On Reserves

I’ve recently joined the MiniatureRulesDesign Yahoo! Group, and stumbled into the midst of a lively discussion about the keeping of reserves, and why wargamers do not do so.

A variety of opinions have been presented, along with suggested rule changes that would ensure that wargamers would keep reserves. For example, it was suggested that a reserve must be designated and that it would not be available to be used before turn X of the game. All well and good, I agree, except that reserves were used when they were needed. I mean, can you imagine an aide-de-camp riding up to Napoleon and saying ‘Sorry, Emperor, the Imperial Guard won’t be available for another hour…’

Now, the fact is that wargamers do not usually keep reserves, and I suspect that this is partly due to rules not rewarding the use thereof, and partly due to the ‘6 year old soccer syndrome’.

Try this experiment. Go out to a sports field and watch a group of 6 year old boys play soccer (football). I suspect that you will find the 2 goalkeepers by their posts, and the other 20 players all running after the ball. I certain circumstances, they can look rather like a comet, with a small head and a long streaming tail.

Now watch a game of professional football. The players are spread out all over the ground, often in a recognisable formation. They will be in two lines (given a 4-4-2 formation) and while these may bend around the field, you will usually see them reform in fairly quick order. We might ask the question ‘why to professionals play like this when the 6 year olds do not?’

We might also ask the question ‘what has this got to do with wargame rules?’ but bear with me.

The junior football team is an ‘all hands to the fight’ type of thing. The whole idea of most of the players is to get hold of the ball and score; in a sense it is the last bastion of heroism available to many of us in our world today. The professionals, however, have a different approach. They aim to interact as a team, to outmanoeuvre the opposition, to conserve as much energy as possible and, if you watch a really good team, attack and counterattack as such speed that the opposition cannot react in time.

I suspect that something like this is happening in our wargames. We want to get all hands to the fight, and our rules encourage this. There is little point in having significant reserves at anything but a tactical level. My first line unit may get beaten, routed, removed from play, whatever, but all I really need is another to plug the gap. This is not a grand tactical reserve, but a local one.

I think that this comes about because of a number of factors. Firstly, our units last too long in combat. There are few rules that I know of which force units back and out of the fight through exhaustion, having been in combat for too long. The psychology of men in combat is not particularly well understood, but it seems to me to indicate that individuals, and hence units, are quickly drained by combat. Bases of toy soldiers seem to carry on for ages.

Secondly, it is possible that our moves, for units in combat, are too long, and for those not in combat are too short, or, possibly, the other way around. Someone pointed out that it took the Confederate units 3 hours to assemble for Pickett’s charge at Gettysburg, but 20 minutes to cover the same distance in the charge itself. This seems to suggest different move rates for different functions for the same units. I suppose that drawing the units up, dressing the lines and so on takes a lot longer than we suppose.

That said, consider Marlborough’s battles. In quite a few of them he shifted significant forces across the battlefield and into action to wrong-foot his opponent and score stunning victories. This suggests fast movement out of combat for the reserves.

Furthermore, not all armies did keep reserves, and even if they did they were not necessarily useful. At Naseby, the New Model Army had no real reserve. I assume that Fairfax and his colleagues reckoned that with a near 2:1 advantage in troops, they could draw on troops from the second line where the fighting was less intense if they needed to, as indeed happened.

The Royalists, however, did have a fairly large reserve, but it did need to be committed early to try to shore up the line on the left to allow the right to continue its success. It failed, and the result is, as you might say, history.

As a last point, I’d also note some of the battles on the Western Front in World War 1. The first waves were often successful, but by the time they had seized their objectives were spent. The second wave usually ran into heavy artillery barrages as they tried to move forward, and so the exhausted first wave were easily rolled back by fresh enemy troops. The point is that the plans called for the first wave simply to achieve a single objective; after that it was recognised that they would be unable to even hold what they had gained.

Which brings me back to my original point. Do we just allow units in wargames to fight on for too long? I don’t think, certainly in pre-WW1 days, that the actual casualty rate would cause the units to become ‘spent’, but the sheer terror of close fighting could well have this effect.

If our units were not so enthusiastic as to fight for prolonged periods without relief, perhaps we, as wargamers, would be keener on having reserves. But there is still the more hands to the fight argument to worry about; numbers, in the end (c.f. Naseby again) count.

Models and Representations

Just to prove that I don’t write in an entire vacuum, I saw recently a debate on the ancmed yahoo group about modelling, in which it was asserted that someone had said that we do not have models in wargaming, but wargames are simply representations of ancient battles.

My problem is that I’m not sure I know what that means.

Now, obviously I have some idea of what a model is. A model is, in some way, a metaphor for a real life event. The Bohr theory of the atom is a metaphor – it does not represent a real thing, but it enables us to do some calculations and get some answers which we can relate to the real world. However useful it is for that, and for teaching people about atomic theory, it is not an atom, or a representation of an atom; it makes no ontological claims.

So, what is a representation?

At one level, of course, our toy soldiers themselves are representations. An Assyrian warrior (at least when painted by a better hand than mine), is a representation of a seventh century BC warrior. It is a scale model, in terms of what I have said before. But I do not think that this is what was meant by the comment.

I suspect (although I do not know), that what is meant is that an ancients wargame vaguely resembles an ancient battle, and no more than vaguely. This is not a scale model, in any way, shape, or form, but perhaps what is meant is that the wargame is a poem or a play set around the original battle; not accurate in its form or activity, but giving a flavour of the action.

I think this is for a variety of reasons. Firstly, the sources are not really adequate to a reconstruction. I’ve moaned before about the inadequacies of the accounts we have of both Marathon and Issus, and the situation for most other battles of the period is no better. Even if we do have more than one account of a battle, usually they are contradictory, at least in part. Of course, if they agree in every detail, they are presumed to derive, one from the other. There is no pleasing us, sometimes.

Secondly, of course, we cannot reduce all the reality to the table top. The smaller details of terrain that affect the action, the overall flow and interlocking of the decision making and so on are not fully modellable. When this is added to the problem of the lack of knowledge of the action already referred to, I think we can see that there is only a small chance of really simulating the battle.

Is a representation, a flavour of a battle, all we can aspire to, then? I suspect that here things get a little complex.

If, as I have surmised, a ‘representation’ of an ancient battle is akin to saying that the wargame is like a metaphor or play or poem of the real thing, then all this is really saying is that it is a model of the original, but in a different form.

Let me try to explain.

Poems and plays (and, for that matter, literary novels) rely on metaphors for their effect. In fact, our entire language, or at least large chunks of it rely on metaphor to work, even though we rarely notice the fact. For example, to say ‘I understand’ is easily understood (ahem), but actually means ‘I stand under’, which is a metaphor for knowing intimately. This is not just the case in English, incidentally.

Now, a poem or play is a set of metaphors to describe something, in terms of something else. That is what metaphors do. Thus, we can say that even a poem or play is a model of some event, because models are extended metaphors, or sets of metaphors which describe something in the world (like an atom) in terms of something else (a solar system, in the case of the Bohr atom).

Therefore, it would seem that a representation of a battle, no matter how vague, is in fact nothing more than another model of the battle, but expressed in different terms. I’m not sure, on that basis, if the person who tried the draw the distinction was actually making any helpful progress towards understanding how wargames function.

As I’ve tried to hint above, the problem with this sort of discussion is that it all turns on our understanding of language and meaning. In spite of Shannon’s work over 50 years ago, and Wittgenstein’s arguments a little bit later, we still have this idea that language, the words we use are, or at least should be, transparent. It would be nice (but probably boring) if that were the case.

Language is a little bit like the three laws of thermodynamics. The simplified version of those reads:

1. You can’t win.
2. You can’t break even.
3. You can’t get out of the game.

Language is a bit like that. We can never be entirely precise – that sort of precision is the remit of logic and mathematics, so we cannot win. We can never even be sufficiently good at language that we do not need metaphors, as the language is metaphors in large part, so we cannot break even. Finally, we need language, as we cannot communicate otherwise, so we cannot get out of the game.

Language, therefore, enables us to speak, even if what we speak about are ‘only’ models. In fact, when taken seriously, much of what we do is done via model and metaphor, and these are representations of reality, seeing a ‘this’ as a ‘that’. The only alternative, as both Wittgenstein and St Augustine acknowledge, is silence.

Painting and Psychology

Just to prove that I do some normal wargamer type things, as well as apply Wittgenstein to the hobby, I have been painting so toy soldiers recently. Greek hoplites, as it happens, some of Mr Berry’s finest. It has raised a few questions for me, which I thought I’d share (or witter on about anyway), at least as a change from talking about models and metaphors.

Although toy soldiers are, of course, scale models.

I suppose that I’m not alone in struggling with painting. After all, if painting was easy, none of us would have those piles of unpainted lead that any brief surf around the internet can find many wargamers bemoaning. I suspect that, concealed in most wargamer’s cupboards there is at least one unpainted army.

Neither my inner critic, nor my external ones, is happy with me playing with what we might call ‘grey armies’. Ever since I got my first proper wargame army, 15 mm Peter Laing ECW, as it happens, at a Christmas long, long ago, using grey, unpainted, armies has been frowned upon. Mind you, I did rope my sister in to helping me paint them then.

This year, Santa delivered again, and I am now the proud possessor of a 15 mm ground scale, 6 mm DBA Classical Greek army. For those of you for whom that is gibberish, the idea is to crowd as many 6 mm figures as possible onto a DBA base, so even a 12 base force looks like a proper army; try it, it works.

Now, I’m not a good painter, I know that, and I’m also slow. So I have to go with my line of least resistance, which means for me a 40 mm base frontage and the minimal number of figures per base I can get away with, in my case, for Greek hoplites, that means 8, or 2 strips of toy soldiers.

This is not the strongest selling point of 6 mm figures, I have to admit. I did try putting four strips on a base, once, for my Pontic pikemen. I confess, they do look quite nice, but painting that many figures for that few bases nearly killed me. You can see them in the middle of the photograph.

Painting, for me at least, is thus a very psychological process. I look at a set of 20 strips of unpainted soldiers on lollipop sticks and get depressed. In fact, with the Greeks I only managed 10 strips, or 5 bases, which is a quarter of the army I want to produce. That in itself is a bit demoralising, so I struggle to find the motivation to actually do any painting at all.

The idea of painting 6 mm figures is a little different from other figures. It is impossible that someone would pick up and examine an individual figure for its correct cuff colour, which can be done with a 15 mm or 25 mm figure. So the painting to achieve an impression of rightness, not rightness itself.

This is supposed to be achieved, according to the Baccus painting guide, by dry brushing. Now, I’m sure that this is very successful for some people, but I do not seem to be one of them. It may just be that I’m too heavy handed, or too light handed, I’m not sure, but I can never get the coverage right. I land up just painting, because that is all I can manage.

Another problem I have is with detail. Some people manage to paint the white stripe around the tricorn hats of 6 mm eighteenth century soldiers. If I try that, I get white hats. It is a similar issue with shield designs on ancients figures.

In this case, I think I have a good excuse. I’m an asthmatic, and the drugs used to treat asthma these days actually impart a slight tremble to your hands (and, indeed, other muscles). No one believed me when I explained this, but then I met someone who was a GP who said ‘Yes, that is a well-known side effect’. So, that is my excuse, and I’m sticking to it.

I suppose that all this is working up to the question of what our toy soldiers really are. In some cases, mine for example, they are counters in a game. I like my counters to be representative and nicely presented, and I do the best I can, but they are still counters.

For other people, they are miniature works of art, beautifully painted and based and even, with 6 mm folk, based as little dioramas with skirmishers to the front, officers conferring and scrubby bushes all on the base. Wonderful, but very labour intensive, I should imagine. And it must be a wee bit frustrating if such a work of art is removed from the fray by a random cannon shot first move.

So, for all the pondering the imponderables of wargaming, I’m not very good at painting; indirectly, of course, that explains the lack of ‘eye candy’ on this blog, which may itself explain the relative lack of popularity of the blog compared to some out there. It is a good excuse, anyway, and means that my lack of marketing of it can be quietly left to one side….

Game Mechanics, Models and Metaphors

One of the things which most opinion is agreed upon is that metaphors are irreducible. By this is meant that a metaphor cannot be reduced to some ordinary language.

For example, the metaphor ‘war is a game of chess’ cannot be reduced to something like ‘war is a thing with two sides and different pieces with different capabilities lined up against each other…’

The attempt at reducing the metaphor to ordinary language (if there is such a thing) does not capture the resonances, links and differences between war and chess in the same way that the metaphor itself does. Metaphor is not simply a rhetorical flourish; meaning is added.

If this is the case for metaphor, it must also be the case for models, only more so. I mentioned before that multiple models are useful for capturing the full picture of a real thing or process.

The most obvious example is in physics, with wave-particle duality. One way to model photons is to describe them as waves; another way to model them is to describe them as particles. Which is correct? Actually, the question is meaningless – photons are both, and neither.

Modelling photons in these different ways helps us to understand, to predict physical behaviour, but does not give us an ontological commitment to them being waves or particles. The underlying reality is closed off to us and we cannot tell what a photon ‘really’ is.

Now, consider a set of wargame rules. They will need, say, some system of command and control, otherwise the toy soldiers will simply respond to every whim of a wargame general’s mood. The command and control mechanism, be it DBA PiPs, or Polemos Tempo points, or written orders or whatever, is a model. It is also a game mechanic.

Now, someone had a mild go at the PM: SPQR tempo points command system a while ago. As I understand the point, it was that the tempo points system is just a game mechanic and that it penalised unnecessarily ‘unformed’ troops. Now, in PM: SPQR terms, unformed troops are those not formed in ranks and with less clear command structures. The person in question argued that Germans, in the Roman period could have been ‘formed’, we just do not know. The rules, however, penalise them for being unformed.

Now, there are a number of points I could make here about epistemology (i.e. we don’t know how German troops behaved) and about the flexibility of the rules in question (which I know ‘cos I wrote them). On the latter point, they do actually say somewhere that German tribes can consist of formed foot (the reference is to a slightly obscure sentence in Tacitus), but that is not the point I’d really like to make here.

The tempo bidding process and use to move troop bases in PM is, admittedly, a game mechanic. It is also a hallmark of the Polemos rules sets. If it was not there, I would not have been writing a Polemos rule set. But I do not think it is ‘just’ a game mechanic.

The tempo use actually is an attempt to model the command and control processes in a battle. This is not to argue that the tempo processes reflect real battlefield processes. In order to do that, we would have to dress the wargamer in armour, create a large amount of noise and convince them that the other side were out to kill them. But we can attempt to model some of the aspects of the overall outcomes of the command process.

Now, actually, the tempo process consists of two parts: you bid to control the tempo of the battle, and then you use the rest of your allocation to order units about. The second part is, I admit, reasonably conventional – DBA uses its PiPs for a similar purpose, although I submit that DBA bases are too dependent on PiPs; if you haven’t got the points, the bases stop.

The tempo bidding phase is, however, a bit more subtle, and aims to capture the overall generalship processes. Here is some skulduggery, attempting to outthink and outwit the opponent. The balance has to be struck between seizing the initiative and being able to order troops about.

This, I think, is the irreducible part of the Polemos command process and this, therefore, raises it a little above the ‘mere game mechanic’ category.

Now, I do need to be careful here not to start any special pleading. If I had started out with a completely blank slate for the rules, I would probably not have had a similar sort of command system. Nevertheless, I do think that the tempo system attempts to model something in reality.

Now, as discussed above, just because we model something in a certain way, it does not mean that we have an ontological commitment to that thing being there in reality. Photons are modelled as waves but are not thought to be waves.

In a similar way, tempo points model command commitments and generalship, but they are not the reality. The reality was almost certainly messier and more fluid than any model could capture. What the tempo points model does add to, say, a PiP or orders based system is the interaction of the two commanders in trying to control the pace and location of the battle.

Now, it could be argued that the tempo system tries to do too much, in both reflecting the initiative control in the battle and the command and control system. Perhaps that is true, but it does, I think, reflect the resource allocation problems that generals might have had, at least in some senses. Again, this is an attempt to model this aspect – in model or metaphor terms it is and is isn’t a reflection of the reality.

Overall, then, I don’t think that any part of a rule set, a model of a certain reality, can be dismissed as just a game mechanic. However abstract, the parts of the rule sets are trying to model some sort of underlying reality. Whether they do that successfully or not is another matter, and that is why we need multiple models, multiple rule sets.

More on Models

Not being one to not beat a dead horse if at all possible, I’m going to talk a bit more about models this time.

Models, as I’ve said, are terribly useful sorts of things. In fact, they pervade much of our lives, even though we do not often notice them.

It is noted by some philosophers of language that, in fact, a lot of language is dead metaphor. The word ‘tradition’, for example, is a metaphor for ‘that which is handed down’, but we do not notice, or indeed know that about the word, while we do understand the meaning of tradition.

Models, in fact, capture the salient points of what we are trying to discuss across subjects. In science, social sciences and the arts, models, or their simpler cousins, metaphors, permeate our language and thinking. To learn something new we often (or, according to some, usually) use a model or metaphor to compare and contrast the new thing with something we do know and understand.

I’m sure you must realise by now that there is a ‘but’, here. In fact, more than one.

The first but is that, in most cases, a single model is inadequate. Models capture some aspects of the thing to be studied. A model cannot be the thing itself. This is for two reasons. Firstly, if the model were the thing itself it would not be a model, it would be a replica. Secondly, the whole aim of modelling is to simplify things sufficiently to understand them. A model which is a complex as the original would be of no use in that direction.

Therefore, multiple models are necessary to capture the different aspects of the reality we are interested in studying. In terms of a battle and a set of wargame rules for that period, the rules, probably, already contain different models: there are models for movement, command, combat and morale. But the single set of models (or, if you like, the single extended model covering all these things) is insufficient to represent the real event.

The upshot of this seems to be that we do need different sets of wargame rules to cover a single period, because the choices and models, or sets of models chosen by the authors will vary, and hence will capture different aspects of the original, assuming that the authors do their homework.

The result is that we should be very wary of any rule set that ‘rules the roost’ in terms of dominating the market. No single model is adequate to representing a battle in its entire nuance, even in the fairly limited set of things we wish to represent in a set of wargame rules. A single paradigm is insufficient to represent reality, and we should not expect it to do so.

The second issue there is with modelling is to look at the dark side of models. I noted above that much of our language is dead metaphor. But our models can die too, and not give us the insight that we should expect.

To an extent, this is related to the point made just above of dominant rule sets. If there is a dominant rule set, then we have a tendency to read back from that into history or reality. In effect, we make an ontological commitment to the model, and claim that the objects in the model are those in real life.

A non-wargaming example would be Maxwell’s equations for the wave motion of light. These worked really very nicely, and the model of light as waves accounted for an awful lot of observed phenomena. The catch was that waves need a medium to travel through, so the model contained this stuff called ‘ether’ for light to behave in.

As you might know, physicists at the end of the nineteenth century spent quite a long time looking for the ether, and failing to find it. Ultimately, Michelson and Morely conducted an experiment to show that it didn’t exist, and eventually, Einstein came along to explain how and why this could happen.

The model contained an ontological commitment: that ether existed. The model actually works without such a commitment, but it implies it strongly, and so ether was presumed to exist.

I think this sort of ontological commitment can also happen when we have dominant paradigm wargame rules. For example, in ancients wargaming, the DB* paradigm is still very much in people’s minds. Now, I’m not one to slate DB* (at least, not in its ancients forms) because it was a breakthrough in terms of wargaming, but that very success can lead to problems.

For example, I was somewhat nin-plussed when a review of Polemos: SPQR commented that ‘there are no warbands in the rules’. My first reaction was ‘why should there be?’ but I think what is going on is a hardening of the DB* model of ancient battles into an expectation. In short, there could be an ontological commitment to the existence of warbands, because the paradigm model is being read back into the history underlying it.

The problem here is that people might look at a new rule set and think ‘it doesn’t have warbands in it, but I know warbands existed, therefore it is a bad rule set’. The model has been substituted for the reality.

So far as I know, no ancient source referred to foot soldiers as ‘warbands’. Warbands are, simply, an element of one model of ancient warfare. Even the name is, I think, somewhat misleading. The foot of tribal armies were not all well trained and well-armed, or committed to the cause. Those who were, were the chiefs comitatus, and they were relatively few in number.

The main bulk of the foot of tribal armies are described in PM:SPQR as ‘tribal foot’, which is, in my view, a more defensible description of them.

Of course, I’m not suggesting that PM:SPQR is the correct paradigm for us to use for ancients wargaming; it is just another model. But perhaps this example does indicate the pitfalls of dominant models, all pervasive rule sets.

Wargame Models

If we want to be a bit pretentious about our hobby, we can claim that wargaming models a real world battle. We can argue that through participation in a wargame, the player can learn something about the history of the time it is set in, the possibilities, the reasons why some things happened and some didn’t.

This is all well and good, and highly educational, and therefore, acceptable in polite society (if such a thing exists any more). Wargaming becomes a vital part of anyone’s life experience. Without a wargame, we are condemned to relive the more violent parts of history. Wargaming is thus part of the moral fabric of our society; to spread the word or the wargame is to engage in a moral crusade, a battle to inform our children of the evils of conflict before it is too late….

I’m sorry. I quite forgot myself.

The problem is, of course, that wargames which set out to educate often land up as rubbish games which no-one wants to play. It is a bit like modern children’s TV, or at least the bits that I see or hear about. Children’s TV used to be about telling stories and singing songs (think Bagpuss). There was no explicit educational agenda. These days, a lot of children’s TV seems to be pushing an agenda, for example telling the children to eat 5 a day fruit and vegetables. This makes for poor viewing (or maybe I’m just an old curmudgeon), aside from the fact that 5 a day has no basis in scientific fact…

But I digress.

The point I am working my way towards is that, yes, in a sense wargames do bear some reality to history, but we do have to recognise in what sense.

We can recognise two sorts of models: a scale model and a ‘disclosure’ model. A scale model is, simply, a reduced version of the original. A miniature car, or a working steam engine are scale models. Our toy soldiers are, as well. The thing about scale models is that you do not learn anything very much about the original by look at them; certainly nothing more than you would learn from looking at the original.

A disclosure model is, in a sense, a metaphor, or at least an extended one. We draw an analogy between the model and the real thing, recognising that the model is and is not the original. By this I mean that we keep in mind that the model is limited in some, fairly well defined ways.

Let’s try an example. Chess is often used as a metaphor or model for war. In chess we have various pieces with different capabilities, a battle field and a fair bit of skulduggery between the players. In this sense, then, chess is a model of a battle. We can talk about strategy and tactics, subterfuge and manoeuvre, and many of the things which war and battles involve.

Now, we also recognise that chess is not a battle. People in armies do not come in such a range of different sizes; conflict is not so simple as landing on another piece’s square and removing them from play. The aim is not always checkmate of the opposing king, even if that is possible on a battle field.

So, in these senses, chess is and is not like a battle. It is sufficiently close to allow us to use the analogy to obtain some understanding of an otherwise closed experience to us, but is also sufficiently different to persuade us to work at the model, to understand the differences that there are.

Now, wargames have occasionally been described as ‘chess with a thousand pieces’, but if I start talking about wargames as a model for chess I’ll get confused. Anyway, I do not think that wargames and chess are particularly analogous, but I do think that wargames and war should be, otherwise we shoud just pack up our toys and slink off into the distance.

The point is that wargmaes, if they are to model some reality, have to be a disclosure model. A scale model of a real battle is ultimately, boring. While are toys may be scale models of the real ones, our battlefields models of the real ones, our command structures the same and so on, a scale model of the battle itself would mean that everything would have to follow from the real one. And this means that no disclosure, no understanding, is possible.

This is where the skill, or craft, or art, or whatever, of writing rules comes in. The aim is provide a model, an extended metaphor, of what happens in a battle. We can only model some aspects of a battle, and so we have to choose them carefully. A rule set that exhaustively covered every aspect would be a scale model, not a disclosure one.

So the aim of, say, refighting the battle of Hastings, or Waterloo, is to provide a set of rules which can aid the understanding, and are also simple and fun to use. For example, a rule set for Hastings should, in some way, allow for the strength of the Saxon position and the attrition of their forces on the hill. That may then permit the Norman player to discover why the battle plan of William was what it was. Other, perhaps more scenario specific details, such as the approach of dusk, will focus the mind more on the problem the Normans actually faced on the day, in a way that a pure scale model of the battle would not manage.

So, maybe, the next time you are tempted to cry out ‘That wouldn’t happen’ when something unexpected on the table does occur, pause for a moment and consider: why not? The rules might be trying to tell you something.

Problems of Interpretation

I’ve mentioned, probably many times before, issues that there are with historical sources. The sources we have from the ancient period (and that is what I’m going to talk about, even though the comments apply more widely) are scarce and, almost certainly, biased.

They also use conventions that we may not understand, and certainly have a relationship with the ‘truth’ (to use a tendentious word) which is not the same as modern historiography would allow.

I’ve been reading the Landmark Arrian, which is a very nice book, well produced, with handy maps, notes and pictures. As you probably know, Arrian chronicled the campaigns of Alexander III of Macedon, commonly known as ‘the Great’. So far I’ve just about got beyond the Battle of Issus, but even at this relatively early stage, problems are becoming clear.

What sort of problems am I talking about? Well, there are problems of interpretation, and these are multi-layered.

For example: Why did Alexander invade Asia?

Arrian’s answer to this question is that Alexander invaded Asia to punish the Persian Empire for invading Greece. Now, Alexander was, of course, a politician, and he knew that such an answer would play well in Greece and the Greek cities of Asia Minor. Indeed, as he captured the Greek cities he generally granted them their freedom, although exactly what that meant is not clear. Mostly, he seems to have restored the Greek version of democracy to them, and ensured that they allied with Macedonia.
Now, Arrian details Alexander’s march from Greece (he had to capture Thebes before invading Asia) to the Hellespont. His list of cities and rivers crossed is a perfectly reasonable one, so far as anyone can tell. However, it is also notable that his route is the exact opposite of that detailed by Herodotus in his Histories, when he is describing the invasion of Greece by Xerxes.

What are we to make of this?

Clearly, Arrian is trying to make a point here about the invasion. Alexander’s claim to be avenging the invasion of Greece is made later in the book than the actual invasion itself. The itinerary, then, is actually making the same point in advance; foreshadowing, if you like, Alexander’s explicit argument for invasion.

What, then does this do to our interpretation of Arrian’s account of events? Do we believe that Alexander did take this route, or is it a literary trope? If Alexander did not take this route, how did he get from Thebes to the Hellespont? If he did, then did he do it consciously to make a political and diplomatic point, or was it just the best route available?

Now, as wargamers, this may not be a major problem to us, but it does affect how we view the reliability of our author. We have to recall that Arrian, although reliant of now lost sources written nearer the time, wrote in the second century AD. His sources were also probably biased, in favour of the Macedonians. So we are only getting at Alexander as an interpretation (Arrian’s) of an interpretation (his sources) of the events; and that assumes that the sources had clear accounts themselves.

The problem is if, for example, we take seriously the parallelism between Alexander’s invasion of Asia Minor and Xerxes’ of Greece, we land up with a sceptical position about a fair amount of what we read.

For example, if the first encounter of the Persians and Greeks was at Thermopylae, then what does that make of Arrians account of the action on the River Granicus? After all, Granicus was, simply a victory over local Persian forces, defending a river line. But these same forces had been able to ward off the Greeks in Asia Minor for a century or so.

I wouldn’t want to suggest that Granicus did not take place, nor that either Arrian or his sources are making stuff up, but the parallelism seems to me to be a bit suggestive that, perhaps, other work aside from his sources, informed Arrian’s account.

It does get a bit worse, too. Arrian, by his own account, is out to lionise Alexander. He thus presents him in a certain way, describing his abilities in battle (he is always at the front, the vital point) as well as his gracious diplomacy.
But took at a map of the manoeuvers before the Battle of Issus. Who let whom get across their lines of communication? Darius was not cut off from his base, while Alexander seems to have managed that, and not to have command of the sea at the same time. Who now looks like the better general?

Now, as everyone knows, the Macedonian led army won at Issus. It wouldn’t be the first or the last battle when the army in the worst strategic position has fought itself out of it. But the general acclaim of Alexander’s generalship does look a little flaky about here. He may well have been confident of the fighting abilities of his troops, but he probably should not have left them in such a dubious strategic position.

Now, Arrian tries to defend his hero by claiming that Darius should have stayed where he was, on a plain at Sochoi where he could use his numbers. This idea is presented to Darius, on Arrian’s account, by a renegade Greek, who Darius ignores, going with advice from his Persian advisors who Arrian is quite rude about (2.6.3-7).

I think we can see what could be going on here. Darius did not make a mistake but the blame is placed on non-Greeks and Darius himself by Arrian (who also suggests a god had a hand in it), while what actually seems to be the case is that the Macedonian – Greek army simply had to fight its way back along its line of communications. Arrian’s account of the battle also suggests cowardice on Darius’ part, by the way, as he was the first to flee.

This, then, leaves us with an account of Issus which has a slightly dubious relationship with the battle itself. But it is the account which we have to work with. Bother.