Book Review: Eisenhorn Trilogy

Spoiler Status: Very mild spoilers before the fold, full spoilers after the fold

I read the Eisenhorn Trilogy on the recommendation of my friend Raymond Arnold, who bought me the first book as a birthday present, presenting it as worth reading and as providing insight into Lovecraftian horror, and as something that would be fun for us to talk about.

I did not love the books as much as he did, but that is to be expected. What I found was a fun romp through numerous action sequences that slowly starts asking interesting questions in between action sequences. In the Warhammer 40K setting, the warp allows travel through space, and also offers lots of other benefits if you are willing to engage with it, but also has the side effect of being full of mind-corrupting demons that drive you into becoming an insane murderous servant of chaos, so the Empire has to deal with people continuously going insane and forming cults that worship chaos demons and try to kill everyone. In response, they have the Inquisition (think the Spanish Inquisition, except the demons are real, have mind control powers and are the most efficient source of physical power) which goes around trying to keep down chaos, including all the Inquisitors who get a little too close to the chaos and turn insane, which happens a lot, and arguing over how engaged with chaos they should be willing to be in order to fight it.

Inquisitors also go around fighting with energy guns and specially engraved swords, killing huge numbers of chaos-enhanced enemies while mostly dodging any attempts to kill them, and looking dope doing it.

There is an obvious story one would tell about such an Inquisitor in this universe. If you’d like to read that story, the books aren’t bad.

Beyond the fold, SPOILERS in the service of discussing the trilogy’s slash universe’s practical and ethical dilemmas.

‘Have you ever noticed that anybody driving slower than you is an idiot, and anyone going faster than you is a maniac?’ – George Carlin

The Inquisitor’s Dilemma

Anyone who engages more with chaos than you is Extremis Diabolus. The punishment is death. If you can’t prove it yet, then for now they’re just a radical. Anyone who engages with chaos less than you is… the guy putting out the proclamation that you are Extremis Diabolus. They then try to kill you, so the effective punishment is again death. If they can’t prove it, for now they are the Puritan who is preventing you from Doing What Needs To Be Done to keep the galaxy safe and by the Emperor’s name you are not about to let them find out what you are up to, lest they destroy things vital to your work and Let The Terrorists (Chaos) Win.

This is especially awkward as Eisenhorn makes the journey from relative Puritan in Xenos to a (well meaning and it turns out all right, but let’s face it) Extremis Diabolus in Hereticus. There is no way to pretend that his actions are remotely consistent, other than to not think about it, which is of course the path he chooses.

He starts out in the climax of Xenos, fighting well-meaning Inquisitors so he can secure priceless sources of information to help fight Chaos.

He then spends much of Malleus dealing with those who think he is cavorting with a demon, in large part because he kind of is, even if he doesn’t know about it or see it that way. It turns out pretty much anyone in Order Malleus can decide to go kill you if they think you’ve Gone Too Far and there is not much anyone can do about this other than use physical force.

Then in the climax of Malleus he makes a deal with a not-well-meaning-at-all agent of Chaos and he and his friends go rogue in order to fight another well-meaning and also legendary Inquisitor who he thinks has Gone Too Far and has certainly done some pretty bad stuff, but seems to be trying to do an epically good thing and there is no evidence Eisenhorn doesn’t believe that, and he certainly neither shuts up nor multiplies to see if the plan is a good idea. He just goes all Puritan on the guy. He also keeps the deal he made, because he Has A Code.

Finally, in Hereticus, in order to deal with the fact that he kind of unleashed an ancient evil that killed most of the people he knows and loves, and is about to bring destruction upon the whole galaxy, he ends up killing a number of Inquisitors and a lot of completely innocent people (but at least having the decency to feel bad about it, so it’s all right), summoning and binding demons and working with aliens (which is very much not allowed under any circumstances) among other things. This works, and the galaxy is saved, and Eisenhorn somehow manages to come home without being executed.

Needless to say, this system of doing things is not ideal. But can they do better? It is easy to understand why everyone involved is willing to kill everyone they are killing – either they are trying to save a lot of lives, or they are acting in self defense, or often both. When there is a lot at stake, such justifications are easy, but it would be better for everyone if they could cooperate or at least not kill each other.

The Inquisitors are stuck in a hard mode Epistemic Prisoner’s Dilemma. Everyone wants to defeat Chaos and defend the Empire, unless they have gone completely mad from Chaos. They disagree on what actions and methods would advance that cause. The first best solution would be for everyone to cooperate and reach compromises, and create procedures such that whatever chaos knowledge was accessed was done so as safely and profitably as possible, in exchange for there being somewhat less of it than the radicals would like, and for the radicals to be allowed to advocate for and occasionally do reasonably terrible things when there was enough on the line. The alternative is an equilibrium where much of the chaos knowledge and power that is accessed is being used directly against the Inquisition, and the knowledge being used to help can’t be shared. and Thus most of the knowledge the Inquisition gets access to ends up being destroyed, while also being more locally dangerous, and not doing much good. It’s a miracle mankind is still standing.

It is also worth noting, in addition to being willing to use the warp to travel through space, that thousands of souls are sacrificed on a continuous basis to keep the Emperor alive, that procedure not being chaos-based seems at best highly unlikely, and everyone seems all right with that decision.

The easy mode Epistemic Prisoner’s Dilemma is isomorphic to the standard Prisoner’s Dilemma, but it doesn’t feel like it is, because the other person is wrong. Your cause is just and vital, their cause is misguided or worse. They do not know their own utility function. You believe that they would be better off if you got your way, they just don’t know it yet, so you can defect to get your way and they will thank you later. This leads to the instinct that defection is not defection, but that instinct is wrong. Already, each player is maximizing their own utility, and if they value the other player doing well, the right way to handle that is to consider that part of your own utility function. If you think their model is sufficiently wrong, you can reasonably not care at all about them getting what they think they want, but if you actively oppose it, that is only another way of expressing your own preferences, and we already have a place to do that.

The reason the Inquisitor’s Dilemma is hard mode is that chaos drives people crazy, including through explicit mind control. This means that as an Inquisitor gets more radical, there is a greater chance they are not thinking clearly, and a greater chance they have become outright corrupted and are now evil, and they have a greater chance of corrupting anyone and anything they come in contact with. Can you make a deal with chaos? If you do, should you keep that deal?

Before we get to the specific examples of that in Malleus, we should ask if Chaos, as this world’s version of Lovecraftian Horror that drives you mad when you look at it, is actually different in kind from the forces that act in the real world all the time. My instinct says that the answer is basically no, so getting these answers right is important.

If you are drunk with Chaos power, how is that different from the ordinary version of being drunk with power? Looked at another way, how is it different from being drunk with alcohol? Is there knowledge that does not try to twist our minds to its own ends? Chaos is an extreme version of this because you can put on an infected necklace and instantly lose your mind, or use an infected paint bottle and have your paintings start killing people, but wait, don’t both of those literally happen in real life, if usually a bit less dramatically?

Power corrupts. Absolute power corrupts absolutely. This corruption, one could even say ‘warping’, is probabilistic, as some manage to mostly resist, but some amount of corruption is universal. Power always goes to your head to some extent. What do people with power want? More power. They also want other things, but the more power they have and the more power is up for grabs, the more they act as if all they want is more power. Think of the potential! Sounds a lot like chaos.

The metaphor of the alcoholic or drug addict is even easier to see. Connections in the brain physically warp them to seek out their drug of choice at the expense of everything else. They will turn against their friends, and hide their actions. Their friends will, if things go too far and they care, attempt to intervene, and fights ensue. Usually these fights do not involve swords, but the same principles apply.

The physicist probably thinks physics is super important and everyone should study it more and he should prioritize it over other things. The architect probably thinks the same thing about architecture. The Magic: The Gathering professional says the same about Magic. Every start-up founder worth their salt preaches the good news to anyone who will listen, and also has to worry about power. The experts in education think the solution to our problems is more education, the psychologists think the solution is more therapy, unless it’s more drugs.

If we can’t solve the hard mode Epistemic Prisoner’s Dilemma, we can’t cooperate in practice in normal situations, because everyone is warped by what they choose to learn and the power they draw from. It whispers to you and drives you mad, so figure out how to deal with it.

I guess this is why I didn’t get the same sense of horror from the trilogy that Raymond was trying to convey. Yes, the idea that Chaos runes are so dangerous that looking at one can hypnotize you, but that is not different in kind from real hypnosis, just degree. There are books that tell you to kill yourself when you touch them and even someone trained to know what is about to happen almost does it, but again, this is just a more dramatic version to let us play with concepts. One could call this Terror in the Merely Real?

War on Chaos

I also didn’t view the Inquisition, which is obviously The Spanish Inquisition IN SPACE, as alien, either, for similar reasons. If demons are real and looking at them corrupts good people, what else would you do? Extreme problems require extreme solutions. Yes, the organizational structure they choose is more than a little nuts, but there seems to be a reasonable attempt to compromise between due process of law and having a continuous flow of hypnotic super terrorists to deal with.

I doubt America would handle the situation with as much grace.

In fact, America has its own version of this, which it calls the War on Drugs, and it is failing at these problems super hard. Chaos in many contexts is the Substance D of drugs and it also turns you into a hypnotic super terrorist, whereas most of the drugs we are dealing with are addictive and bad for your health but on the mostly-non-addictively mellowing out to one look transforms you into a hypnotic super terrorist scale none of them are that far to the righthand side. We really shouldn’t be putting so many people in prison and destroying so much of the trust in our society that the Spanish Inquisition IN SPACE fighting against real demons comes out looking better, like a more livable alternative.

Having an Actually Evil enemy to fight probably helps a lot with that; Imperial citizens don’t only want to placate Inquisitors because of their absolute power, they also understand the importance of the cause. Still, that’s not much of an excuse.

Extreme Sanction

The other thing one could say is that at least we don’t solve our problems via aerial bombardment, contenting ourselves with the second best solution of rent control. No matter how bad things get no one considers wiping out entire neighborhoods or cities, let alone planets. This certainly is to our credit in important senses, but it is hard to argue with this being the right thing for the Inquisition to be doing when facing something sufficiently dangerous. Better to sacrifice locally than risk things getting worse. If we faced a problem that actually did require extreme sanction, would we be able to use it?

What is the farthest we would be willing to go to contain a deadly plague? Would be wipe out a city if it was the obviously right thing?

If we lived in a paradise but it turned out it all depended on the unhappiness of one child, would we decide this was actually a Young Adult Dystopia and destroy the system?

If the Emperor was the only thing keeping mankind alive, and a few people had to die each day to preserve that, would we do it, or would we get all human rights about it and all perish?

What would we do at the end of Cabin in the Woods? Would we be right?

Could we even ask these questions out loud, or are we too afraid?

AI Boxing and Dealing With The Devil

Should Eisenhorn have kept the deal he made with Pontius Glaw at the end of Malleus?

I would certainly argue it was a profoundly stupid deal. Glaw’s bargaining position wasn’t the worst, but it was pretty bad, and there were many things Eisenhorn could offer that would have been tempting without being so obviously dangerous.

Compared to the default AI Box case, Eisenhorn had huge advantages. There was no chance he would be fooled into thinking Glaw wasn’t evil. Glaw did not outclass him in intelligence by that much. Glaw’s knowledge base was limited. Glaw didn’t have the ability to inflict disutility inside the box (e.g. by simulating people). The box was physically secure. Glaw has a utility function that cares a lot about what happens inside the box. Glaw even played it straight, and gave Eisenhorn instructions to do dark rituals that did what Glaw said they would do! And despite all of that, he still lost. Hard.

Glaw thirsts for knowledge and conversation, and Eisenhorn can wipe him from existence at any time, so at a minimum he should have tried to offer lesser things to get what he felt that he needed. I would argue that if Eisenhorn intended to keep his word, he should have at leastmade a desperate effort to do that, and this is a man who makes a desperate effort about once every twenty pages. Instead he opens his negotiations on letting the known-to-be-evil super-intelligence out of its box and into what at a minimum is a much less secure one. You need not be especially Genre Savvy to know what is going to happen next.

Set that aside, and assume the bargain has been struck. Should you keep that bargain? There is little doubt that Glaw would not keep such a bargain. The decision process of Glaw and that of Eisenhorn are only correlated to the extent that they are predicting and modeling each other, or to the extent that Glaw has corrupted Eisenhorn, and if Glaw has already corrupted Eisenhorn then he really doesn’t want to let Glaw out.

That leaves four arguments I can think of, three of them closely related, in favor of keeping the deal.

You can argue that this is Parfit’s Hitchhiker, that Glaw would (at least probably) know if you did not intend to keep the deal, so you have to be the person who would keep the deal if you want him to give you the information. Given you want to make that deal if you would have to keep it, you therefore make the deal and keep it.

You can argue Inquisitors in general or any given Inquisitor in particular needs a reputation that they keep deals they make with those infected by Chaos, in order to keep making such deals in the future, and therefore deals must be kept.

You can argue from virtue ethics that one should be the type of person who keeps their word even when that seems pretty bad, and even when (as in this case) no one will ever know if you fail to do that. Eisenhorn does invoke this to some extent.

What Eisenhorn mostly argues is option four, which is that Glaw won’t be able to escape and the consequences therefore won’t be that bad, which is also why he was willing to deal in the first place. I can’t even.

The arguments against are rather obvious, and Eisenhorn had a number of options even after making the deal with intent to honor it. One obvious option is to make Glaw a body that cannot defend itself and then destroy it, since he in no way promised not to do that. In general I find invoking Exact Words kind of terrible, but here I’m not sure it even goes against the spirit of the arrangement. Spirit of the arrangement is for negotiations in good faith, not for deals with actual devils.

In the end, I’m not sure what the right thing to do is here, which tells you how far I go towards the ‘honor your word’ camp.

Selective Intelligence

There’s this very evil book full of quite evil but powerful rituals. It tempts those who read it to do things like summon demons. That seems bad. But it seems like some of it is good. In particular, there were runs of banishment whose sole purpose is to banish demons. They are powerful enough that even looking at them hurts. Perhaps we can copy those pages and make them standard issue? Paint them on the sides of Imperial buildings and call it a new art movement, perhaps?

Was Quixos Right?

Villians have a nasty habit of being right. This is true both in fiction and in non-fiction; there are two or more sides to every story. Policy debates should not appear one sided. Sure, there are some selfish, obvious evil ones out there, but the “look at your face” defense that Doctor Strange uses needs to not be a knock-down argument. So, given that ‘go through the proper channels’ seems like it would have been unlikely to get anywhere, does Quixos win the Magneto Memorial Award For Being Right?

That depends on the facts. Was his plan capable of working, or did he have reason to believe that it was capable?

If so, then yes. Clearly yes. Shut up and multiply.

Quixos was attempting to close the Eye of Terror. This is just about the best thing one could hope to do. The Eye of Terror is a source of constant invasions into Imperial space in the best of times, and you do not want to see what the worst of times would look like. Whole sectors of the galaxy are on a permanent military footing, and by permanent we mean thousands of years, trying to keep this at bay. If your plan has even a small chance of closing the freaking Eye of Terror, even if the price of trying is pretty high, you go for it.

Was the price high? Sure. A lot of Inquisitors and civilians died so that the necessary steps could be taken, and the components gathered. A lot of laws were violated. A lot of chaos was used, with the obvious accompanying dangers. Most of that had already happened.

The objection Eisenhorn raises to the final action is that trying it might cause a lot of damage. It might even destroy the planet! To which the correct response is, if you think the attempt might work and blowing up a planet is the worst thing that can happen, in all seriousness who the hell even cares. That objection is dumb. A random mostly uninhabited planet is small potatoes. Imperial actions sacrifice much worse all the time. Given how far things had already come, why not let the man finish his work? Why not at least ask the question?

Living in the Present

What makes all of this seem so normal, more than anything, is the profound lack of imagination in the Warhammer 40k universe’s picture of human civilization.

Medical technology advances, but the effect is several hundred years of life instead of one hundred, and the ability to survive injuries and get artificial limbs. There is an interesting cyborg-related debate underneath the surface, but we mostly ignore it.

Travel is the same thing we have now, except that there is a warp to carry us through interstellar space; the warp is basically The Premise but the effect is to make space function like any other area you can travel through.

War. War never changes. You have armies and navies and guns and swords and if anything it all feels much less advanced than the 2017 US Army. They do manage to make up new nonsense names for new sword techniques, and the sword has lots of cool tech in it, and the guns shoot energy instead of bullets, but that’s about it.

Architecture gets bigger, but it doesn’t get different. Huge Imperial hives don’t have much in the way of innovation. People go to the same parks and forests and bars and stores and districts and factories and jobs that we are used to. For fun people play Regicide, which is just a board game that does not sound especially interesting.

There are “mutants” with all sorts of strange features, usually lots of extra copies of things, occasionally useful but mostly hideous. These could not have come about through “mutation” any more than Wolverine or Professor X, but there is no evidence of genetic engineering. Otherwise, neither evolution nor devolution of humans seems to have taken place on any important axis in forty thousand years, although planets occasionally differ about things like ethnicity or average height.

Mankind has reached the stars, multiplied into the trillions and had forty thousand years, and in every way that isn’t directly impacting militarily-relevant medical science, weapons or space travel or the warp (e.g. The Premise) it seems like things are mostly less advanced than 2017. This isn’t just not-a-singularity or what happened to AI. This is just a lack of imagination.

Conclusion

There are always more things to talk about, but overall I would say that everything felt too familiar. I did not feel uncomfortable at any point, like things were actually strange and alien, despite that being a lot of the goal. The whole where-is-my-mind thing that Eisenhorn does is so clueless compared to even what is actually happening that it feels like intentional cluelessness, and it combines with the constant battles to make it seem like the man simply has on impenetrable Plot Armor. I was hoping that Eisenhorn would end up having gone too far, and his friends would have to kill him and finish the task without him or something similar, but I never thought that had much a chance of happening. Nor did I get the sense that the end-of-the-world threats had any chance of happening either.

I think a key element of Lovecraftian Horror is that you lose. The forces involved are infinitely older and more powerful than you or even all of humanity, so you lose. Your hope is simply not to lose too hard or too fast, and stave off disaster for another day, but even that seems unlikely.

We ourselves face such horrors. The one I worry about most is unfriendly artificial intelligence, but there are others. For our stories to prepare us to handle them, to truly Shut Up and Do The Impossible, we need to understand that we have no plot armor. By default, you lose. The hero dies. Things fall apart, and end. The Imperium does not manage to muddle through and barely turn out all right. You can’t just show up and trust that it’s all going to work out. If five pages from the end it looks like the bad guys are about to win, that does not mean we get to pull a cheap trick and win anyway. It means we all die.