Alternate History Philosophy

Motivation

Alternate histories, Allohistories, Counterfactuals, Contrafactuals, Parallel Timelines and the Many Worlds Interpretation of Quantum Mechanics

Causality

• In the forwards approach you start with a situation and allows it to evolve. In effect you are simulating the universe in your own imagination, choosing the path that seems to you most probable and interesting. (The tradeoff between those last two properties is another question.) This might be called the causal approach: it moves from causes to effects, just like reality. The causal approach is the one with delusions of academic value. This is the method by which the classic question "what if" can be answered, or, more commonly, superficially discussed and fantasised about. The motives here include curiosity, and a desire to understand the processes of history, and a certain mathematical elegance.
• In the backwards approach you start by specifying a final state, and trying to imagine the smallest possible change that could have plausibly led to such a result. This might be called the countercausal approach, since it moves from effects to causes. The countercausal approach is used to provide background and texture to a setting whose fundamental nature is basically determined beforehand. An obvious application is the setting for a work of alternate historical fiction.

Single Point of Divergence

There are various de jure points of divergence for the Dardanelles Victory timeline. Two examples are the choice of a different (and, in my opinion, more competent) commander, and the ab initio inclusion of infantry in an operation that in our timeline was initially purely naval. There's only one de facto point of divergence, however: the rapid success of the Dardanelles operation, which in our timeline stagnated and eventually failed. Everything else funnels through that point.

Before the de facto point of divergence the evolution of the narrative is countercausal. All the de jure points of divergence are motivated by the need to bring about the success of the operation.

After the de facto point of divergence the evolution of the narrative is causal. The differences stem, directly or indirectly, from the de facto point of divergence. So the de facto point of divergence acts as the boundary between the causal and countercausal regimes.

I've said earlier that my approach was almost entirely causal, and that's reflected in the fact that the de facto point of divergence comes very early in the story. The vast bulk of the narrative, therefore, is causal. If I'd been writing countercausally I would have started with a de facto point of divergence late in the narrative. One possibility would have been "how might social structures from before the first world war have survived into the middle of the twentieth century", I could have worked from there to "how might the first world war have been much shorter" and from there to "how could the Dardanelles operation have succeeded".

Butterflies in Orbit

Is it, therefore, impossible to predict the future in any meaningful way, as is concluded by a character in Spielberg's Jurassic Park? Well, obviously not, because we do predict the future with at least some success in some arenas. So a better way of putting the question is "does that mean that the butterfly effect is just worthless technobabble?"

Well, yes and no. Jeff Goldblum's character may have been talking rot but the loophole that makes it rot has an interesting shape. It's tied up with a concept in nonlinear dynamics that mathematicians call an orbit. These are subsets of the space which, once entered, the system tends to stay within for quite a while. So while the system changes state within the orbit frequently, it only hops from one orbit to another occasionally.

The importance for history is the idea that orbits can be associated with persistent features of world geopolitics: the United Nations, for instance, or the Commonwealth of Australia. Major features can disappear quite quickly (Austria-Hungary, for instance, or the Soviet Union) but as a rule if they exist today there's a good chance they'll still exist ten years from now. So we can't make statements about every element of the future, but we can for some elements.

To make an alternate history believable, therefore, we need to make statements about which orbits the system will probably pass through. Any statement about movement within the orbit is there for flavour. This is fine provided our extrapolation is for times shortish compared with the orbit-hopping time. But what if it isn't?

Analogy

The solution I've chosen is to pick the most analogous event that occurred in the historical timeline, and base the success or failure on that, adjusting for any change in circumstance. The historical Brusilov offensive was more successful than you'd expect from an ab initio viewpoint, so the modified Brusilov offensive in the Dardanelles Victory timeline is too. The changes between them all flow, directly or indirectly, from the de facto point of divergence.

I visualise the orbits as stacked rings. I'm being forced to pick a point on a ring, but history never visited this ring: it went instead to a ring a short distance along. So I pick a point adjacent to the point that was visited historically. Never mind that there's no way to get from one point to another (because you can't hop between those rings), it's still a slightly more historical choice than any of the other points I can choose.

And when the event has no analogue? An example is the concentric anti-Austrian offensive of 1917, historically Austria simply never faced such a long front. In this case I make the answer up, and try to choose a result I can defend.

Conclusions