Men at work: hoes, ploughs, and steel

By Razib Khan | December 2, 2010 6:51 pm

Ancient Egyptian farmer ploughing a field

Recently several weblogs have pointed to a new working paper on the role of plough-based agriculture vs. hoe-based agriculture in shaping cultural expectations about male and female labor force participation specifically, and the differentiation of gender roles more generally. My first reaction was: “doesn’t everyone know this already?” I am a cursory reader of the anthropological literature and the assumption that a shift from relatively extensive hoe-based agriculture (e.g., slash & burn, gardening, etc.) to a more intensive plough-based mode of production seems to suffuse the literature. Before touching on the major points in the paper itself I did a quick literature search, on the order of five minutes, and found something from 1928 which already assumed the major parameters which are now being mooted today, The Division of Work According to Sex in African Hoe Culture. I read the whole paper, and it remains surprisingly relevant (though some of the terminology and frameworks are a bit dated naturally). Here’s a selection:

Eduard Han, to whom the ethnological study of economics owes a considerable number of important discoveries which have been published repeatedly and in varying forms, seems to have paid scarcely enough attention to the good work of the scholars who preceded him in the fight for the recognition of the outstanding position of women in the lower forms of soil cultivation. Steinmetz and quite recently Koppers,’ have pointed out that Buckland already attributed to the female sex the invention of the most ancient method of soil cultivation, or hoe culture…Here we find, in particular,a clearer statement of the arguments of Grosse, Bachofen, and others about the connexion of matriarchal society and lower forms of soil cultivation. Matriarchy and hoe culture are assigned to definite chronologically determined stages of civilization (older forms of the so-called ‘two class culture’, and later ones of ‘bow culture’). Koppers, of the Austrian branch associates matriarchy and hoe culture with these two civilizations….

It is not our business here to study in detail the researcheson the zones of culture,which may be regardedas successfulup to a certain point, though we shall have to refer to them incidentally. It suffices to state that a connexion between woman and hoe culture, nay more, between that social system where the woman rules, matriarchal society, and primitive soil cultivation is universally acknowledged to exist.

Ignore some of the terms and concepts which might seem loaded or outmoded today. Rather, observe that in 1928 a distinction between hoe-based and plough-based agriculture was widely accepted in terms of the cultural consequences. Why? Just take a look at an old-fashioned plough vs. hoe (at least old-fashioned compared to the sort of mechanized devices you can find in catalogs today):

Manual plough & hoe, credit Jonathunder

Both the plough and the hoe serve the same general purpose, to prepare and maintain fields for the purpose of agriculture. The plough has more up-front fixed costs, because of its greater size and complexity, as well as a greater maintenance cost in terms of usage, because it requires more effort. But it is far more effectual than the hoe in its task of preparing and maintaining soil integrity for a host of crops. In particular, the two staffs of life in Eurasia which are associated with intensive agriculture: wheat and rice. Additionally, the physical demands of the plough, as well as its coupling with work animals, means that in general men have an advantage in generating productivity from the tool in relation to women. This is just a function of biology, as men tend to be larger and have more upper body strength. In societies where the plough reigns supreme the men dominate the public cultivated fields (I’ve been to rural Bangladesh, and the men perform most of the back-breaking labor in the paddies, while the women often engage in processing at home as well as handicrafts). In contrast it turns out be a truism that in societies where agriculture is “women’s work” the implement of choice tends to be the hoe, where they are not at great disadvantage to men in terms of marginal product (though even in these societies men may clear forests and do other extremely physically demanding tasks to prepare the ground for the women, who will follow in their wake).

civAs implied above, the plough tends to be associated with farming in more populous societies at the heart of the Old World Ecumene, while the hoe is generally found in more “small-scale” societies. Presumably all plough cultures went through though a hoe phase, just as intensive agriculture which is squeezed up against the Malthusian margin generally is preceded by a period of extensive agriculture where land is in surplus to labor. But a key point is that cultures do not emerge and evolve purely through endogenous social forces, but are constrained by exogenous parameters. Not only are some regions more suitable to particular modes of production (e.g., farming, animal husbandry, transhumance, etc.), but the details of production can vary a great deal. Even in a nation as ethnically homogenized by centuries of top-down fiat such as China the divide between the north and south of the country is a structural constant due to different climatic regimes and geological conditions.

So what does the new paper add? I am not familiar with the economic literature in this area, so I do not know if the methods are novel in terms of their application to this subject, but the authors use a formal statistical framework to confirm the correlation between mode of agriculture and gender roles. They do not smoke out novel or new correlations, rather, they place a specific number so as to render the broad-brush observation precise. In this way ironically the economists who bandy about fancy regression models actually make a mental picture of social phenomena more accessible to those without specific “thick” local knowledge.

The Origins of Gender Roles: Women and the Plough:

This paper studies the historic origins of current differences in norms and beliefs about the role of women in society. We show that, consistent with anthropological hypotheses, societies with a tradition of plough agriculture tend to have the belief that the natural place for women is inside the home and the natural place for men is outside the home. Looking across countries, subnational districts, ethnic groups and individuals, we identify a link between historic plough-use and a number of outcomes today, including female labor force participation, female participation in politics, female ownership of firms, the sex ratio and self-expressed attitudes about the role of women in society. Our identification exploits variation in the historic suitability of the environment of ancestors for growing crops that differentially benefited from the adoption of the plough. We examine culture as a mechanism by looking at first and second generation immigrants with different cultural backgrounds living within the US.

In the paper they cite a 1970 book, Womans Role in Economic Development, for the hypothesis they’re testing. I just want to emphasize again that all the pieces of the model were already widely circulating in anthropological circles implicity, as is made clear by the 1928 paper I mentioned above. It is important to remember that we come to the new paper with a strong expectation in the direction of the conclusions which they found.

Below are two maps which show the distributions of plough vs. hoe societies:

no images were found

The first map illustrates usage by ethnic group in their database. It is obvious that there are many more hoe farming populations than plough farmings. But, many more as individuals are from plough farming cultures. Why the difference? In the latter case the analogy can be made between the replacement of hunting & gathering and farming; plough farmers simply had greater per unit productivity and so could support much higher population levels than hoe farmers in regions where crops amenable to their intensive techniques could flourish. In their sample “86% of the ethnicities did not introduce the plough, in 12.18% of the societies, the plough was used, and in 1.5% of the plough was not initially used, but it was adopted after European contact.” In Brazil today the vast majority of ethnic groups were not plough farmers in 1500, but the majority of Brazilians are descended from plough farmers (even counting Africans as hoe farmers it seems likely that there is a preponderance of European ancestral contribution to Brazil in relation to African). Remember that even if you count the dozens of Portuguese speaking ethno-racial constructs among Brazilians as ethnic groups they’re still outnumbered by the small-scale societies of the Amazon!

Though the real yield of the paper is its general quantitative finding, which I will get to, it has to be framed in the broader area-specific patterns. For example, Northern India and Southern India traditionally exhibit very different gender relations patterns, as well as agricultural traditions. From the paper:

Boserup reports data for regional differences in women’s activities in different regions of India. She finds that in the South, where the plough was not present, there is a high fraction of women in the labor force compared to the North (36% versus 26%). This difference is present not only in agriculture (40% versus 29%), but also in other sectors such as construction (17% versus 6%), trade (17% versus 6%), transport and services (27% versus 13%) and other industries more generally (17% versus 6%). The author also notes that “in India the connection between the work of women and the direction of marriage payment is close and unmistakable. . . In region where women do most of the agricultural work it is the bridegroom who pay bride wealth, but where women are less actively engaged in agriculture, marriage payments come usually from the girl’s family” also “In some of the farming communities in Northern India, where women do little work in agriculture and the parents know that a daughter will in due course cost them the payment of a dowry, it was customary in earlier time to limit the number of surviving daughters by infanticide. This practice has disappeared but nevertheless the ratio of female to male population in these districts continues to be abnormal compared to other regions of India and to tribes with working women living in the same region”.

428px-Aishwarya_Rai_CannesIn the 20th century India has exhibited a relative trend toward cultural homogenization. But ethnology and anthropology were well developed enough early in the 20th century that practices which would decline or disappear were attested for many mainstream groups in south India. The Nairs and Bunts for example practiced (and some continue to pactice) matrilineal inheritance (Aishwarya Rai is a Bunt). Interestingly, there are groups in northeast India of Tibeto-Burma or Austro-Asiatic origin which practice both matrilineality and matrilocality. Not only is inheritance defined through the maternal line, but men move to the villages of their wives. This is in sharp contrast to the patriarchal bias evident in the Indo-Aryan speaking societies of North India. The “evil mother in law” who torments a young woman who marries into the family, who eventually becomes the “evil mother in law” in her own right after she gives birth to a son who will inherit, seems almost a cultural trope in these societies. Women are “strangers” who are coopted into the patrilineages through their sons, and ultimately become some of the most egregious enforcers of the norms which perpetuate the hyper-patriarchy of North India.

A quick inspection of the map above will show that it is the regions where hoe agriculture is also practiced which are less patriarchal. The highlands of Meghalaya, the the states of Nagaland, Manipur, Tripura and Mizoram, are really extensions of the broader Southeast Asian zone of slash & burn extensive agriculture, so the trend exists outside of Southeast Asia. The “hill people” in Burma, Thailand, Laos, and Thailand, occupy culturally marginalized positions, and probably are a snapshot of a dynamic which was common across much of the Old World.

In any case, speaking of maps, look at this one which compares sex ratio and literacy rates in India:


South India has higher literacy rates, and more normal sex ratios, indicative of relatively equitable gender relations. But, there are more developed regions of North India, like Punjab, which have massive sex ratio skews, while the tribal dominated regions of Central India combine low levels of development along with relatively equitable sex relations. What you tend to see is that cultures which are the most developed and least developed have the most equitable relations between the genders, while those in the middle are generally more conventionally male-dominated.

ploughtable1That’s where the statistical framework presented in the paper is so useful: one can control for confounding variables. Basically what they seem to have done is integrated modern geodata with older ethnographic information. Not only that, but they weighted the geographical variation by population density. So what the paper is popping out are spatial correlations of the variables of interest. To the left you see the output from an OLS regression. The dependent variable is the extent of female labor force participation as the dependent variable (as six categories). The first row shows the effect of the introduction of the plough, with the second two columns introducing controls. Remember that more plough societies have become developed than hoe societies, and developed societies tend to have more formal female labor force participation. The data set is broken down by ethnicity here. Note that the R-squared is capturing the proportion of the variance of Y explainable by X(s).

Obviously ethnic group level analysis is going to be somewhat problematic, because the data sets aren’t as well attested on this scale (e.g., labor force participation of Naga women as a percentage value?). On the other hand, they had better measures on the country and district and individual level, because the World Values Survey, and a host of statistical organizations, track numbers internationally which can be related together.

no images were found

The pre-print is freely available, so I’ll point you to that to make sense of the numbers in the above tables if they mystify you. Basically they’re supporting the general anthropological hunch that plough farming societies tend to be more patriarchal, all things equal. The final results section shows that immigrants to the United States impart to their descendants the same values. I do not find this surprising, though again it is useful to peg the exact magnitude of the correlations as they do. The majority of the world’s population are no longer primary producers, but most are recent descendants of primary producers. (around 50% of Americans lived on family farms just before 1900).

Ultimately this goes back to the foragers & farmers debate. I have argued for years that the “traditional” and “conservative” values which emerged after the rise of agriculture, and crystallized during the Axial Age, are actually cultural adaptations to existence in the Malthusian mass societies which arose as the farmers pushed up against the production frontier. As the world “filled up” there was a necessary switch from extensive to intensive agriculture, and social controls needed to be more powerful so as to keep the masses of humanity in some sort of meta-stable equilibrium. The rise of institutional religions, conscript armies, and national identities, all bubbled up as adaptations to a world where a few controlled the many, and the many persisted on the barest margin of subsistence.

But this is not to say that the world before the ploughman was a paradise. I believe it too was Malthusian. Humans are animals, and naturally hit a “carrying capacity.” But there’s Malthusian, and then there’s Malthusian. There were differences between the lives of hunter-gatherers on the margin, and farmers on the margin. Some scholars hypothesize that interpersonal violence was more common, not less, among hunter-gatherers, because there was no central authority regulating conflicts. More frequent unexpected deaths by hunter-gatherers may have constantly relieved pressures on small bands (as well as the fact that there was a check on births due to the problems of moving more than one infant at a time in a roving band). In contrast, it may be that famine and plague were the means by which nature re-calibrated the populations of farmers, who culled themselves off in the course of their lives through violence far less often. This is not because farmers necessarily were pacific angels, rather, the powerful men who rose with the mass societies valued individuals as sources of tax. Too much conflict which reduced population would naturally leave less aggregate rent, so there was a strong incentive for the thugs in charge keep their sheep alive and producing.

The above is a rather materialist economic reading of power relations. One could create a narrative of moral evolution over time, and the expansion of the arc of humanity with the spread of universal religions. I think the two variables are related, and in any case the description of what happened remains the same. But now we’re in a third age. We’re not trapped by Malthusian parameters, because gains in economic productivity haven’t been swamped by  population growth. Rather, on the contrary a demographic transition has occurred across much of the world, producing mass affluence.

397px-Body_artWith mass affluence has come liberalism, post-materialism, and all sorts of ideas and movements predicated on self-actualization. The converse is that the traditional values and social controls necessary for the proper maintenance of human civilization during the millennia of the ploughman have come under attack, and those who defend them style themselves conservatives and reactionaries. Ironically the flowering of the individualist ethos has resulted in a reversion to the relative lack of mass conformity, which replicates the diversity across small-scale societies. I say mass conformity because individuals still conform within their own subcultures, it is just that the mass culture has shattered into many smaller constituent pieces (e.g., the rise of “urban tribes”).

We should not proliferate categories beyond what is needed. But the story of hoe vs. plough agriculturalists shows that a simple hunter-gatherer vs. farmer narrative does not suffice. In some ways the hoe agriculturalist remains more like the hunter-gatherer, and in some ways more like his or her fellow agriculturalist. The most polygynous societies for example are arguably those of hoe based agriculturalists, as well as nomads. In contrast, hunter-gatherers and ploughman tend to be more monogamous, at least in a genetic sense.

We need to evaluate human nature and society at its true joints. That may require more complexity than is pithy, but so be it.

Note: I used the British spelling “plough” because it looked right to me.

Image credits: Nomadtales, Georges Biard, kris krüg

CATEGORIZED UNDER: Anthroplogy, Economics

Comments (8)

  1. Zach K

    “Women and hoe culture”? What’s changed? 😛

  2. Chet Snicker

    Sir! Comport yourself! This is no place for foul imputed obscenities.


    C. V. Snicker

  3. A fine analysis. One nit and one deeper point.

    The nit:

    “The most polygynous societies for example are arguably those of hoe based agriculturalists, as well as NOMADS. In contrast, hunter-gatherers and ploughman tend to be more monogamous, at least in a genetic sense.”

    Many (but not all) hunter-gatherer societies are nomadic. You seem to mean to say “nomadic herders” or “herders” or “pastoralists” where you say “nomads.” (It isn’t clear that the polygyny practices of relatively settled herders differ much from those who are nomadic, but perhaps it does.)

    It would also be interesting to see where transhumane food producers fit on the continum.

    The deeper point:

    What is remarkable about this particular study, compared to what was known in 1928, is not just that it quantifies the strength of the relationship. It is that it documents the persistance of the cultural norms long after their materialist economic basis has ceased to exist in many cases, a point similarly made in Malcolm Gladwell’s book “Outliers.”

    It is hardly surprising that there are stark cultural differences between societies that currently plough (or hoe) or did so a generation or so ago.

    It is remarkable, however, that this is still the case three or four generations later, when very large proportions of the populations in these societies are not engaged in food production at all, use machines that make biological differences associated gender less relevant when they do still engage in agriculture in most cases, and in many casesmost of the population has never personally know a family member who was a farmer or herder of any kind. The world of 2010 is far more removed from its agricultural roots than the world of 1928 was, especially in the developing and undeveloped world, yet the cultural patterns persist.

    The endurance of these cultural patterns is astonishing strong, something dramatically illustrated by American political history, for example. The liberal v. conservative voting patterns of U.S. counties, for example, are almost identical to what the were 140 years ago. The political leanings of states as hawkish or dovish in foreign policy were almost identical in the 1780s to what they are today, yet in each case, the material economic reasons for those differences to arise in the first place dissolved long ago.

    This is highly counterintuitive, particularly compared to the patterns of immigrant assimilation in America. While first generation immigrants (i.e. those born abroad), are very culturally distinct from other Americans, the cultural and demographic differences between second generation immigrants (i.e. the first generation born in the U.S.) and the general population, is remarkably modest, and by the third generation (i.e. the children of U.S. born children of immigrants), almost indistinguishable apart from a few residual food or holiday traditions and their genetically inherited component of their physical appearance.

    The implication is that culture lives not so much in individuals, who can leap from one set of cultural beliefs to another in a generation or two, but in communities, who effectively assimilated newcomers by both birth and migration, over long periods of time, by means largely social rather than through government (which has also homogenized much faster than cultural norms have changed).

    It is also somewhat at odds with changes in fertility patterns in countries undergoing economic development, which typically see dramatic declines in the numbers of children born (and co-incident changes in intensity of religious belief) in periods of time less than a full generation (see, e.g., changing fertility rates in 20th century Mexico and Brazil and (more coercively) in China), which again suggest that it is possible for cultural norms to change very rapidly when the conditions are right.

  4. Douglas Knight

    Some scholars hypothesize that interpersonal violence was more common, not less, among hunter-gatherers, because there was no central authority regulating conflicts.

    Does your invocation of theory mean that you are not impressed by attempts to measure? Perhaps it is a topic for another post, but I would like to know your assessment of the attempts to measure violence.

    I’m not convinced by this theoretical argument because, while a central authority would want to suppress local violence, it might want war.

  5. The picture of the Egyptian plowing interested me. The details of the plow itself closely match the plows I saw in use in the highlands in Central America, in Latin, not Indian villages. A few wealthier village people there had advanced to John Deere-style moldboard plows, still usually drawn by oxen. The harness is attached to a yoke tied to the horns, rather than a chest-mounted collar. This is several agricultural revolutions behind state-of-the-art US agriculture in the early 19th century.
    These plows require a lot of human muscle power to use, shoulders and arms. I tried. It was hard work, even with the oxen doing the pulling. The men universally considered farm work too hard and dangerous work for women, even the manual hoeing, using hoes almost identical to those pictured above.

    As to the question of the reduction in violence, I thought this was pretty much resolved some decades ago. Violence has declined sharply, at least in Western nations, over the last few centuries. I am curious about the earlier times. How did violence wax and wane during the periods between, say, the paleolithic, neolithic and stages of early civilization.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar