Saturday 16 August 2014

Abductive Wargame Rules

The problem with wargame rules is that they are written down. On the face of it, this is not too much of a problem, because without being written down, they would not constitute wargame rules as we usually understand them, but writing stuff down tends to do a variety of things, not all of them positive.

Firstly, we find ourselves in hoc to language and its variety of understandings. Wittgenstein showed a while ago that language is not a simple, transparent thing with a single reading, meaning and understanding. I guess that anyone who has written a set of wargame rules, even for private use, has come across the problem of multiple readings and misunderstandings.

Secondly, there is the problem of, as Polanyi said, ‘we know more than we can tell’. A set of wargame rules is a result of the author’s understanding of something, in the case of historical wargames, some aspect of military history. This understanding will form the background to the rules, but will almost certainly not be spelt out in them. I may have the opinion that Greek generalship before the Peloponnesian Wars was non-existent, and I may have my reasons, from my reading of history, for believing that to be the case, but they will not find their way into the rules where, for example, Greek armies before 450 BC are not given a general.

Thirdly, and perhaps more subtly and importantly, there are differences in the way we hold beliefs. One way of looking at this is to suggest that there are three types of belief. The first is logical:
All men are mortal.
Socrates is a man.
Therefore Socrates is mortal.
Given the premises, the conclusion follows. The problem here is that, given the premises, the conclusion follows. We do not, actually, learn anything new here. All we do is affirm something that, in the simple form given above, is obvious.

The second way of forming beliefs is that of induction. Every swan I have ever encountered is white; therefore the next swan I encounter will be white as well. Now, clearly, this is not the case, but actually it forms the basis for our scientific understanding. Every electron we have encountered so far is negatively charged, therefore the next one will be negatively charged. This works, of course, but is always open to the possibility that the next encounter will break the rule. Furthermore, we do not learn anything particularly new here. We know that, in general, infantry squares hold against cavalry, so we know that the next infantry square which finds itself to be charges by a load of blokes on horses will probably hold. This is simply firming up our already existent belief, not learning anything new.

New stuff, therefore, has to come from somewhere else. C. S. Pierce called this abduction, John Henry Newman called it the illative sense. I suppose that, in modern philosophy of science speak, it would be called the hypothetico-deductive method. The problem is that, as mentioned above, neither inference nor induction actually generates anything new. However, knowledge is expanding, new things are being found out and verified, so there must be some way of actually figuring out new stuff. This is abduction.

The idea here is that I can think something like ‘the reason squares held against cavalry is because of the morale of the infantry’. I thus have a hypothesis, which I can do two things with. Firstly, I can search of existing evidence to back my claim up or disprove it, and secondly I can implement it in a set of wargame rules and see if it gives me sensible answers. But the point is that the hypothesis is not held like a deduced law of logic, nor is it an inductive rule. Both of these may come into the affirming or rejecting of the hypothesis, but they are not part of forming the belief, the hypothesis, themselves.

An abduction, therefore, is not a rule in the same sense as a logical deduction or a inductive rule. It is held in a much less rigid sense than either of these. It is a conditional, and the things that derive from it in either of the other two senses always has (or should have) an implied ‘if…’ in front of it, as in ‘if the morale of the infantry is decisive in the survival of a square, then I need to check morale each time a square is charged’.

I am sure you can see where this is going. Many wargame rules are abductive. The example of the square is just one example from one particular period. However, that is not how the rule or rules are actually presented.  I have written before, and I dare say I will do so again, that wargame rules require certainty. A set of rules which started each rule with an ‘if…’ statement, such as ‘if you, too, believe that the morale of a square….’ would be a far longer work than most people would put up with, be very boring and probably really hard to use in a wargame.

The issue is, then, that wargame rules run the risk, indeed the likelihood, of being mistaken for either deductive or inductive rules when they are, in fact, held in the conditional, abductive, sense by the author. We can thus land up in situations where there are disputes between the users of different sets of rules because the authors disagree in this abductive sense, while the disputes are disagreeing in an inductive or inferential sense. Writing an abductive rule down, as we have to in a rule set, makes it look like one of the other forms.

When this is added to the other problems of the written word, of course, it becomes remarkable that we can, in fact, agree on what most rule sets say, most of the time. The human mind is much better at working out meanings from ambiguous text that most philosophers of language are willing to give it credit for. At least, that is the way I read them.


  1. Wittgenstein showed a while ago that language is not a simple, transparent thing with a single reading, meaning and understanding.

    I was not aware he had played Advanced Squad Leader.

    1. Interestingly, the physicist Claude Shannon said the same sort of thing, at about the same time, but for very different reasons.

      as for ASL, all i can say is 'more is not necessarily better'.

  2. Excellent post - as ever.

    I did have a few thoughts, though they may add little to your discussion - as ever.

    Firstly, it seems that the actual nature of language enters into this. I have considerably less relish or stamina than you have for studying philosophical ideas - one reason I have trouble with this is because there seems to be a paradox implicit in conveying exact ideas by means of language which, for the most part, is itself not exact at all - not without special constructs or nested definitions which make it less ambiguous but much harder to absorb. One aspect of this, as a poorish example, is that writing and reading are linear activities, yet a set of war-game rules is pretty complex, and has links and cross references all over the shop. It is possible, of course, to go to some trouble to cope with that in a linear document, but it also has to be a document in which people can find what they are looking for, and that in itself adds another set of links and indices. I have a set of rules for the ECW in my house which cover everything you need, explain it in an agreeable manner, yet are almost impossible to find anything in if you wish to check a point - structurally useless.

    It also seems that definitions come into this. If we define an electron as a negatively charged particle, or a swan as a very large white bird, then the classification of our belief changes somewhat. I don't like war-game rules which say "cavalry cannot break a square" - intuitively, i prefer rules which allow squares to break, but make it so unlikely that you are advised not to try it. That seems more like a game, though the difference in effect may be slight.


    1. I think the fact that most wargame rules have not gone down the non-linear document route which was all the vogue when HTML arrived is an indicator of something, but I'm not sure what. innate conservatism, probably, plus the fact that most people don't want computers cluttering up their wargaming space.

      As for language, then yes, there is a paradox. Deridda wrote that there was no authorial intention outside the text. He seems to have written it with a straight face and expected us to believe it. He seems to have inadvertently given us an escape clause from post-structuralism, in that we don't have to believe him, because as author he has no control over our interpretation of his text.

      I guess the cavalry and squares thing comes back to how constraining we would like our rules to be: don't do that or you can do that, but consider the cost. The latter is more subtle, but whether it is better (in any particular sense) is a bit moot.

  3. Fantasy/SF rule writers cannot be using abduction, induction or deduction because there is no data set. Yet they produce a finished product that is indistinguishable from historical rules in any functional sense.

    1. Don't they have a sort of false deduction going on? Something like this:
      "(If there were such a character as a Hero who was a strong warrior with good armour) it takes 5 wounds to kill a Hero. (If there was such a thing as a wizard with magical powers) a wizard could cast a healing spell to cancel out one wound. Therefore it takes 6 wounds to kill a hero who is helped by a wizard".

      There is a sort of logic to them once you accept the initial leaps of faith (or invention of a data set).

    2. I think that SF fantasy rules still do have all the logics, because they are human constructs and we expect things to go in a certain way. If we shoot at something and hit it, we expect to do some damage. Hume would argue that there is no necessary connection between cause and effect, but induction usually works.

      When you work backwards far enough, most things start from unjustifiable premises, anyway. Whether F/SF writers set out their premises in an explicit way is a different matter, but having contributed to some SF rules I can assure you that logic is still there....