Wild Mountain Thyme: Roman Fall, Byzantine Failure to Rise

It is going to be 30 degrees in Vancouver this weekend, time enough to think about the cool, high mountains of summer.

Oh, and in the course of an interminable thread over at Delong's, some guy, who probably doesn't deserve to be called out, given that I find the whole "the Byzantines had the wrong institutions" thing was riding high through the whole comment thread, proposes that since Turkey produces a lot of coal now, it couldn't have been a lack of coal that held the ancient east Romans back from industrialising. I'd say that runaway deflation is probably more important than coal supplies. In fact, I will say that. That will be my point. But, first, on the subject of coal,

Zonguldak to Eregli

On November 2, 1914, in one of the earliest of many indications that Turkey's declaration of war would beat out impressive competition from Russia  and Austria-Hungary to be the most foolish and tragic decision to join the clusterfuck that was WWI, a flotilla of Russian destroyers bombarded the coal port of Zonguldak before catching a small coastal convoy at sinking it. At that point, if not before, the Young Turks migth have noticed that Istanbul, and its fleet base, were dependent on sea coal brought from either the new town of Zonguldak or Eregli, the once-Heraclea Pontica. While there is a great deal of winnable coal in Turkey under modern logistics regimes, the coal mine at Kandilli, a town about 10 km north east of Eregli and about 40 km west of Zonguldak city, and 1 km from the coast of the Black Sea, was the only one then in use. Short of men, horses, industry and men as well as coal, the Ottoman Empire would not make a very impressive showing in the First World War. The question is, why?

No, it's not that Kandilli is a mountain top town serviced only by a funicular railway, making its deposits completely uneconomical before the Steam Age, silly. It's because reasons. I now introduce a guest speaker, John William Draper (1811--1882), writing in his History of the Intellectual Development of Europe, Although you'll hear that Byzantine history was neglected in the old days, you will be glad to hear that Draper covers the Byzantine inheritance of classical philosophy in a solid progression from narrative to climax. The narrative begins, more-or-less, with the final collaps of Justinian's attempt to "restore" the Roman Empire, which would be Phocas' coup against Maurice. This led to Chosroe's campaign of revenge against the east Roman Grand Guignol emperor (old timey allusion for the win!); which led to Heraclius' last-minute rescue of the east Rome 

Heraclius defeats Khosrau, Piero dell Francesca, "c. 1452"

and the resulting collapse of the Persian Empire; which led to the Islamic conquests, which resulted in the separation of eastern and western church; which resulted in. . . 

You can read it for yourself here, but I hope I'm not doing an injustice to the "thesis" when I summarise it as saying that it turns out that it's all about the Iconoclasm. After the failure of Leo the Isaurian, the true last Roman, Byzantium was in for a solid eight centuries of superstition. It turns out that drawing a picture of the Virgin Mary drives the science right out of your brain. Drawing Baby Jesus is like anti-matter to doing science's matter, except that instead of an enormous explosion, you get an irresistable compulsion to torture Galileo. 

If John William Draper's name does not ring a bell, perhaps it will help if I mention that it was during his vanity session at the 1860 Oxford meeting of the Association for the Advancement of Science that Bishop Wilberforce made some comment now lost to history, prompting Tom Huxley to drop that line about rather having a monkey for a grandfather than a bishop. (It's important because Draper wasn an American in England at  a time when they weren't yet everywhere, and because the American Civil War was incipient, and you can probably get an explanation of the connection between the having/not-having monkey ancestors and the Civil War from any given Trump fan.), 

If you're stil not saying, "Oh, that John William Draper," it might be because of fatigue with the whole thing that he stands behind. History of the Intellectual Development of Europe was a bestseller back in the day, but it was just a prelude to the sequel-rewrite, History of the Conflict of Religion and Science. Someone has got to be the poster-child for any good, interminable debate/rehash, and as far as the whole religion-stops-science thing goes, Ol' J.W. is the mutton-chopped poster child. 

If you're wondering how Draper gets to be at the AAS talking about "the intellectual development of Europe," it's because he is i): rich; and ii) actually has scientific street cred. J.W. showed up with his mother in Virginia at the age of 21, and is totally the same guy as the "John William Draper" who was a student for two years at Woodhouse Grove in London and the University College, London. I can't believe anyone would suggest differently. They just decided to move all the way from London to the Virginia backwoods when his Dad died, as people used to do in the 1830s. And then he married the "daughter of Daniel Gardner, a court physician to John VI of Portugal and Charlotte of Spain, who was born in Brazil after the royal family fled Portugal, to an unknown mother, and who was sent to live in London with her brother Daniel in "about 1830." (Thus Wikipedia.)

Anyway, in 1837, Draper was called to be head of chemistry at the medical school at the new New York University, although he actually ended up being a full professor at the New York University Medical School from 1840 to 1850. There, he did, again per Wikipedia, "important research in photochemistry," and "made portrait photography possible by his improvements (1839) on Louis Daguerre's process," and published a series of textbooks. 

It will probably not surprise anyone to discover that Draper only "made portrait photography possible" in the minds of family historians and patent lawyers. It will  probably also not surprise anyone that all his descendants became university professors except the ones who became fabulously wealthy merchant bankers, and that the House of Draper has since flourished exceedingly since. I haven't been able to confirm --because I am lazy and have not gone beyond Wikipedia-- that pioneering venture capitalist William Henry Draper, Jr. is the grandson of John William and not of some clan of drapers. It's just too precious to trace the role of American patent law in the multi-generational fortune of a family that originates with a man who made up tortuous explanations about how "intellectual development" goes wrong. 

Shorter: Unless you get your social institutions and culture just exactly right, you will start thinking wrong about science, and then you will never invent the steam engine ! Now where's my software patent royalties? I have a unicorn to invest in!

Funicular railways are things people build to make cities in mountains "legible." Yes, that's a strained allusion to Scott, and I wouldn't be doing it if I weren't coming back to The Art of Not Being Governed in a much more forceful way below. 

Spoleto, in the province of Perugia in the region of Umbria, is 20km south of Trevi, 29km north of Terni, at 396m elevation on the eastern branch of the Via Flaminia, which forks at Narnia (240m; seriously). 

In the Liber pontificalis, Spoleto appears as a constant threat to the Rome-Ravenna axis of communication, something that this map makes a bit more clear. It is a threat, and it is on one of the routes, but, at least as the view from 30,000ft has it, it is a great deal more important as a marketing town for the hills around it. 

first mentioned as a Latin colony of 241BC, it was a pretty obscure neck of the woods until some Lombards arrived in town in 572 and established the Duchy of Spoleto, one of the two southern Lombard duchies that historians of the early Middle Ages always mention in the context of the Lombard conquest of Italy without ever explaining further, probably because they're a bit hazy on the details, too.

   Constantine, Algeria, the principal town of eastern Algeria, with a population in 2016 of 448,000, is at 694m elevation. It is not so well treated on Wikipedia by the camera, although this is quite picturesque. (Credit where credit is due: I, Szaten's "bridges of Constantine" collection is great!)

Also:

Picture by Mariusz Kluzniak, flickr

Constantine, formerly Cirta, is a Phoenician establishment, but best known as the capital of the Numidian kings between the Third Punic War (149--146), through the Jugurthan War, until the removal of Juba with the purge of Pompey's successors. After that, it languishes in obscurity, never quite managing to attract a romantic prince nicknamed "the Falcon of This or That" to brood over its mountain fastness(ness?). So it's a surprise to learn that when the French arrived at the Algerian sea level and found ruins and desolation, the vast majority of the actually quite-large population of eastern Algeria were living up around Constantine. Meanwhile, Constantine's "natural" ports, Skikde where Constantine's river Rhumel breaks through to the sea, is built on the ruins of ancient Rusicade, but that town had been abandoned about a century afterValentinian and Valens built granaries there,Just east in al-Annaba, the town built on Augustine's Hippo, one finds the “land of the jujube tree,” where the rivers Seybouse and Boudjimah reach to the sea and create a coastal plain. Four thousand people lived within the walls of the neat, port city of Bône, but the French found only 25,000 inhabitants, and even they depended heavily on a tiony, isolated range of 1000 meter-high mountains. The French thought it odd, and typically "native," that there were no houses in the mountains, and made no connection with the intensively-managed cork oak forests that provided local exports and even acorn bread as a famine resort. This obliviousness helped when it came time to portion out the "vacant" land to Pieds noir. Unlike the mountains around Constantine, this coastal range could be safely appropriated, and was.*

Tripoli, Greece, is also not well served by panoramists, although this shot does at least give us the mountains. By Ddogas - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=5761912

Tripoli, Greece, formerly Drobolitsa, the place of oaks, and before that Tegea and Mantineia, is the smallest of the three cities, but not inconsiderable in the context of Greece's rapidly depopulating Peloponnese at 47,000. It is in the virtuous middle so far as elevation goes, hngher than Spoleto and lower than Constantine at 650m. Tegea, in particular, was a well-known, old Greek polis, contributing ships to the Trojan war. (That's quite the portage!) When I last visited Tegea/Tripoli, in honour of the abortive idea of a Grexit and a fine meal, I also linked to some Norwegian work that provides lonely support to the idea that Archaic Greeks were building in wattle-and-daub, and that might be why it is so hard to find masonry structures from that period --a claim that goes to the issue of population decline in the early Iron Age. This underlines what makes Greece special in this story. Its history is long enough to have two declines. 

Finally, I could talk about Toledo (529m), but Spain put its capital in nearby Madrid a long time ago, which makes Toledo look a bit less like a Late Antique anomaly. , 

So the basic point I want to make here is that there are major cities in the Mediterranean littoral at a strikingly high altitude, that their history can be obscure compared with the port metropoli of the coast, and that they emerged from obscurity with a vengeance in the late sixth century. (Or Spoleto did, anyway. The thesis isn't quite watertight.) The high altitude thing is worth more attention than it gets. California is often described as having a Mediterranean climate, and it has a lot of mountains, but the mountains around San Francisco and Los Angeles are basically all national forest. Where's the summer capital? Is there even a city in California above 2000ft with a population of more than 60,000? And, if so, is it a non-resort community?

The second point is about obscurity. Spoleto is the key examplar, although its historical trajectory in the late Antique is identical to that of Toledo: going from rarely-mentioned colony to the capital of a major power.  Even the identity of the state is in question. I notice that Spoleto had a "Byzantine" garrison commander, a "Balkan warlord" named Herodianus, who surrendered Spoleto to Totila's Goths in 543 and "became" a Goth. Depending on whether you agree with Patric Amory or his critics, this is either typical of the way that "identitiy was constructed" in Late Antique Italy, or a near-unique exception to poorly-documented norm. (You can see where I fall.) Since, apart from Procopius, Agathias, and Menander the Guardsman, our main source is the Early Medieval historian, Paul the Deacon, who is all about the "constructed" nature of the Lombard nation, we're pretty much hooped, here.  Paul of Monte Cassino got his start in life in the circles of Arechis II, Lombard Duke of Benevento, and there was no equivalent in Spoleto. We know, or "know" that Benevento was  established by a Lombard adventurer named  Zotto, who made his own way to Italy a few years before  Alboin's Lombard army. but Faroald, first Duke of Spoleto,  either arrived in Spoleto before Alboin or after, depending on the source you prefer. The point is: either Lombard invasion from outside, or Lombard ethnogenesis in situ. You take your choice as to whether the Duchy of Spoleto was an exogenous creation of Germanic conquerors, or a local reaction to changing circumstances, articulated in typically Antique terms as the consequence of a migration of nations.

With Tripoli, the situation is quite different. "Dobrolitsa" might be Slavic, and the Byzantine chroniclers are pretty clear that "Avars and Slavs" overran "the Balkans" at various times in the late Sixth Century. The Chronicle of Monemvasia, an odd but clearly important text, narrows the dates down much more precisely. Slavs and Avars overran Greece in 591, driving the inhabiting Greeks to flee to various places, including the newly founded coastal fortress of Monemvasia proper, where they could defend themselves more easily, and monpolise the export of Malmsey. On the other hand, the whole thing might be a thirteenth century fabrication intended to support the Bishop of Patras against his episcopal superior, Corinth. At the very least, we can say that the Emperor Maurice, who founded the Exarchate of Ravenna in 584, was by this time deeply embroiled in a war with the Slavs and Avars in the Balkans. This went successfully until, in 602, "dealing with a lack of money," he ordered his army to go into winter quarters north of the Danube. The army responded by mutinying under Phocas, which brings me back to the start of Draper's story. 

Decapitation porn-themed art not credited.

What seems clear enough in this version is that the Emperor Justinian has overdone it in his attempt to reconquer the Roman Empire. Pride goes before a fall, imperial overstretch before retrenchment. 

With historiography, you always seem to be dancing at the abyss, ready to plunge contemporary politics. Did I write "Justinian?" I meant to say something about the Afghan surve. I told David Sibley it would be a mistake. (At least, in my version of events.)

If that is the case, though, things get bewildering. Sending an army into winter quarters north of the Danube translates as sending it to winter in Moldova.   There's a reason you don't hear about hte ancient urban civilisation of Ukraine. That sounds a lot like desperation. In fact, there seems to be a trope of late Roman army detachments being told to go into winter quarters, because their payroll won't be forthcoming, and their responding by "going barbarian." Unfortuantely, I may be relying too heavily on Matthew Innes for that. 

Second, Phocas's Persian war is catastrophic, which is about what you expect in a situation like this. But, then, immediately afterwards, Heraclius arrives from Justinian-reconquered Africa and proceeds to defeat the Persians so solidly that their state is overthrown in its turn. Crisis? What crisis? Then, just before we proclaim Heraclius to be a world historical genius . . . 

. . . The Muslims arrive, and overrun everything. Because overstretch is back. Or possibly another overstretch? Then we need to jump forward through the whole Heraclian dynasty, and, above all, Constans II taking an army to Italy to fight the Lombards, visiting Rome (the only "eastern Roman emperor" to do so), and then removing to Syracuse to rule the Byzantine Empire from Sicily for five years. Certainly you can read this as Constans getting out of Istanbul because it was just that bad, but this isn't exactly an obvious interpretation of events. 

Next, we get to the"Twenty Years' Anarchy." Just as Maurice had been the last of Justinian's dynasty, and his overthrow had triggered catastrophe, so the overthrow of the last Heraclian, Justinian II (don't worry, he got better) led to more . . . I would like to say "catastrophe," but apart from the rapid succession of emperors, there are not the outward signs of total derangement that followed Maurice's death. Leo the Isaurian seems to have done a good job of rebuilding the empire. Draper is neither the first nor the last modern historian to make him a personal hero, and so perhaps we owe it to Leo to paint the "anarchy" he ended in the most lurid terms. Unfortunately, we "owe" it to him for the same reason that Draper found him so attractive. Leo is the original Iconoclast protagonist of the century-and-a-half-long Iconomachy, when the Byzantines had a truly Byzantine fight over religious images. Or something, the problem being that the "Iconophiles" won, and then rewrote the history of the period, and we are entirely dependent on these revisionist Iconophiles of the 900s for the history of not only the 730-842 period, but for  everything from about 600 on. This failure of the Classical/Latin/Roman/Byzantine historical tradition is comparable to the one in that spans the Third Century Crisis, and probably needs explanation. 

So do we move the "crisis" back to the early 600s, and so explain why Agathias can continue Procopius, and Menander, Agathias, silence falls when the pen of Maurice's bodyguard is stilled? No, because as weak as my framework is, I need the abandment of Carthage in favour of Tunis in 698. Tunis (and Kairouan) are not, you may have noticed, Constantine. The medina of Tunis is only 15km from the ruins of old Carthage, although there is a 41m elevation gain along the way.  It is still a huge move inland and towards the plain, if not the mountains.

So. A century of crisis? Matthew Innes has a plausible interpretation --essentially Pirenne with the objectionable bits removed, but I just want to digress here.

i) Emperor Maurice was thr first reigning emperor to have a son "born to the purple" since Theodosius II in 401. This is significant, since in one theory of hereditary succession, only sons born during the reign count as heirs.

ii)  Theodosius was married by his father to an otherwise anonymous daughter of one Germanus.

iii) This Germanus was probably the cousin of Justinian who married Matasuntha, granddaughter of Thederic, the Ostrogothic king of Italy. Per Jordanes (as interpreted by Theodor Mommsen), Germanus was the grandson of Anicia Juliana. 

iv) Anicia Juliana, apart from being the woman who erected St. Polyeuctus, the church that Justinian built the Hagia Sophia to overshadow, was the younger daughter of Emperor Valentinian III.

v) Valentinian III was the son of Galla Placidia.

vi) Galla Placidia was the daughter of Thedosius I, by his second wife, Galla, for whom he set aside the mothers of his sons for.

vii) Which he did because Galla was the daughter of  Valentinian I by his second wife, Justina.

vii) Justina's antecedents are unclear, but she is descended through her mother from Constantius and Helena. 

viii) Although the founder of the Constantine dynasty needs no introduction here, he might be descended, taking dynastic propaganda at face value, from an earlier emperor, Claudius Gothicus. If so, it is through his mother. 

Have you read Frank Herbert's The Godmakers? I'm not exactly going to recommend the incoherent little "fixup," but I am going to point you to the third chapter, originally published as "Operation Haystack" in the May, 1959 issue of Astounding Science Fiction. Or if you haven't the time, spoilers ahead:

Lewis Orne, the chief character in the stories, is an operative of the I-A (Investigation and Adjustment Agency) of a galactic government, which is rebuilding civilization after a disastrous war in which all contact between planets has been lost. His job is to watch for the embers of war on rediscovered planets and, if necessary, call in force to stamp them out. In the course of his work, Orne has to rely heavily on intuition and on heightened perception of the nonverbal indicators of a warlike mentality. In one story, "Operation Haystack," he ferrets out a secret society of women, the remnants of a once-proud race who lost power in the Rim Wars. These Nathians landed on the planets of their enemies, and, using special techniques to breed only women, embarked on a devious, 500-year plan to gain ascendancy by masterminding the political careers of their husbands. Orne discovers that he himself is one of the Nathians, the rare male child they have allowed, intending him to play a special role in their plan. Like Paul, however, he has escaped from their control, the random factor that becomes the ruin of their schemes.

 Given the stereotypes of clannish Balkan politics, I'm a little surprised that no-one's ever pointed this out and imagined a Clan Constantine family stronghold somewhere in modern Serbia, with a council of clan elders orchestrating generations of political marriages; and to this cloud cuckooland fantasia, I will pile on the chief tenant, a bailiff who traces his ancestry back to the "Goths" that Claudius Gothicus settled on the land so long ago --and to the father of Theoderic, too. I know, I know. There's no reason not to think of the descent of the Constantinians as thoroughly deracinated, cosmopolitan elites: but if, instead, we allow them to be rooted in their ancestral estates, the decision to send the unsupplied army across the Danube to fend for themselves in the winter of 602  makes sense in a way that would thoroughly enrage me as a soldier in that army. Bring on Phocas! 

Fantasies inspired by bad science fiction novels aside, Innes proposes a three-stage collapse of the Late Antique economy. It begins with it functioning "normally" in the late 300s. The annona, a state-subsidised redistribution of grain and other necessities, forms the backbone of all long distance trade. It is, after all, subsidised transport, and tax-exempt. The annona that fed the population of Rome at its peak. This is plausibly estimated by various specialists as between 450,000 and 680,000,** and came largely from North Africa. This makes the Vandal conquest of North Africa a crisis for the Empire, albeit not a well-attested one, especially given the continuing prevalence of North African Red-Slip Ware around the Mediterranean basin.

Innes resolves this point by first pointing out that the annona that provided for Istanbul and other eastern cities was derived from Egypt and Syria, not North Africa. The Vandal kingdom continnued to export various goods, including grain to Rome, and the Roman subsistence crisis is put off to the era of the Gothic Wars, which is assumed to have been a real and total social crisis in Italy, on the basis that Procopius knew what he was talking about, and wasn't just creating an elaborate justification for the reign of Germanus that never happened. The real crisis that matters, then, is the one that follows the cutting-off of the eastern annona, and this happens with the fall of Egypt to the Persians in 615, and not the Islamic conquest. Our lack of imperial historians after 600, then, is the consequence of a gradual (or rapid!) emptying-out of the Imperial capital. Throw in an epidemic and your imperial overstretch, and you have a crisis that begins in 615, and which gradually plays itself out over the next [number omitted] years. There is also some interesting archaeology taken to indicate that Roman aristocrcatic households are shifting their luxury consumption from imported to local goods in the late 600s, but I notice that Innes' account depends heavily on finds of glass shards, and the idea that these are spoils for tempering local production imight be past its vogue in British archaeology, so I am not sure how much to make of it. 

So much, then, for literary evidence mixed with archaeology old enough to make it into survey texts. Does Bench Grass have anything to offer? What does upland and dryland pasture have to do with the fall of the Roman Empire and the Byzantine failure-to-rise?

Lameen calls our attention to Elizabeth Fentress's work. I'll throw in some more new work, Francesco Carrer on "Herding Strategies, Dairy Economy and Seasonal Sites in the Southern Alps: Ethnoarchaeological Inferences and Archaeological Implications."(***)

Fentress sees a North African crisis of the late 400s, predating the larger Mediterranean one, and preparing the way for the Vandal Conquest. This "Moorish" incursion is to be explained, she thinks, by a movement from the deep Sahara into the North African tell. For her, this movement is mediated by the Garamantan kingdom of the Fezzan. As precious western practitioners of foggara agriculture, the Garamantans are necessarily intermediaries in long-distance Saharan trade: foggara agriculture is not self-sufficient. The old theory sees the Fezzan already performing its later role in the trans-Saharan trade with tropical Africa, and this theory has always struggled with the lack of Roman goods south of the desert. It seems more plausible to me to see the trade as infra-Saharan, and to suggest that the Garamantans were drawing on less archaeologically-visible cultures of the desert wetlands which preceded the foggara oases of later times. 

In that case we have . . . Well, we have something. Something that is driving a Moorish "invasion." An ecological/agronomic crisis is possible. Perhaps the water was drying up. I'm a little skeptical: "fossil groundwater" is like "deforestation," "overfishing" and "eroding topsoil"-- real things that actually happen, but which also increase the value of existing stocks to the beneficial owners, who are often the ones most vehemently certain that the crisis is actually happening. On the other hand, the trading system is very definitely collapsing. 

Francesco Carrer may have an explanation:

This paper investigates one of the main issues of the archaeology of pastoralism in the southern Alps: the decrease of upland archaeological evidence in the late Iron Age (IA) and the Roman period. A novel interpretation is proposed using ethnoarchaeological, ethnographic and ethnohistoric data collected in different areas of the French and Italian Alps. These data document the existence of two pastoral strategies: Alpine pastoralism (focused on dairying activity) and transhumance (mainly focused on the exploitation of wool and primary products of livestock). The sites related to the first strategy are usually permanent and complex, while those related to the second are simple and ephemeral. On the basis of this information, the decrease of upland archaeological evidence during the late IA and the Roman period has been interpreted as a consequence of the decrease of dairy activity and increase of transhumance. These inferences allow us to rethink the evolution of human exploitation of high mountain environments, as well as to tackle the bias of archaeological invisibility in specific pastoral contexts. Furthermore, they also confirm the important role of ethnoarchaeology as a method of improving archaeological interpretation

I hope he doesn't mind if I scrape his nice pictures: 

Here's a more modern summer cottage in the cool, high mountains. It's a bit of a fixer-upper, but with a little elbow grease. . . :

"Ethnoarchaeology" in this case means that Francesco Carrer is arguing that they did things like they do today, back in the old days. Specifically, an enclosure is consistent with a dairy-based subsistence strategy, in which the young folk of the community take the herds up from mid-level winter pasture around their mid-level communities, and turn the high alpine grass into cheese and butter for winter consumption. 

The lack of enclosures datable to the Late Iron/Roman Age (400BC--c. 400AD) is not consistent with botanical evidence of continuing human use of the high pastures, leading Carrer to propose a livestock-centred transhumance strategy. Animals are driven up from low altitude wetlands to high altitude pasture, then brought down in the winter, and the increase of the herds sold. Although he uses evidence for the Roman wool trade in northeastern Italy to argue for this pattern during the Empire, so things are a bit more complicated than that.

The key point here is that the transhumance strategy is market oriented, while the dairying strategy is subsistence-oriented. .

This is a very cheesy post.

It is also a "legible" enterprise. As Pseudo-Aristotle tells us, while kings and cities make money from the business of coins (including excises and sales taxes), satraps make their returns from land and cattle, and the way to tax the increase of the pastures is very simple: roadblocks on the path of the migrating herds take the royal portion of the increase. Local surveys in Britain around the Pennines show a baffling number of  small fortresses: those would be your roadblocks, and that would be how the legions support themselves. 

That is, until the cows stop coming down the hill, because the herders are not getting a good price on the market, and your army stops being able to pay for itself.

*So I found a monograph: a little knowledge is a dangerous thing, but the (European-language) historiography of North Africa does strike me as a little thin. David Prochaska, Making Algeria French: Colonialism in Bône, 1870—1920 (Cambridge: Cambridge University Press, 2004).

**Specifically, Nicholas Morley, Although the argument tends to turn around the insulae, ancient Roman apartment buildings that sound an awful lot like late Nineteenth Century Brooklyn apartment blocks, the key point of departure seems to me to be the average population density. Remember that at a mere 150,000, ancient Rome's population would have been "only" 10,000 per square kilometer.

***Elizabeth Fentress's paper is a colloqium contribution that has been uploaded to Academia.Elizabeth Fentress and Andrew Wilson, 'The Saharan Berber diaspora and the southern frontiers of Vandal and Byzantine North Africa', J. Conant and S. Stevens (eds),North Africa under Byzantium and Early Islam, ca. 500 – ca. 800 (Dumbarton Oaks Byzantine Symposia and Colloquia. Washington, D.C.)Carrer is in Journal of Mediterranean Archaeology 28.1 (2015) 3-22. The article is online, but it is in one of those new pdf formats that's hard to link to without auto-downloading.