Value capture is monocropping for the soul

The Score, Part IV

Welcome back to our philosophy of technology book club.

In May, we’re reading The Score by C. Thi Nguyen. Here’s the reading schedule for that book:

• May 4: Chapters 1-4

• May 11: Chapters 5-11

• May 15: Paid Subscriber Zoom Call, 8 PM Eastern (recording available)

• May 18: Chapters 12-18

• May 25: Chapters 19-24

• May 29: Chapters 25-29

• May 31: Paid Subscriber Q&A with C. Thi Nguyen, 3 PM Eastern

Notice that the final post is on Friday, May 29. That’s to give us a little more time to finish discussing the book by breaking up that week’s reading and having an additional post. Notice also that we’ll be joined by C. Thi Nguyen on May 31, where you’ll be able to ask him questions about the book directly.

In June, we’ll be reading The Ethics of Authenticity by Charles Taylor, and in July, we’ll be reading Pattern Recognition by William Gibson.

Paraphrasing the scholar of play Johan Huizinga, Nguyen writes that: ‘Play is a magic circle—a space segregated from the rest of the world.’ When we play, we enter an island of meaning. This is a place without consequences. How we behave, what our actions mean, is somehow segregated from the rest of the world, or, as Nguyen puts it on the next page, ‘The meanings of what happens in the game are quarantined from the rest of life.’ He uses a few other words to characterize this phenomenon, calling games ‘tightly bound’ and appealing to the fact that there are many games for many purposes—so, when we choose to play a game, we are choosing to temporarily inhabit one of these islands of meaning, where our actions taction on new significance, and what we do on one of these islands does not bleed out into the rest of the world.

It was in this chapter that I became skeptical of some of Nguyen’s claims, more skeptical than I’ve been throughout reading The Score. Chapter 19 is a short chapter and a romantic one, giving us an idyllic look into how Nguyen sees games from a highly personal point of view. And I think we can fairly say that games are romantic for Nguyen. They are little havens where he can adopt a new personality and a new set of values. They are a place of repose. But where I became skeptical is when Nguyen appeals back to his distinction between the striving player and the achieving player to explain how this quarantining works.

Let’s look to one of Nguyen’s examples here to work through my worries. Nguyen tells us that he often spends an evening playing a board game with his wife after their children have been put to bed. More dramatically: ‘We spend the rest of the night trying to kill each other. We play a board game.’ He’s taken by the fact that within the game, they can be fiercely competitive, while outside of the game, they can be nothing but friendly, even making drinks for each other. They can do this because they are both striving players, and within the struggle they can both get what they want, so in fact by trying to ‘kill’ each other they are playing the game better—they’re having more fun. The achievement player can’t do this, because all the achievement player enjoys is winning.

But I am not sure that the distinction is particularly explanatory here. Suppose Alice and Bob both like to struggle when they play; they’re really into striving play, just as Nguyen wants them to be. Suppose further they are playing the sort of board game that relies on duplicity, some sort of game where the point is to deceive your opponent. And let’s say that Alice is really good at it. I can imagine scenarios where Bob is troubled by the behavior that Alice displays through this gameplay. She’s just a little too good at lying, he starts to think. Even though Alice is playing by the rules of the game – she’s not revealing Bob’s most embarrassing secrets, she’s not stomping on his toes – Bob might be troubled by it. The quarantining is far from perfect, even when both are committed to striving play.

Here’s an embarrassing example taken from my own life. Over a decade ago, I lived in Boston when I was attending Boston University. A few friends of mine moved to Boston to attend graduate programs at other schools. One of them – I’ll call him A. – had just moved to the area to attend Harvard Divinity School, where he was going to train as a secular humanist chaplain. While he was in Boston and his family was in Texas, his mother died. We got together to drink and play games to make him feel better. We are all millennials, so forgive the cringe for a moment: we played Cards Against Humanity. The point of Cards Against Humanity is to be offensive and funny. It’s to push boundaries. We were all playing it well. But I was dealt a card that I didn’t know what to do with: Dead Parents. I didn’t know what to do, but I was a bit drunk (we all were), and I played that card. The action in the game, which was in line with the rules and the spirit of the game, broke us out of this little island of meaning, and I knew I had hurt A.

I later apologized, saying it was thoughtless—and it was. To his credit, A. forgave me.

What broke the quarantine of the game wasn’t that we were achievement players. There’s no point in being an achievement player in Cards Against Humanity. It barely requires any skill, and it doesn’t even require a sense of humor. We broke quarantine for, I suspect, two reasons: i) it reminded A. of something he was trying to forget, and (ii) I inadvertently revealed something dark about myself. I’ve known that my sense of humor can be mean—I’ve hurt people with jokes before. I didn’t know I’d make a joke about somebody’s recently deceased mother.

And I wouldn’t have made it if we weren’t in the game. I think that within the game, I saw it as a way to get us to laugh at tragedy, which is my preferred response. But it was too much and too soon, and it broke us out of the game.

I think we played Settlers of Catan after that, which helped the night go better.

Games are special, Nguyen will tell us, because they allow us to inhabit these islands of meaning, but these islands are apparently quite fragile ecosystems, and I’d like to have a fuller explanation of the limits and boundaries of games.

Metrics, on the other hand, ‘are an unbounded scoring system.’ While in games we can mostly freely adopt and freely discard these new rules and values, in an institutional mechanical scoring system, we find that we are always playing. Your credit score is a metric. It’s a number you probably want to make go up. While ‘games are a space where meanings can wander free,’ institutional mechanical scoring systems are games we can’t escape.

What happens when we want to use these metrics to make sense of the world? Nguyen uses the work of James C. Scott, particularly his book Seeing Like a State, to explain this.

When states as we understand them today, which are highly bureaucratic, centralized entities, emerged, they faced a problem. In order to exercise some control over a domain, you need to understand it. So, states began processes of gathering information, and they found that to better process information, it helped to simplify the domain.

Simplifications are useful. In Chapter 22, Nguyen briefly mentions that scientific explanations rely on simplified models in order to explain the world. A model that included all of the information about the domain would not be a good model; it would be too cumbersome to use, but it would also lack predictive power. (I am not a philosopher of science, so I am sure there is a more substantive story to tell about model-building in science; it has been years since I read on the subject.) But statecraft didn’t just build simpler models; states wanted to simplify the domain.

We all know the phrase ‘The map is not the territory.’ The model isn’t the domain, and you should never confuse them. But in early modern statecraft, Scott says, a centralizing force would try to re-render the domain so that it better reflects an easily processed model. The keyword here is legibility. Scott writes in his book:

Legibility is a condition of manipulation…Whatever the units being manipulated, they must be organized in a manner that permits them to be identified, observed, recorded, counted, aggregated, and monitored. (Seeing Like a State, pg. 183)

So, states found ways to better organize the units in the domain. One paraphrased example from Scott helps here. A Welshman appearing in an English court was asked to identify himself, and he did as he always had, using the patronymic prefix ap. Your ‘last name’ is your father’s name after an ap in this example, and you can string them together: Owen ap Arthur ap Llewelyn ap… Through a series of these clauses, the Welshman could uniquely identify himself while also telling his ancestral story. In his village or town, this is an information-rich description, but it makes for cumbersome processing. The English court gave him a standardized surname. States around the world had to standardize practices like measurement and naming to better process the units. They also found that in order to better process/understand agricultural production, monocropping was preferred over polycropping. In forestry, if you want to estimate timber yields, it would be convenient for the forest to have only one sort of tree—so some states went about ‘standardizing’ the trees in their forests. Suddenly, the domain was more easily understood, and you could ‘score’ forests based on timber yields, but in order for this to happen, you had to change the forest.

We are the forest in this example. Through standardized metrics, our values are simplified: better able to travel across contexts, better able to be ranked and compared. But we lose something; we’ve explored this over and over in the book. As Nguyen writes: ‘Value capture is monocropping for the soul.’

Source

Before we move on to the question of what we do about it, which Nguyen gets to in the final few chapters of this week’s reading, let me raise one other concern about the argument. I’ve found that as Nguyen has continued to return to his exploration of games, he’s said some things that he does not adequately defend.

I thought this especially when Nguyen discusses the rules skeptic on pages 240-241. He describes this person as being unwilling to enter new worlds – these are social worlds, which I think we can say are partly constituted by the rules governing them – and calls them ‘stodgy’ and ‘unwilling to try anything new.’ He even says that rules skeptic won’t try new hobbies or arts. I’m really not so sure—and it was only through interacting with people during this book club that I came to be skeptical.

Adam is a regular commenter on Commonplace Philosophy, and a few weeks ago he wrote that he does not enjoy games. I posited last week that this is because he is particularly sensitive to being manipulated. There’s no need to stress the negative connotation of ‘manipulation’ here—often it means being controlled in a duplicitous manner, but it doesn’t have to be. You manipulate dials on a radio, and in medical contexts, ‘manipulate’ means something like ‘treating a condition using only your hands.’ I think that some people are highly resistant to this sort of manipulation, even when it isn’t duplicitous; that’s why Adam doesn’t enjoy games. Perhaps he has this sense that when playing a game his desires are not his own. You could even say that this stems from a valuing of authenticity. But I’d like to think Adam is still able to be exploratory and creative—he just doesn’t want to do it via games. (Adam told me in our last Zoom call that he quite likes hiking.)

Which makes me think that Nguyen hasn’t adequately explored why it is that some people are so resistant, and his proposed explanation verges on a misdiagnosis.

---

Let’s turn to the question of ‘What do we do?’ Nguyen gives us three options:

1. Build better metrics. We could call this optimization.

2. Give up on metrics and their attendant centralizing institutions. Let’s call this decentralization.

3. Embrace more pluralism and try our best to keep metrics in their proper place. Nguyen calls this value federalism.

Nguyen’s preferred path is (3). He argues for this largely in the negative—(1) seems doomed to the same value capture cycle, and (2) involves embracing a radical statelessness that would doom many people, like Nguyen himself due to chronic health conditions. We want the virtues of large, powerful institutions without selling our souls to them and their metrics. Thus, value federalism.

Why is (1) doomed to fail? It helps to skip ahead to Chapters 23 and 24 before returning to value federalism.

Metrics have a way of presenting themselves as objective and clear. Nguyen calls the former objectivity laundering and the latter he describes as the seduction of clarity. Metrics are objectivity laundering, Nguyen says, when value-laden decisions are being obscured; these value-laden decisions are often made in the decisions of what to measure or in the swapping out of a fuzzier value for the more precise value. His example of weight and health is a good one. ‘Health’ is a fuzzy term, and it can be sensitive to our interests. Health also always involves tradeoffs, too; you can only optimize for one thing at a time, and you have to choose want to optimize for; reasonable people will make different optimization decisions. A swimmer will prioritize leanness in a way a rock-climber might not. Some people want to optimize for years lived, while others want to focus on the fuzzier value of quality of life. But it is all too common for being healthy to be practically equated with losing weight. You don’t have to be a radical to see why this is misleading. One, you can lose too much weight and make yourself quite ill. Second, maintaining a healthy weight is only one component of living a healthy lifestyle. But we often think that the number on the scale is the measure of our health.

I lost a lot of weight in the last few years. And I found that as I lost weight, I didn’t really care about the number on the scale. I cared about things like being able to play with my son for longer periods of time, or not having to buy new jeans so often. I did care about a new number, though: blood pressure. Every man on both sides of my family has high blood pressure, and for the first time in years I’m in the normal range. I care about that because I don’t want to die of a heart attack in my fifties. But blood pressure isn’t only improved by losing weight—that’s why I do more cardio now, too. What has happened as I’ve come to have a healthier view of my body is that my values are multi-layered and more subtle, and this better reflects the domain. That’s true even though these values aren’t so clear, and thus aren’t so easily turned into metrics. But as Nguyen says: ‘Sometimes vague language is better because it expresses the truth that things are unclear or unsettled.’

Now back to value federalism. It helps to remember what metrics are for. Metrics enable cross-domain communication. They also allow us to outsource for practicality’s sake: we outsource because ‘we just don’t have enough time to think about everything.’ Metrics allow domain-specific experts to give us an easily understood judgment about a domain, and often we do need to defer to those. I mentioned it before, but I always check something like Wirecutter if I need to buy a new appliance. Why? Because I don’t want to buy four vacuum cleaners, test them, and then figure out which one really suits my needs. Wirecutter does that for me, and so I outsource (part of) my decision-making to them. What we want to avoid is outsourcing our whole souls. When we defer to others, though, we are putting ourselves in a vulnerable position—but that vulnerability is part of trusting someone else.

So, under value federalism, we outsource that which is appropriate to outsource to centralizing forces, and we outsource nothing else.

I was thinking about the practical upshot of this, and one simple idea came to mind. Next time you need to make a low-stakes decision, try to do it without using metrics. Pick a restaurant without looking at the reviews. Pick a book without looking at Goodreads. Ask a friend for a recommendation, or pick up a book based on what it says on the back cover. It’s a little way of not outsourcing our values.

---

Here are some of my favorite comments from last week.

Given that we returned to value-ladenness this week, I want to highlight Jordan’s comment:

I found the example on pg. 210 about value-laden standards particularly illuminating:

“Diurnal time and standardized clock time serve deeply different purposes and ways of life. This is why the choice of timing systems is value-laden.”

Standards like this produce a convergence that I’m often not aware of. It feels like a neutral system, yet there is clearly a worldview promoted and facilitated by our agreement on that standard. It’s been fascinating to unpack some of those assumptions, like the mapping sound quality example he gives on pg. 212.

Nguyen gave me another example in conversation. Think about maps you would use to navigate American highways. These are information-rich, but they leave out so much. They do not, for instance, have a way of representing beautiful views; I don’t think any of them represent something as nebulous as the pleasure of the drive or the quality of the roads. There are reasons for these decisions—and we have to remember that they are decisions.

And this, from The Alchemist of Life:

The line about scoring systems producing convergence, not merely discovering it, feels important far beyond games. Once we decide what counts, people start bending themselves toward being countable.

You can see it in fitness, business, social media, even self-improvement. The metric begins as a tool, then quietly becomes the definition of success. Steps become health. Revenue becomes value. Likes become impact. None of those are useless, but they flatten the thing they were supposed to illuminate.

Maybe the question is not “are metrics bad?” but “what part of the human experience does this metric make harder to see?”

This theme is continued in Nguyen’s remarks on ‘outsourcing the soul.’ And I think the final suggestion of that comment is on the right track. I’d also ask: ‘What part of the human experience do we wish to keep fuzzy?’

I’ll admit I found this comment from Artie funny, so here it is:

I love sorting and categorizing immensely, possibly more than playing games. I am thankful to Thi and this book for making me feel safe/validated in wanting to explore why I love and hate games, and love and hate tracking and metrics.

For example, a few months ago I decided to find every game I own on PC (or for which I hold a temporary license, à la Steam/Epic) and categorize them into discrete categories, so that I could stop complaining that “I have nothing to play.” I started with very clear-cut tags, like “Completed”, for games I had completed all the way through, and “Games with Friends”, for games that are co-op or multiplayer only (like Left 4 Dead 2); but soon I ran into a roadblock. How could I compare an unplayed deckbuilder that I could play for 300 hours, and an unplayed puzzle game that might last only 10? If the goal of this was to find games to play, I had to make tags that reflected my purposes.

So, I resorted to less measurable, but much more valuable categories: things like, “I want to play this”, for games I am actually excited at the prospect of playing, and “I’ve had my fill”, for games I’ve played a bit, and I’m not interested in returning to. I ended up with 9 categories: some were discrete and rigid, others were much more qualitative and subtle.

I probably worked on this for a week or two, with ferocious intensity. I would get home from work, put on some music, and crank out a bunch of games. It was such a blast, to know I was making something so helpful to myself, and putting everything into nice, neat boxes.

Since then, I’ve played maybe... 3 games from the “I want to play this” category, in as many months. I haven’t even finished one game! I’m curious as to why I pursued the sorting with such vigour, yet I couldn’t muster the same energy to play the games themselves!