Oddly Influenced

Welcome to oddly influenced, a podcast about how people have applied ideas from *outside* software *to* software. Episode 13: David Graeber and gift economies, or: yeah, open source culture ain’t that simple.

I have two primary sources for this episode. The first is anthropologist David Graeber’s /Toward an Anthropological Theory of Value/. The second is chapter 5 of his later book /Debt: The First 5000 Years/.

Graeber’s /Debt/ and his later /The Dawn of Everything/ (with David Wengrow) were somewhat controversial. He’s been accused of drawing conclusions that go beyond the evidence and some sloppy use of citations. However, those accusations seem to be about him dipping into other fields (like the history of the Enlightenment). I’m not aware of him being slammed when speaking about his field of expertise, anthropology. In particular, the first book predates his time as a “public intellectual”, which was the role he’d consciously assumed for /Debt/ and /Dawn/. That first book was a book by an anthropologist mostly for other anthropologists, so less prone to oversimplifications.

Another thing about him you might want to know is that he was a left anarchist of the sort that protested the 1999 World Trade Organization meeting in Seattle. He was also one of the leading figures of the Occupy Wall Street movement of 2011.

So, you might guess that he was not a fan of what’s sometimes called “the neoliberal consensus” and its underlying worldview is that humanity is best seen as utility-maximizing individuals. I’m going to adopt his point of view as I cover his work. You’ve been warned.

I’m going to use Eric Raymond’s essay, “Homesteading the Noosphere” as a jumping-off point. Although I’m going to criticize it, I do recommend you read it. Even if you’re as skeptical of his conclusions as I am, it has some value as – and I hope this isn’t seen as too insulting – an example of an insider explaining his own society. As Graeber remarks in both of his books, people who do that typically do an inadequate job. Quote: “Bordieu has long drawn attention to the fact – always a matter of frustration to anthropologists – that a truly artful social actor is almost guaranteed not to be able to offer a clear explanation of the principles underlying her whole artistry.”

Or, a longer quote:

“[F]eudalism was a notoriously messy and complicated business, but whenever Medieval thinkers generalized about it, they reduced all its ranks and orders into one simple formula in which each order contributed its share: “Some pray, some fight, still others work.” Even hierarchy was seen as ultimately reciprocal, despite this formula having virtually nothing to do with how real relations between priests, knights, and peasants operated on the ground. Anthropologists are familiar with the phenomenon: it’s only when people who have never had occasion to really think about their society or culture as a whole, who probably weren’t even aware they were living inside something other people considered a “society” or a “culture”, are asked to explain how everything works that they say things like, “This is how we repay our mothers for the pain of having raised us,” or puzzle over conceptual diagrams in which clan A gives their women in marriage to clan B who gives theirs to clan C, who gives theirs back to A again, but which never seem to quite correspond to what real people actually do.”

In any case, Raymond describes open source culture as being one where quote “participants compete for prestige by giving time, energy, and creativity away.” He describes that as a gift culture, using as an exemplar the *potlatches* observed in the Kwakiutl society in the late 1800s and early 1900s, when individuals competed for what Raymond calls “reputation” in elaborate feasts where aristocrats competed to see who could give away the most stuff. He describes gift cultures this way, quote:

“Gift cultures are adaptations not to scarcity but to abundance. They arise in populations that do not have significant material-scarcity problems with survival goods. We can observe gift cultures in action among aboriginal cultures living in ecozones with mild climates and abundant food. We can also observe them in certain strata of our own society, especially in show business and among the very wealthy.”

and, quote:

"Abundance makes command relationships difficult to sustain and exchange relationships an almost pointless game. In gift cultures, social status is determined not by what you control but by *what you give away*. […] thus the hacker's long hours of effort to produce high-quality open-source code.”

This is mostly wrong. While the Pacific Northwest *was* an extraordinarily abundant ecosystem, gift cultures are much more widespread than that. And the Kwakiutl society most definitely had command economies. It was in fact what Graeber calls a “heroic society”, one in which a quote “Big Man” – someone wealthy and powerful – is surrounded by flunkies, sycophants, poets he gives money to, dinner guests who just won’t go away, and so on. One example that leaps to my mind is Odysseus of “The Odyssey”. He was, not to put too fine a point on it, a complete asshole by today’s standards. But as a leader in a heroic society, his role was to accumulate stuff, give it away to followers, and be *extremely* boastful while doing so. In return – though that’s not actually the right phrase, we’ll get to that in the next episode – In return, the followers were tools Odysseus could use however he wanted.

Another example would be rich end-of-Republic Romans, whose day started off receiving clients. Patrons would give followers money or grant favors. (The more enterprising citizens would have multiple patrons and rush from patron to patron to collect as much as possible in a morning.) In return, the clients, among other things, escorted the patron to the forum: the bigger the crowd of clients, the more important the patron. Especially in the later Republic, they also formed a useful mob that the patron could wield in the constant jockeying for power and “virtus” that made up the life of the Roman aristocracy. (“Virtus” is a particular Roman virtue that combines attributes like valor, manliness, excellence, courage, character, and worth.)

Another issue in Raymond’s analogy to the potlatch is the relationship between aristocrats. In open source, Rich Hickey’s reputation is entirely (or almost entirely) independent of José Valim’s. Open source is mostly a positive-sum game.

That’s not actually how heroic cultures work. To them, gift giving is akin to war. The Kwakiutl aristocrats Raymond compares to open-source project leaders referred to potlatches as “fighting with food”. They spoke of themselves as great mountains from which gifts rolled like gigantic boulders, and of how conquered rivals were reduced to slaves.

I *hope* that’s not the attitude Mr. Hickey and Mr. Valim have toward each other.

To my mind, the most grievous error of Raymond’s account is that Kwakiutl society was in the middle of a major crisis when the reports we rely on were written. Like other native populations in North America, a *huge* percentage of the Kwakiutl died from disease when European colonists came in contact with them. One estimate I’ve seen is that around 30% of the native population of the Northwest died from disease, though we can never really know.

Stop for a minute. Suppose 1 out of 3 people you know had all of a sudden died. That might have shaken how your society worked and how you behaved in it, no?

And don’t forget, your country is also occupied by impossibly powerful invaders who are intensely against your potlatch custom and will ban it fairly soon.

One bit of chaos was that there were now more aristocratic positions than aristocrats. It seems commoners were getting pretty uppity and wanting to occupy aristocrat positions. The aristocracy resisted. It was perhaps more able to resist because contact with Europeans allowed the already rich to get vastly richer. (Beaver pelts were involved.) So the aristocracy were simultaneously engaged in competing for titles with each other and preventing commoners from getting those titles. Things got out of hand. Let’s just say we know the reports are not of a culture in a steady state.

We do not know how these cultures worked before they were disrupted. Even during the disruption, we don’t know what the commoners were doing, how they interacted with each other, if they used gifts. It’s all about what the aristocrats were doing.

Commoners keep society fed and clothed. I think Raymond errs in a very typical way by focusing on the motivations of software project owners and, um, Big Men. It seems to me open source projects rely on “commoners” to adopt them and contribute to them. Focusing on encouraging many people to produce a few pull requests each might – *might* – be more important than focusing on the major contributors. I’m not sure. In my only truly successful open source project I was bad at it. But I see people like José Valim being really good at it. One gift he’s good at giving is gratitude for things like simple documentation improvements – not the way a Big Man behaves.

I don’t think Raymond has picked a great model for open source. Let’s take a look at how some other gift economies worked.

Another famous gifting culture was the Kula Ring of 18 islands in Papua, New Guinea. Like the potlatch, this is not a society-wide practice but is primarily for the upper classes. What people do is take canoe trips hundreds of kilometers to exchange gifts. As the first description of the Kula ring asked, quote "Why would men risk life and limb to travel across huge expanses of dangerous ocean to give away what appear to be worthless trinkets?" The answer is that the gifts’ value is not in what they’re good for or how hard they are to make.

The most important gifts are red shell-disk necklaces and white shell armbands. The necklaces travel clockwise around the ring, while the armbands travel counterclockwise. So I might travel to your island and gift you an armband and possibly other things. Some amount of other trading, including pure barter, might take place. The closing gift would be from you to me and would be a necklace. Then I’d go home.

Neither of us would keep our gifts. After a customary time, you would be obliged to go counterclockwise and pass on the armband I gave you to someone else. I would do the same. In this way, gifts travel around and around the ring.

Once we’d exchanged gifts, we’d have certain obligations of hospitality, protection, and assistance to each other. That’s a first-order reason to bother trading gifts, but not the only reason.

Some items have names and histories that are well-known. Possessing such an item enhances the (temporary) possessor’s status. “Status” isn’t really a great metaphor, though. By owning a valuable Kula, you attach your name to it. As the Kula moves on, your influence or potency – your ability to act on the world – moves with it. You have greater control over the world because people you’ve never met consider your name important, your actions significant.

Pause for a moment. This is a different way of thinking of one’s place in the world than the one you’re probably used to. Reputation isn’t something you *contain*; it’s something you *distribute*.

The Kula ring strikes me as fitting open source culture maybe a bit better than potlatches.

1. The idea of extending your influence or potency in the world via the thing you give away seems to match open source rather nicely.

2. So does the idea of sending your gift into the world and having its value circulate around back to you, increasing because of its possession by other people. And also: the person who gives more gets more: actual concrete gifts, not just reputation.

You get to participate in a big open source ecosystem where your contributions come back to you, not as reputation points, but as tools you can *use*. You cast your seed upon the ground, and it comes back to you a hundred fold.

3. Raymond describes the modesty of open source “alpha hackers”. A potlatch aristocracy is very much the opposite of modest. When the aristocrats gave a gift (or, though Raymond doesn’t mention this, ostentatiously destroy their possessions), they made sure you knew they could do that because of how great they were. However, the giver of a Kula gift always downplays the actual value of the gift, in a very exaggerated way. That’s far from unusual in gift cultures. Here’s an example from the Inuit.

quote

“The old man laughed. “Some people don’t know much. I am such a poor hunter and my wife a terrible cook who ruins everything. I don’t have much, but I think there is a piece of meat outside. It might still be there as the dogs have refused it several times.” This was such a recommendation in the Eskimo way of backwards bragging that everyone’s mouths began to water.”

However, the modesty Raymond depicts among open source developers is characterized as genuine versus the knowing wink of many gift cultures, so maybe this isn’t a good fit.

4. There is a notion of permanent(ish) ownership in the Kula ring. There is an actual person (or group of people) who owns the item, no matter where it happens to be. They can sell it or destroy it (presuming they’re the person holding it at the time, I think). This is independent of its travels around the ring. However, it appears to affect the people who (temporarily) possess the item: they should take care of it because it *belongs* to someone else. That might be akin to how Clojure *belongs* to Rich Hickey even though I can have all its source on my own hard drive.

All that given, I want to repeat that both the potlatch and Kula ring gift cultures are about the aristocracy. Most of us who use and contribute to open source are not aristocracy, so maybe we need some analogies that aren’t all about the motivations of project leaders / benevolent dictators.

Another thing that’s worth noting is that first book about the Kula ring was Malinowski’s /Argonauts of the South Pacific/, published in 1922, one hundred years ago. There’s been vigorous debate about what gift economies *mean* ever since, and I don’t get the feeling that the topic is settled among anthropologists. So it’s the height of presumption for us in software to think we can map anthropological understanding (which is contested) onto our own naive, internal understanding of what we do.

But “height of presumption” is probably part of the definition of “computer programmer”, so let’s continue. This time with a kind of semi-gift culture that’s reasonably common.

Here’s a quote

“A day or two after we reached Vana we found one of the natives very ill with pneumonia. Comber treated him and kept him alive on strong fowl-soup; a great deal of careful nursing and attention was visited on him, for his house was beside the camp. When we were ready to go on our way again, the man was well. To our astonishment he came and asked us for a present, and was as astonished and disgusted as he had made us to be, when we declined giving it. We suggested that it was his place to bring us a present and to show some gratitude. He said to us, ‘Well indeed! You white men have no shame!”

That seems bizarre to us, well, me. But there is a logic behind it, according to Graeber. Suppose that saving someone’s life puts you in a relationship like brother-to-brother. One brother, the saved one, then observes that the other brother is enormously rich in comparison. So when the rich brother is about to leave forever, is it unreasonable to ask him to give you something worth little to him that would be worth so much to you?

I’ve been peripherally involved in open source projects where the dynamic was something like that: aristocrat gives away this wonderful software system; users indicate it’s great but partly inadequate and say ‘this really oughta kinda be fixed’ (with at least a bit of an implication of a demand), and the aristocrat goes nonlinear in outrage that the result of his huge gift would be demands about how he should spend his time.

I’m not going to make a judgment about who’s right in abstract, logical terms, if that even makes sense. Perhaps in the next episode. For now, I just suggest that it seems we have here a part of human nature that won’t be fixed by getting mad at it.

I want to finish off with an example of what is the most common sort of gift culture, I think: gifts with delayed reciprocations, where the reciprocations deliberately don’t close out the relationship.

I once helped my neighbor fix his bicycle’s brakes.

More recently, he mowed our lawn while we were on vacation.

Do we maintain internal registers where we consider whether his debt to me is wiped out? Do we know the exchange rate between bicycle repair and lawn mowing? No, we do not. In fact, it’s important that we don’t. The exchange of gifts is something that creates and re-creates social structure; in this case, the social role of “neighbor” in whitebread middle-class American culture.

Here’s a more exotic example:

quote

Laura Bohannan writes about arriving in a Tiv community in rural Nigeria; neighbors immediately began arriving bearing little gifts: “two ears [of] corn, one vegetable marrow, one chicken, five tomatoes, one handful [of] peanuts.” Having no idea what was expected of her, she thanked them and wrote down in a notebook their names and what they had brought. Eventually, two women adopted her and explained that all such gifts did have to be returned. It would be entirely inappropriate to simply accept three eggs from a neighbor and never bring anything back. One did not have to bring back eggs, but one should bring something back of approximately the same value. One could even bring money – there was nothing inappropriate in that – provided one did so at a discrete interval and above all, that one did not bring the exact cost of the eggs. It had to be either a bit more or a bit less. To bring back nothing at all would be to cast oneself as an exploiter or a parasite. To bring back an exact equivalent would be to suggest that one no longer wishes to have anything to do with the neighbor. Tiv women, she learned, might spend a good part of the day walking for miles to distant homesteads to return a handful of okra or a tiny bit of change, “in an endless circle of gifts to which no one ever handed over the precise value of the object last received” – and in doing so, they were continually creating their society.

end quote

If it is in fact a good idea to say that there is *a* way open source projects work – and I’m not at all convinced there is, versus descriptions of how *this* open source project works – it seems to me better to focus on the way that a project is a society of sorts, and how *everyone’s* gifts help to sustain it. But then, I would, as I’m temperamentally allergic to pumping up the visibility of Big Men at the expense of all the smaller people who keep things running, and I’m especially allergic to the notion that status or reputation is a single quantity; one, it seems to me, modeled after the very special – but limited – role money takes in our larger society. People don’t have reputation that are usefully quantifiable. They have reputations *for* things. And they have different reputations for different things. There’s no universal medium of exchange here.

Graeber describes all societies as being made up of a mixture of three different types of economies. Different societies use them differently, in different proportions, but always in a way that continually re-creates the society (or, sometimes, changes it). All three economies contain gifts, but in different ways.

I’ll talk about those three economies next time and muse a little bit about how understanding them might affect your understanding of the little society of your software team. Or help change it.

Thank you for listening.