There is no center

Daniel Estrada
Nov 11, 2016 · 12 min read
Image from Andris et al, (2015) The Rise of Partisanship and Super-Cooperators in the U.S. House of Representatives

In my streams, the proposed alternative to despair is to return to our previous efforts reinvigorated, to learn our lessons and do better next time. Michael Moore’s popular to-do list is a good example. He says we should fire the party leadership and the experts and politicians, presumably to replace them with people who can make better calls. The presumption is that politics will go on as before, it was just staffed by the wrong team making the wrong plays. Change the roster, change our attitudes, and our outcomes will improve.

I can’t blame someone for feeling despair, but I think it’s a huge mistake to return to what has demonstrably failed to work in the past. The mistake is understandable — go with what you know — but the only thing we really know right now is that we know nothing. We are in the initial conditions of a new period of history, a new chapter of textbooks that haven’t been written. And as we all know from grade school, with a new textbook chapter comes new vocabulary lists, key concepts, and important figures. That might be a bore for the students in a few generations, but we’re not studying history, we’re making it. Right now those lists are blank, and it is our job to fill them in.

The challenge for us, right now, is to make a clean break from the previous chapters of history, to avoid repeating its worst mistakes, and to stake a claim of our own. If history teaches anything, it is that our worst mistakes can and will be topped, so we have quite a challenge ahead of ourselves. For instance, it is tempting to frame this new era of politics in traditional terms of parties, laws, elections, constitutional powers and so on. These are the terms we’ve been using for literally hundreds of years. Why stop now?

Well, because these concepts are completely ill-suited to the tasks they hope to accomplish, which is to constrain the powerful in the interests of the people. Imposing democratic constraints on power was the original motivation and substance of the US constitution, and it has been the cornerstone of all democratic movements since. Trump turned out voters who, among many other things, were angry at the abuse of power from governments and other wealthy elites. Now everyone else is concerned about how Trump will abuse the power he’s somehow grabbed. No matter where you sit in these chaotic times, you have an interest in keeping the powerful in check.

This is the basic work of politics: we, the people, must impose checks on power. The power of the rich to steal from the worker. The power of the police to steal black lives. The power of rapists to assault women. The power of racists to assault families and communities. The power of the military to kill and torture by remote control. The power of the political class to steal from the environment and protect its own. This is why people are protesting in the streets tonight and have been protesting consistently so for a decade. Imposing a democratic check on power is what political organizing has always been about. This is the perpetual task before us.

So let’s assume, for the sake of discussion, that everything about our existing framework has failed in realizing this basic goal. Powerful parties, corrupt politicians, and broken election procedures can easily evade their supposed constitutional constraints. Our concepts and theories are bankrupt, our institutions have collapsed. What resources does this new chapter of history have to make sense of our time and return some semblance of democratic control?

I’d like to propose one key concept to add to the history textbooks: dynamics.

Dynamics is a general word to mean “change over time”. There is very little recognition in our politics about how political systems change over time, how they grow and decay and evolve. The constitution does describe mechanisms for changing and revising laws, but both the laws and the mechanisms for change are imagined as static, discrete artifacts that need occasional replacing, like a horseshoe, rather than dynamical social systems that take on a life of their own, like a child.

I’d like to suggest that if we starting thinking about politics in dynamical terms, we might have a better appreciation for what is going on, and how best to respond to our situation. But the language of dynamical systems theory can get technical, which makes a difficult discussion that much harder. So let me try to introduce the issue with some crude, pseudo-mathematical drawings instead.

Imagine we live in a simplified political universe with just one dimension of value and one dimension of control. In other words, imagine that all the things we do to control our political fates (voting, yes, but also public debating and community organizing, passing laws, joining protests, lobbying politicians, etc.) could be represented in a single dimension of control, as moving us either left or right in Fig. 1. You shouldn’t think about these moves as giving “more” or “less” control. Instead, think of these moves as changing the “configuration” of control: changing the laws, who has power, what we do, etc. You can make big or small changes to the configuration, but only by moving to the left or the right.

Fig. 1: A simple model of political compromise

The curved line in the diagram above describes how different configurations of control change the “value” of the system, where “value” means how happy people are in the system as determined by some objective measure. Again, these are gross simplifications: people value things differently, there are no simple linear orderings of the control configurations, and there are many dimensions of control and value relevant to politics that can’t be represented independently. Things are much, much more complicated than this model allows.

But the model does give an intuitive representation of the dynamics of political compromise, at least as it is typically understood. In Fig 1, we can see there are two configurations that yield roughly equivalent value peaks at points A and C, and a smaller peak at B. In other words, there are a few distinct configurations of control that yield good political arrangements, one slightly less optimal than the others. We can imagine that some people in the population prefer A and work to push the system to the left, while others prefer configuration C and try to push the system right. These combinations of interests might leave us at point B, a so-called “local optimum”. It isn’t the best value we can get in the system, so the people working towards either A or C will be dissatisfied. But B still has the best value in its immediate neighborhood, and any small moves left or right will decrease the overall value. So if we find ourselves at point B, we’ll tend to prefer staying there rather than moving left or right. In this situation, we’d call B an “attractor”.

This is, more or less, our intuitive understanding of political compromise. The leftists would prefer A, and the rightists C, but since we can’t make everyone happy we’re ultimately attracted to the compromise position B. Everyone is a bit dissatisfied by this arrangement, but it is better than any nearby alternative. On this model, the centrist is the person who advocates explicitly for B, despite it being sub-optimal, precisely for the sake of the better compromise.

In Fig. 1 I’ve generously drawn points A and C as equivalent in value, to make the compromise position appear as neutral and attractive as possible. But it is also unrealistic for describing our situation. While trying to remain neutral, we might give a more accurate description of the real world as follows:

In Fig. 2 we’re dealing with the same simplified model of control and value, but a more realistic political landscape. Point A is technically a local optima, but its neighborhood is almost flat, and small changes left or right don’t do much to affect the value overall. However, this plateau of value is flanked on both sides by dramatic plunges in value, labeled “pits of despair”. In this picture, politics isn’t so much about optimizing value as it is about keeping us from plunging into despair.

In other words, most political work is about keeping us within the brackets labeled “politics as usual”. While this confines us to dismally suboptimal performance, it still looks a lot better than the pits of despair. At the same time, fear of falling into these pits keeps us from pursuing anything else of greater value on the other side. In this more realistic setting, the centrist is the person who advocates for politics as usual, not for the sake of sup-optimal compromise as before, but simply to avoid despair.

It’s fairly easy to understand why people would be frustrated and upset for getting stuck in a rut like this. It is also easy to see that while centrism is a much more reasonable option in this scenario than in Fig 1, people in a political rut will be much more likely to radicalize and become extremists. Despite the risks, there’s the possibility for great rewards. With this picture, it is easy to explain how Trump’s election win has pushed us outside the brackets of “politics as usual”, and is threatening to send us down a pit of despair, while still explaining how it can seem so attractive an opportunity to some, just for the chance at anything else.

The natural question at this point is: how do we return to a stable position of value and avoid despair? Do we lean into the radicalization and try to make it over the canyon and up the hill? Or do we make a hasty backtrack and try to return to the frustrating but relatively safe political norm within the brackets?

If these are our options, the situation looks hopeless. There’s no reason to expect that we can avoid the pits of despair. Given the people attempting to knock us out of this rut, accelerationism is not a viable option. And as I said at the beginning, attempting to return to a comfortable norm is unlikely to work, and Fig 2 gives some intuitions why. If we’re rolling to point B and want to reverse our course, there’s not a sufficiently distinguished attractor at A to expect we’ll stop there again as a stable state. It looks equally likely that we’ll blow past the “center” and fall into the pits on the other side.

I think this has been a clear and important lesson from Clinton’s loss on Tuesday: the center does not hold. I was convinced that Clinton would win not just because all the sources in the media said she would, but because I though the center was the dominant, default position in the discourse, and Clinton its archetype representative. I thought the center was a strong and stable enough attractor to overcome the radicals at the extremes; the Democratic primary results were as good a confirmation of this assumption as any.

But the centrists lost big, and the stability we’ve come to expect from politics as usual has vanished. All bets are off. The center has shown that it doesn’t hold and can’t be trusted for support. So in a political landscape like in Fig 2, moving back to point A is not an option. Are there any options left?

Yes, of course. This is a new chapter. All bets are off.

History is ours to write as we see fit, regardless of traditions and customs of chapters past. And the dynamical discussion above, simplified as it was, gives some obvious suggestions for improvement beyond the limitations of our existing politics.

For instance, one problem with the model of compromise described above is that we’re all trying to optimize the configuration value of the entire system together, all at once! For instance, in Fig 1, we’re either at points A, B, or C. The attraction of compromise position B arose because, the argument goes, we can’t have both A and C simultaneously, because we can’t have two distinct configurations of power at once.

But why can’t we? This is our chapter of history, damnit! Why can’t our political system accept multiple, conflicting configurations of power and control at once, to allow for a diversity of values and structures to be represented? In other words, why can’t we build a political landscape like this:

Fig 3. Decentralization means no center. No center means no compromise is necessary.

The simplified model of control and value is the same as the previous images, but again the political landscape has changed. In Fig 3, there are multiple peaks AND valleys, describing multiple, configurations of control yielding roughly equivalent value. With Fig 1 we also had multiple peaks and valleys, but we assumed that we must decide on a single, universal configuration, which motivates the attraction of compromise. Abandoning the assumption of a universal configuration can be thought of as a way of “decentralizing” the political space. The alternative depicted in Fig 3 shows multiple distinct peaks of value, each of which might be considered a viable configuration of political control.

In this landscape, the goal of politics is not to unify around any single configuration, to rally around one peak at the expense of the rest. In this space there is no reason to prefer one configuration over any other. Instead, politics becomes an effort to protect the configurations around each peak, so that users attracted to that peak don’t fall off into the nearby valleys. In other words, multiple distinct peaks of value gives rise to multiple distinct forms of “usual politics” that bound the peaks and guard against the valleys. We can call these local politics “zones of protections” (or “safe spaces”, if you will) wherein the configuration of values particular to each peak is cultivated uniquely. In each bracketed space are a range of control configurations with relatively optimal value, and politics becomes the local affair of nudging that community around that specific peak. And the same goes with the communities attracted to every other peak.

In this landscape, not only are there more local optima to decide between, there are also a lot more pits of despair! A centrist whose primary concern is avoiding despair might find this new political landscape intolerable, because the ways of falling into despair have multiplied substantially, and the resources to prevent despair must be divided among more protected zones. In this sense, the new political landscape seems a lot more dangerous than either of the previous figures.

Furthermore, in this landscape, there is no universal center to advocate for. There are only local optima, and no universal standards by which to prefer one optima over another. So there’s no simple compromise position to advocate for, and no obvious attractors beyond someone’s particular interests. In this landscape, the work of the universal centrist must be transformed into the work of a border particularist, scrutinizing the minutia of control around each protected political zone. This is much harder and less appreciated work to conduct.

The centrist’s goals are upended in this new political landscape, but so are the extremists! Indeed, extremism gives no clear orientation on this graph. Unlike either Fig 1 or 2, moving to the extreme left or right in Fig 3 doesn’t, on the whole, improve one’s situation very much. To do well in this landscape you don’t need extremism, you need specificity. Pick some particular peak to explore and optimize. Making wild moves in any direction is unlikely to help.

Even on this simple model, the political landscape represented in Fig 3 is much more chaotic and confusing than the previous configurations. Maintaining multiple distinct configurations of viable political protection simultaneously is a lot of hard work. But I hope this discussion gives some intuitive reasons for thinking that it might be worth the effort. It keeps us from falling in a political rut, and it undermines the compulsion to lazy centrist compromises. It also limits the effectiveness of extremism, and undermines the motivation to radicalize, while still fully respecting the specific and unique values of distinct communities.

These all sound like positive things to me. These possibilities become available by thinking about politics from a dynamical point of view, and without the constraints of the convention wisdom, which can clearly no longer support the load. What other insights might accrue from thinking about politics in a new light? What alternatives can we construct?

Daniel Estrada

Written by

robot. made of robots.