Still, I have my schtick, the "substructural history of strategy," and some actual evidence from the era that suggests itself for reinterpretation in the light of the failure of the financial markets in 2008..
So the grand explanation is this: Mounted warriors replaced foot soldiers at the "right of the line." The Seventeenth Century rolls over, cocks an eye, then goes back to sleep.
Field-Marshal Viscount Wolseley is not just a name you throw out when someone posits that before the great watershed of modernity, armies were dominated by cavalry generals. (Because he was an engineer, like most of the famous Nineteenth Century commanders-in-chief before the turn of the twentieth century. Someone should look into that, although I think Eric Dorn Brose has actually begun to do so.)
He's also the author of this:
The daily ration for all horses in the field is 12 lbs. of oats and 12 lbs. of hay. Horses employed on heavy draught work are allowed 2 lbs. of oats and 2 lbs. of hay extra [in the Crimea, 12 lbs. of barley and 16 lbs. of hay were substituted]. . .
In Turkey, where green forage was issued, 28 lbs. was given in lieu of 10 lbs. of hay or chopped straw. When no grain is to be had, the ration of hay should be 32 lbs., or 20 lbs. of unthreshed corn forms a good ration..." [Garnet Wolselely, Complete Soldier's Pocket Handbook, 86.]
So now you have a taste of Woseley's authorial voice. At one point in the pocket handbook, he explains how to conduct a fire fight. At another, he gives a good recipe for campfire plum pudding. And that's why people make fun of him.
But there's a cheesy excuse for this song directed at Wolseley. The character lampooned is a different Major-General, and that is telling, because back before Major-General was just a step on the promotion ladder, actual Major-Generals had to know how to solve equations quadratical . Why? Because a hundred horses will eat, in a single day, a solid acre's mow, but you can't just take them to the acre, cut down the grass, and turn them loose. Ideally, they should be fed continuously, in small quantities. Organising war means organising and controlling the landscape, and is generally a task for cavalry --and surveyors. So, ironically, enough, using cavalry in war means having even more cavalry to secure the means of subsistence for the cavalry, and the horses of the train.
Thanks to the revisionists who have written on the subject in recent years, and in particular to Robert Drews, we're getting a better picture of the impact and novelty of cavalry at the dawn of Classical warfare. On the one hand, Mediterranean armies appear to have had very small ratios of cavalry to infantry corresponding to a security detachment. On the other, at least by the time of the Carthaginian wars, both Romans and Carthaginians used larger forces of mercenary cavalry suitable for more ambitious operational uses. The Empire distinguished legions from auxiliaries that included many cavalry regiments, including many stationed in Britain. Still, our understanding of this distinction in Roman armies is evolving, and I wouldn't want to make too much of it.
What I would make a fuss about is ...economic geography. It would be nice if an historian studied this subject in detail, and it would also be nice if a horse person would talk about these things in a way that makes sense or is useful to non-horse people who do not speak the "money is no object" language of the hobbyist. Still, we have the situation at the outbreak of WWI as our pointer, and so we know that Russia had unlimited access to horses; that Germany, Austria-Hungary and Britain had adequate studs; that France was pushed to the limit in providing its armies with horses, and that of countries in the Mediterranean climate zone, only Turkey had a close-to-adequate stud; that India, in spite of being almost entirely unable to raise its own horses outside the Rohilla country (something else for a historian to investigate!), could field a substantial cavalry thanks to its ability to call upon the resources of Inner Eurasia.
Getting away from handwaves in the direction of something magical called "climate," it is easy enough to see that good horse-raising areas lack well-defined dry seasons and correlate this with the anyways perfectly well known productivity of Ireland, Northumbria, Lothian, Lower Germany, Lower Poland, Pannonia, the Anatolian plateau, and Inner Eurasia. This might be because of rainy and mild weather, heavy seasonal riverine flooding, or both. Either way, whether it is rainy Scottish weather or the annual flooding of the Volga, there is plenty of water available to replenish the grass. That these regions remained strategic assets in 1914 does not tell us that raising horses is expensive. That is not news. What is telling is that it was still so difficult for France in particular to raise enough horses. This, to me, suggests the limits of the power of a modern state even in its holy of holies, war-making. The bare implication is that as late as the dawn of the oil age, stud horse industries existed where they made sense for the farmers, not the strategists. As an extrapolation, the supply of skilled troopers is also likely to be geographically limited. You do not get good riders without putting horses together with humans.
Now assume that the same restraints fell upon the Romans. It was not always, necessarily, a severe restraint. Strategy is dialectic, and the Romans only needed enough cavalry to keep the field against their likely enemies. So the problem here would be, it seems to me the Persian menace, as while one might well find a strong, persistent cavalry menace across the European frontiers, but I am not aware of one at the moment. The Persians have a strong cavalry because they can draw on the Zagros stud and all Inner Eurasia beyond.
Where do the Romans get their cavalry? It does not seem that they make any particularly serious attempt to get at Inner Eurasian horses, so I am going to go with exactly those historic studs that the Romans actually controlled. If the deployment presented in the Notitia Dignitatum is correct as to numbers as opposed to nominal strengths, the cavalry is deployed in regions where it would be cheap to maintain it, just as one might expect it to be. (And this addresses the cynicism of those who fail to see the Scotch as such an existential menace as to require so large a garrison. The Roman Army is in Britain in such strength because it is so cheap to keep it there, just like the oversized Irish establishment of the old British army. It also succeeds in making Hadrian's Wall even more of a mystery than it already is. Why not incorporate Lothian in the Empire? How do we know that it didn't?)
So the problem is that you have your cavalry along the Tyne, where there is a distinct lack of Sassanian invaders. That's not so much of a problem, however, if you go with recent writers who see the situation along the Iraqi frontier as more one of recurrent Roman aggression than of a need for defence. (Will you buy that this is an all-too-brief summary of a recent H-Net review of a book that I can't recall at the moment? I'll fix this, but I'm running a little short of time right now.) Beginning at least with Trajan, the pattern is for a Roman emperor interested in making a victorious war against a worthwhile foe to assemble an army at one end of the empire and march it via the Rhine, Danube, Bithynia, and Cappadocia to the Euphrates valley. The route has ample forage along the way, as evidenced by its continuing use as late as the Crusades.
All this said, can I make a case for the route, and this strategic nexus of east and west intensifying as a result of the rise of the Sassanian state, and particularly the disaster of 260? I think so. The "crisis of the empire" that is seen as overlapping the local crisis entailed by the rise of the Sassanian state is now interpreted by some as a problem of patronage. (I'm going to credit that argument as soon as I remember where I got it from.) So as the Sassanians rise, at the same time we are seeing the distribution of the spoils of Imperial politics as an emergent crisis, resolved by the so-called Dominate, Domitian's new system of governance. And also a system of patronage that is robust against these end-to-end of Empire movements?
Here's the thing: what do you get if you try to tie together an empire that does not have a common economy? I'm going to say monetary problems. I'm going to see if the rise of African Red Slip Ware (and barrels, if and when we ever see Liquid Continent) and of hoard-deposition activity as evidence of this. Of what? That the Great Military Road is transmitting fiscal distortions through the state and producing fiscal crises of supply and demand locally. See? I said this was going to be more about the Zeitgeist than about Rome! The Failure of Lehmanus Fraterni and all that.
If such a crisis occurred, what might be the consequences? How about social transformation as local elites seek to make themselves opaque to the surveillance of the coercive state? That actually comes out of what is apparently a libertarian account of the fall of Rome, but it will also fit Wickham's interpretation without his insistence on autonomous collectives. Here I see the possibility of explaining the barbarian invasions not as a genuine demographic transition, but as a move in the politics of identity, even if I may be retracting these words after I've actually read Peter Heather's latest.
And, in the long run, if all war in the future is going to be unable to evade the need for cavalry, we are going to have to see the emergence of an agricultural economy that gives the farmer a stake in his horse stock. Conceded that there is a late Roman demographic transition, the need to feed all the horses presents one possible explanation for a period of below-replacement fertility. (Assuming that we can argue that population change rates are rationally chosen.)