Dwarf Fortress: A Marxist Analysis

One of the centerpieces of Marxist theorizing about history, that philosophy of history generally known as ‘historical materialism’, is the succession of modes of production. Each mode of production is essentially a more or less integrated totality of social relations, one that has a stability and continuity determined by its particular division of labor and techniques of production. Each is reproduced on the basis of the ‘laws of motion’ of that particular division of labor and that particular set of technologies, never mind the mental conceptions of society and the role of each functional part within it that form such an essential part of the continuity and stability the material process of social reproduction has.

The classic periodization of history according to this model in Marx’s own day was the succession from savagery (or ‘primitive communism’) to ancient society, feudalism, and then capitalism, with the non-European societies making a detour via the ‘Asiatic mode of production’. After a century and a half of historical analysis and refinement, little of this edifice is now standing. But the two most studied modes of production are perhaps also the most generally accepted and most stable concepts: feudalism and capitalism. But how to understand these? For capitalism, one need look no further than Capital itself and the vast literature that has followed in its wake. Feudalism, on the other hand, is much less well described in its own characteristics – the central point of contention in the literature has been the debate of transition from feudalism to capitalism, much less describing feudalism itself with anything like the rigour of Marx’s masterpiece on capitalism.

Nonetheless, the game Dwarf Fortress is amenable to a Marxist analysis precisely by understanding its relationship to the central characteristics of the feudal mode of production.(1) This can be grasped by looking at the main dynamics of the game: the division of labor, the reproduction of dwarven society, the economic system and trade, and their integration. The first characteristic is the organization of dwarven labor. Starting out with a small exploratory team of settlers, one takes a particular section of the world that is more or less taken as terra nullius and makes it amenable to dwarven homesteading and regular life. This is the beginning of every game in Dwarf Fortress (assuming fortress mode), and it already gives a clear indication of the nature of social expansion under feudalism. The fact of the low population density of the world allowing such expansion is one such observation, but more important is the purpose of settlement itself.

What one does in Dwarf Fortress is create a colony of an existing dwarven fortress – you’re always sent out as a team from a much larger existing stronghold elsewhere, and your foreign relations with other dwarves are limited to that particular fortress, on the whole. Even though your settlement is independent and self-governing, and the relations with the mother fortress mostly those of trade, the purpose of the game in all its open-endedness can be nothing other than to create oneself in the image of the previous fortress. In other words, fundamentally in Dwarf Fortress you reproduce the existing structure of dwarven society on a merely quantitatively expanded scale. Allowing for the different resources in this or that part of the world, this resembles nothing so much as the colonies of the city-states of the ancient world, or the processes of settlement enforced on pagan Eastern Europe by the Franco-German feudal societies of the high Middle Ages. The goblins and kobolds who regularly harrass your fortress but do not impinge on your world as an equal counter-society are analogous to the more or less loose relations of early Medieval chieftainship that still prevailed in the lands not yet subject to Frankish reconstruction.

Now the organization of labor in a given fortress is essentially revealing of the nature of feudal society. Each dwarf, male or female, can equally be a worker at any task, and who does what is mainly a question of establishing a strict set of social conventions early on that limit each dwarf to a number of possible economic activities reproducing the whole. They live and die within this limited sphere of labor, and are identified entirely by it, being a ‘miller’, ‘miner’, ‘cheesemaker’, ‘planter’, or whatever. Unlike under capitalism, this process is a matter of a more or less organic arrangement enforced by the player as a top-down set of strictures, without the least competition between dwarves, let alone the appearance of such a thing as a labor market. In fact, in the present version dwarves are not paid for their work in money, but rather demand customary rewards in kind, such as high quality food and drink, decent living quarters, and valuable and pleasing decorations and furniture throughout the fortress. This is characteristic of feudal society’s bounding of needs by custom and convention and the strong role of reciprocity in maintaining the division of labor, especially given the technological constraints on mobility and on adjustment of production.

This aspect can also be seen in the hierarchy within this division of labor. Dwarven society is more egalitarian than the corresponding human feudal societies were, but nonetheless both an explicit and an implicit hierarchy is visibile. Certain social functions are accorded noble status, which in turn raises the dwarf’s expectations regarding the standard of living, expressed still purely in consumption and not at all in the accumulation of money as an expression of value, like under capitalism. Rather, the fortress’ manager, broker, mayor and similar notables will demand correspondingly more grand living quarters, dedicated dining rooms and offices, and so forth. This is none too different from the palaces and grand meals arrogated by those who fulfilled corresponding functions in our feudal society: the (higher) clergy, knights (always willing to turn robber to enforce their claims to status), landowners and the court, and so forth. In fact, a sufficiently large fortress will obtain the right from the mother fortress to appoint a baron, who is purely parasitical and has correspondingly greater consumption demands still – and in fact, refusing to make this step towards the deepening of feudal relations (essentially corresponding to moving from the early to the high Middle Ages) on the part of a colony is offensive to the home fort!

It is telling in this regard that previous versions of the game used to have an internal monetary economy, with dwarves receiving payment for labor and buying commodities collectively produced in turn. But contradictions occurred between the dwarves’ collective possession of the means of production and the commodified basis for such an economy, and it became unworkable within the game. The game’s designers have abandoned it, and its only remnant is the ability to produce coinage, fulfilling no real purpose. (Indeed, this right of monarchs only achieves true significance beyond questions of prestige in the mercantilist period of absolutism, arguably.) Trade exists, of course, as it does in all feudal societies. This takes the classically feudal form of long distance trade undertaken in dedicated seasons with definite trading partners, the trades themselves taking place at the equivalents of the yearly fair-sites of high medieval France and Italy. This trade is mainly a trade in valuable resources, luxury goods, and basic stocks, and is highly profitable to the mercantile capital engaged in it without in the least affecting the reproduction of the societies in question in the vast majority of cases.

Indeed, although the various goods do have prices denoted in a commonly accepted currency (say, dwarf pieces), in accordance with David Graeber’s observations on the nature of credit and exchange in precapitalist societies, this is merely money as a medium of account. No actual money changes hands, it merely serves to account for the commensurability between concretely different use values required for the purposes of such long distance trade. Of course, as one would expect for a society where even formal subsumption under capital has not yet taken place, this trade is almost exclusively a trade in excess goods – production over and above the necessities the fortress needs for its own reproduction. Occasionally, some essential raw material will be absent in a given area and be traded for, but this too is merely a case of equal exchange, without further effect for the division of labor itself. Trade taking place not just with other dwarves, but also with other peoples such as elves and humans complicates and diversifies this, but does not change its fundamentals.

The military structure of dwarven fortress is one somewhat removed from the most common feudal ones of Europe, perhaps corresponding more to the equivalent experiences of societies other than Western European and Japanese feudalism. This is to say, rather than particular military services being performed in exchange for land held as fiefdoms, the basis of dwarven military organization is conscription of dwarves normally engaged in other functions within the everyday economy, and is usually done according to a certain dispensability – so that in case of significant losses, as happens often in the low tech battles of this world, the social reproduction should not be too affected. Indeed, the captain of the guard has noble rank and makes demands on that basis, but exercises no particular power over the immediate production of goods in the society itself. He or she is more like a military retainer than a semi-independent feudal intermediary lord, more a janissary than a duke. In the military as with dwarven society in general, it is the aptitude and training of this or that dwarf that determines their role in the social reproduction process.

These skills often predate the founding of the fortress itself, and one can surmise therefore that they form part of long lines of the exercise of similar roles in dwarven fortresses by particular families. It is important to note here that dwarven fortresses, being colonies or outposts, expand more through migration and settlement than through natural expansion, although dwarves have many children and are sturdier against disease and child mortality than their human counterparts have been in the feudal era. Child labor is common, of course, although dwarves only come of age at a certain point in their teenage years and are not fixed in their social roles until then, which gives the dwarven society a certain leniency and flexibility for the young not found in most human precapitalist communities. This may, however, be owing to the nature of the game’s fortresses as colonies rather than the main settlements. These colonies attract migrants of great skill as well as many fortune seekers of little skill, and therefore differentiate quite quickly.

In effect, the dwarven class system is somewhat more similar to a caste structure, be it a less rigid one than those of the human world. Those fulfilling functions more or less essential to the fortress’ ongoing reproduction are often freed from any other duties and specialize in these higher roles, such as metalsmithing, gem-setting, mining, and woodworking, whereas the many dwarves on the lower rungs are engaged in a wider variety of menial tasks (cleaning, hauling, woodcutting). As with all feudal societies, it it the function of the caste that determines one’s status, not competition between individuals. Craft skills are the determining factor for individuals in the fortress, and a dwarven outpost prefers losing nine haulers to one master weaponsmith. One may presume that this leads to the corresponding social and religious differentiation between these classes over time.

In all this, the one holding the society together is the player. The player in Dwarf Fortress fulfils the Durkheimian role of the collective mental structure and representation of society and its division of labor, whether religious or in some other ideological form, that makes possible the society itself as a totality and structures it. The long count of centuries forming feudal society’s division of labor is shortened by the activity of the player in assigning the dwarves their roles in the colony. Similarly, the player incarnates, as it were, the central Durkheimian roles in feudal society of the court and the clergy. There is little religion in the game other than references to the dwarven pantheon, but the cultural norms of dwarven burial (they are very strict on this) and its dietary strictures and practices indicate some kind of presence of this kind.

It is however the player who appoints dwarves to the noble roles and who coordinates the military efforts as best she can, and the player who needs to make sure dwarves are sufficiently happy within their existing and essentially static cultural-normative framework that no excess of alienation or anomie might destroy the fortress – as indeed happens if conditions deviate too strongly from the dwarven expectations about their customary rights and positions! Also, in accordance with the always incomplete ‘fit’ of any society with its normative framework that is reproduced on an ongoing basis, it is impossible for the player to control the dwarves too directly. Rather, the division of labor and the specific tasks are set, but the dwarves themselves have to carry them out to the best of their ability, which is not always a lot. The player is not a god, but the concept of the gods, and is not the commander of the fortress, but the ideological authority of its social relations and the need to reproduce them. In effect, the player is lordship – in Marc Bloch’s conception of it as the social and ideological glue holding feudalism together.

There is no slow transition towards a capitalism emanating from the urban trade according to the views of the world systems theorists. Because the production of food and farming are simply a single part of the division of labor and its surplus is collectively shared as with all the produce of the fortress, this also does not form the basis for the emergence of a large feudal landowning class coming into conflict with the peasantry. In this sense, as mentioned above, conditions resemble rather those of early feudalism in Europe or of certain social formations in Asia than the late feudal agricultural specificity beloved of the Brenner school. This has closed off the two routes to capitalism identified by the contending factions of Marxist historiography, and assures a dwarven future that remains feudal, albeit a more egalitarian and flexible feudalism than was characteristic of Europe. Dwarves do eat, of course, but it is a mining-based feudalism more than an agricultural one, and perhaps this accords it its more collectivist structure – it is no doubt vastly more difficult to spatially differentiate a hierarchy within a mine.

However, perhaps the dwarves need to heed a warning from Engels: “While the chaotic battles among the dominant feudal nobility were filling the Middle Ages with sound and fury, the quiet labours of the oppressed classes all over Western Europe were undermining the feudal system and creating a state of affairs in which there was less and less room for the feudal lords. True, in the countryside, the feudality might still assert itself, torturing the serfs, flourishing on their sweat, riding down their crops, ravishing their wives and daughters. But cities were rising everywhere… In all cases, they were ringed by protective walls and moats, fortresses far stronger than the castles of the nobility because they could be taken only by large armies. Behind these walls and moats, medieval craft production, guild-bound and petty though it was, developed; capital accumulation began; the need for trade with other cities and with the rest of the world arose; and, gradually, with the need there also arose the means of protecting this trade… Judged by today’s standards, all these advances in production and exchange were of a very limited scope. Production remained confined within the pattern of guild craftsmanship, and thus itself retained feudal characteristics; trade continued to be restricted… Yet, petty though industry and the businessman remained, they were adequate to overturn feudal society; and they at least remained in motion, while the nobility stagnated.”(2)

Is this the image of the dwarven future? For now, its social relations and lack of monetization, its mining collectivism have yet prevented it from arising in full. Dwarven society takes on these characteristics in its developed form, but remains within the personal bounds of early feudalism in its political organization. But it seems a plausible perspective. We have a feudal Dwarf Fortress now – will we have an absolutist Dwarf Nation soon?

1) Throughout, the latest version of Dwarf Fortress (0.34.11) is assumed, unless explicitly stated otherwise.
2) Friedrich Engels, “The Decline of Feudalism and the Rise of the Bourgeoisie” (1884) https://www.marxists.org/archive/marx/works/1884/decline/index.htm


Amazing, very good read. From a long time DF player(6yr) i have some short points to add.

1. Dwarven culture is marginally protected from capitalism by the cultural desire to work ones craft.

There is no Farmer Dwarf looking back onto the fortress at the end of the day wishing he got the chance to bend wood or beat metal.

Its hard to explain exactly what im trying to say but essentially, how we see lower and upper tier classes does not exist(as strongly) in the fortress.

The Farmer will refine his craft, and be happy to be a farmer. Try and image the same social strength stereotypes and nobility of craft in a Farmer.

4. Each weapon can be seen as a craft, and you can become a lord of that craft to never be expected to do anything for the rest of your life besides filling that role.

BUT if the whole fortress has been poisoned by a forgotten beast that makes your blood boil until you explode, and the only people left is a Spearlord, metal-worker and 8 kids.

The Adults would get to work fulfilling their new roles as food providers. (And not be very happy about it)

3. Dwarves can commonly change craft in wartime for new needs or the craft is producing to much for the community to operate correctly.

Five Miners making stacks of raw ores behind the smelters and having 2 Dwarves doing their best to upkeep the income of metals. A Farmer may be offered the opportunity to change his craft.

Please also remember in DF terms THIS IS A MASSIVE LIFE DECISION.

You are no longer “Farmer Daton” you are “Smelter Daton”.

Thanks for the read.

I loved this and I’d love to see a similar analysis of the ideological underpinnings of Civilization which, esp in light of the Engels quote, seems to introduce cities + city trade way too early to human development — not to mention additional wrinkles vis-a-vis specialized merchant citizens, agrarian collectivism, the simultaneous presence of economic stages in tech.

Leave a Comment

Your email address will not be published. Required fields are marked *