We have to start not with tactics or the physics of shouting orders, but with cultural expectations. First, we need to establish some foundations here. First, in a pre-modern battle (arguably in any battle) morale is the most critical element of the battle; battles are not won by killing all of the enemies, but by making the enemies run away. They are thus won and lost in the minds of the soldiers (whose minds are, of course, heavily influenced by the likelihood that they will be killed or the battle lost, which is why all of the tactics still matter). Second, and we’ve actually discussed this before, it is important to remember that the average soldier in the army likely has no idea if the plan of battle is good or not or even if the battle is going well or not; he cannot see those things because his vision is likely blocked by all of his fellow soldiers all around him and because (as discussed last time) the battlefield is so large that even with unobstructed vision it would be hard to get a sense of it.

So instead of assessing a battle plan – which they cannot observe – soldiers tend to assess battle commanders. And they are going to assess commanders not against abstract first principles (nor can they just check their character sheet to see how many “stars” they have next to “command”), but against their idea of what a “good general” looks like. And that idea is – as we’re about to demonstrate – going to be pretty dependent on their culture because different cultures import very different assumptions about war. As I noted back in the Helm’s Deep series, “an American general who slaughtered a goat in front of his army before battle would not reassure his men; a Greek general who failed to do so might well panic them.” An extreme example to be sure, but not an absurd one. In essence then, a general who does the things his culture expects from him is effectively performing leadership as we’ve defined it above.

But the inverse of this expectation held by the soldiers is that generals are not generally free to command however they’d like, even if they wanted to (though of course most generals are going to have the same culturally embedded sense of what good generalship is as their soldiers). Precisely because a general knows his soldiers are watching him for signs that he is their idea of a “good general”, the general is under pressure to perform generalship, whatever that may look like in this cultural context. That is going to be particularly true because almost all of the common models of generalship demand that the general be conspicuous, be available to be seen and observed by his soldiers. As a result, cultural ideals are going to heavily constrain what the general can do on the battlefield, especially if they demand that the general engage personally in combat.

Different sorts of generals

We can actually get a sense of a good part of the range simply by detailing the different expectations for generalship in ancient Greek, Macedonian and Roman societies and how they evolved (which has the added benefit of sticking within my area of expertise!).

On one end, we have what we might call the “warrior-hero general”. This is, for instance, the style of leadership that shows up in Homer (particularly in the Iliad), but this model is common more broadly. For Homer, the leaders were among the promachoi – “fore-fighters”, who fought in the front ranks or even beyond them, skirmishing with the enemy in the space between their formations (which makes more sense, spatially, if you imagine Homeric armies mostly engaging in longer range missile exchanges in pitched battle like many “first system” armies).

The idea here is not (as with the heroes of Homer) that the warrior-hero general simply defeats the army on his own, but rather that he is motivating his soldiers by his own conspicuous bravery, “leading by example”. This kind of leadership, of course, isn’t limited to just Homer; you may recall Bertran de Born praising it as well:

And I am as well pleased by a lord
when he is first in the attack,
armed, upon his horse, unafraid,
so he makes his men take heart
by his own brave lordliness.

On the opposite end of the spectrum, there is the pure “general as commander” ideal, where the commanding general (who may have subordinates, of course, who may even in later armies have “general” in the name of their rank) is expected to stay well clear of the actual fighting and instead be a coordinating figure. This style […] is fairly rare in the pre-gunpowder era, but becomes common afterwards. Because in this model the general’s role is seen primarily in terms of coordinating various independently maneuvering elements of an army; a general that is “stuck in” personally cannot do this effectively. And it may seem strange, but violating these norms with excessive bravery can provoke a negative response in the army; confederate general Robert E. Lee attempted to advance with an attack by the Texas Brigade at the Battle of the Wilderness (May 6, 1864) only to have his own soldiers refuse to advance until he retired to a more protected position. Of course this sort of pure coordination model is common in tactical video games which only infrequently put the player-as-general on the battlefield (or even if the “general” of the army is represented on the battlefield, the survival of that figure is in no way connected to the player’s ability to coordinate the army).

In practice, pre-modern (which is to say, pre-gunpowder) generals almost never adopt this pure coordination model of generalship. The issue here is that effective control of a gunpowder army both demands and allows for a lot more coordination. Because units are not in melee contact, engagements are less decisive (units advance, receive fire, break, fall back and then often reform to advance again; by contrast a formation defeated in a shock engagement tends not to reform because it is chased by the troops that defeated it), giving more space for units to maneuver in substantially longer battles. Moreover, units under fire can maneuver, whereas units in shock generally cannot, which is to say that a formation receiving musket or artillery fire can still be controlled and moved about the field, but a unit receiving sword strikes is largely beyond effective command except for “retreat!”

In between these two extremes sits variations on what Wheeler terms a “battle manager”, which is a bit more complex and we’ll return to it in a moment.

What I want to note here is that these expectations are going to impact where the general is on the battlefield and thus what he can do to exert command. A general in a culture which expects its leaders to be at the front leading the army has the advantage of being seen by at least some of his soldiers (indeed that is the point – they need to see him performing heroic leadership), but once engaged, he cannot go anywhere or command anyone. This is also true, by the by, in cultures where the general is expected to be on foot to show that they share in the difficulties and dangers of the infantry; this is fairly rare but for much of the Archaic and Classical periods, this was expected of Greek generals. Even if a general on foot isn’t in combat directly, their ability to see or move about the battlefield is going to be extremely limited.

On the flipside, a general who is following the “commander” ideal is likely to be in the rear, perhaps in an elevated position for observation. The obvious limitation here is that such a commander is going to struggle to display leadership because no one can see them (everyone is facing towards the enemy, after all). But that also impacts their ability to command – no one is looking at them so if they want to change their plans on the fly they need to send word somehow to subordinate officers who are with or in front of the battle line who can then use their visibility to communicate those orders to the troops.

Bret Devereaux, “Collections: Total Generalship: Commanding Pre-Modern Armies, Part II: Commands”, A Collection of Unmitigated Pedantry, 2022-06-03.