevensey may very well look uphill, and its role may be as simple as to levy the salt tax. If the tax is paid in kind, then there you have an explanation for a walled enclosure --which we have known for a long time faced inland, and not out towards the water. That focus has been taken as an accident of survival, but perhaps it should be taken as diagnostic. The linked study of a saltern with "late Roman features" at the Stanford Wharf Nature Reserve in Essex shows the use of charcoal primarily from small, young roundwoods, that is, the product of coppiced, managed forests. This is in contrast to the saltern excavated at Morton Fen, Lincolnshire, where the fuel seems to have been scavenged from a wide range of waste sources.

1. In the deep Roman past, there was a "fleet of Britain," known to us less for its military achievements than for building iron foundries in east Sussex.

2. By the middle of the Third Century, this industry was supported by managed woodlands, where there was considerable potential for conflict with herd owners.

3. Abruptly, during the crisis, the iron-making business was suppressed.

4. By the end of the crisis, a "pirate-suppressing fleet" was active in the Channel, which mutinied and established a breakaway regime with broad popularity on both sides of the Channel, and which issued a very large quantity of very good bullion coins.

5. The restored Diocletianic regime, anomalous coin find evidence notwithstanding, established control of the mainland in 293.

6. The leader of the breakaway British regime was then assassinated and replaced by his treasurer.

7. A fort was built at Pevensey, perhaps to levy the salt tax. Given that the east Sussex woodlands were no longer being used by "the fleet" to make iorn, charcoal would have been available in large quantities to make salt, instead, so there is a priori reason to think that that the industry was growing.

8. The restored Imperial regime continued to operate Anderita, and may have built a number of similar sites on both sides of the Channel.

9. By the end of the century, this had become a region of special interest to the regime, such that the "Count" commanding this "Saxon Shore" was a notable late-Roman dignitary.

The archaeology of charcoal does not --yet-- lead to a very strong argument that it's all about grazing and salt for cattle, but the use of managed fuel sources, and the shift of charcoal away from iron to salt is at least very likely.  from other uses to saltmaking. On the other hand, Pevensey Castle is pretty clearly an extraordinary exercise, and  its very rapid construction suggests an intersection of fiscal and strategic exigencies. Maybe the breakaway British regime really, really needed the revenues from the salt tax? So what about the archaeology of Roman salt making? It turns out that there are known Roman salt-making sites in the protected archaeological area around Fishbourne in West Sussex, at Chidham and Thornham Boat Yard. The latest conservancy report ends with this puzzler:

"[The Fishbourne area] was also used for salt making in the Roman period. It is not clear why salt production seems to cease in the 2nd century, just when such production should be increasing."
If ithe puzzling cessation in saltmaking activity in West Sussex wasn't a century prior to my theorised beginnings in east Sussex, this post would be ending on a not of triumph, instead of puzzlement.

Now, as promised, the full version of the "coal statistics"

You'll notice that the wage comparison line is missing. It comes from here:

The suggestion that the industry is working marginal pits and using lots of overtime seems persuasive to me. The statistics are not going to tell us whether or not Jimmy Saville was a Bevan Boy or not, but it seems as though more of the "Boys" absented themselves from the pits than not, and no-one was looking for them very hard. (The fact that the Army wasn't, suggests that the selection process might not have sent the cream of Britain's youth into the mines in the first place.

Hmm. You have to wonder if some grizzled old recruiting sergeant saw something in Saville that the Beeb missed. No, wait. Actually, I don't wonder at all.

* Grandazzi, The Foundation of Rome: Myth and History Trans. Jane Marie Todd (Ithaca: Cornell University Press, 1999)
(70, 76-8) It's all down to the Lower Tiber’s primeval salt industry. So why did the herders ‘stop on their long travel from the hills” at Rome? It is a natural and necessary stop in the trip laid out by the routes of the Via Latina and Via Salaria, the route from hill to salt. 


  1. Off topic, and thanks for the response, but I think you'll like this lone data point:

  2. > "it seems that the mystery requires us to see the early settlers as speakers of many languages or at least Norse dialects, in order that they will require a koine to communicate with each other"

    I don't see how that follows. No doubt there were at least some Irish speakers around, and no doubt they did start out speaking different dialects of Norse - Norway is not a country that lends itself to dialectal homogeneity even today. But the fact that they managed to avoid dialect variation would be remarkable even if they had started out perfectly linguistically homogeneous. In the presence of barriers to travel, diversification is what we expect; the maintenance of homogeneity requires an explanation no less than homogenisation does.

  3. Yes: clearly by far the most interesting aspect of what's going on with Icelandic is that linguistic homogeneity is established and then maintained over several centuries during the high medieval period prior to the establishment of Norwegian-Danish sovereignity, and maintained without a state, or, by the saga-authorised narrative, church.

    That just doesn't tell us very much about Iceland's prehistory, the key pre-870 period which is still arguably not a period of Icelandic human history at all. Just like you warned me! I'm just salvaging what I can from the discussion by drawing the conclusion that Leonard's case works better, the more founding lingustic diversity is allowed.

    At the moment, a saga-authorised history of Iceland that incorporates what we actually know of early Norwegian history would suggest that the founding population came from a relatively limited part of even Norway, and that although many of these people had lived in much more exotic places before moving to Iceland, their connections still went back to the future Norwegian core territories around Bergen and Trondheim.