This page gives more detail on the methodology and key data for the "Has life gotten better?" blog post series.
On the types of evidence I've sought out
Emphasis on systematically collected data
I generally focus on systematically collected data. This means, essentially, that someone applied one consistent set of rules to scoring different periods of time on some metric. My main source is Our World in Data, which seems to aspire to collect and present as much of this sort of data as possible, and tends to be very clear and detailed about where its data is coming from.
When asking "Has life gotten better?", there is lots of evidence you could consider that doesn't fit this description. You could pore over depictions of different eras written by historians; you could review journals and diaries; you could reason about what life probably was like based on e.g. intuitions about how TV and the Internet make you feel today. My problem with this kind of analysis is that:
- It leaves a huge amount of room for interpretation. I expect it to be heavily colored by the pre-existing worldview and emotional orientation of the person doing it.
- I think the amount of variety and richness in past people's lives is beyond our ability to really understand and imagine it. I'm worried about romanticizing or demonizing past eras based on thin clues here and there.
- I think in order to really feel confident that we were interpreting the clues we have reasonably - and building a picture of the past anywhere near rich enough to compare with our picture of the present - one would need to spend years (maybe lifetimes) just to get a good picture of some ~20-year period in some particular geography.
Focusing on systematic data leaves out a lot - in fact, for most eras it leaves out some of the most fundamentally important questions, like "How happy were people?" and "How was mental health?" But I can gather and look at all of the data that's relevant and available, and readers can check my work, and we can know that we're all looking at the same non-cherry-picked sample of evidence, and trying to make an uncertain best guess. And to date, I haven't seen any other analysis that seems reliable enough to change the basic conclusion I've arrived at this way.
Ethnographic and archaeological data
When trying to understand quality of life a very long time ago (thousands of years), most of the systematic data seems to be in one of these two categories:
- Ethnographic data: close observation (by anthropologists) of the relatively few people who maintain or maintained a "forager" (synonymous with "hunter-gatherer") lifestyle in modern times. An advantage of this data is that it allows direct observations of existing societies; a disadvantage is that it requires assuming that the millions of pre-agriculture years resembled what we can observe from modern foraging societies. I have seen a number of theories of how today's forager societies might differ from history's:
- Steven Pinker writes that they "have survived as hunter-gatherers only because they inhabit remote parts of the globe that no one else wants. As such they are not a representative sample of our anarchic ancestors, who may have enjoyed flusher environments. Until recently other foragers parked themselves in valleys and rivers that were teeming with fish and game and that supported a more affluent, complex, and sedentary lifestyle. The Indians of the Pacific Northwest, known for their totem poles and potlatches, are a familiar example." (From The Better Angels of our Nature, chapter 2 - location ~1200)
- They often have access to alcohol, which prehistoric societies likely didn't. From chapter 1 of War, Peace and Human Nature: "Increased aggression due to intoxication has been noted for the Hadza (chapter 14), Casiguran Agta (Headland, 1989; Griffin, 2000), Slave (Helm, 1961), Micmac (Le Clercque, 1910), Mardu (chapter 13), among others, and the fact that the nonviolent Paliyan of South India purposefully shun alcohol to avoid its negative effects on self-restraint makes the same point (chapter 15). Needless to say, alcohol was not available in the [pre-agriculture environment]."
- They may be interacting with modern societies in a number of ways. The effects of these may include violence reduction - see the two "!Kung Bushmen" rows of this table, which show a fairly dramatic drop in violent death rates "after the establishment of a state authority."
- Archaeological data: fossils and other remains of ancient societies, which we can try to make inferences about. I have generally put less weight on this evidence than on ethnographic evidence; while ethnographic evidence has major limitations, archaeological evidence seems still harder to draw confident conclusions from.
- From what I've seen of archaeological evidence, it's often hard to know whether there are "selection biases" in what sorts of remains were found at a given site. For example, there may have been burial sites specifically for people who died in war, or not in war - for an example of what I'm concerned about, search this page for "Crow Creek."
- Even for seemingly straightforward uses of archaeological evidence, like determining the share of deaths due to violence, I've had trouble finding sources that clearly explain how they deal with concerns such as (a) fatal vs. non-fatal wounds; (b) deliberate vs. non-deliberate wounds; (c) the "selection bias" concern above. I've also seen some sources that seem to make big leaps from their archaeological observations to conclusions about lifestyles (for example, going from "Neanderthals in a particular area don't seem to have eaten the sorts of foods traditionally gathered by women" to "These Neanderthals had egalitarian gender relations").
Some key sources I've used
Our World in Data
Our World in Data seems to aspire to collect and present as much systematic data as possible, and tends to be very clear and detailed about where its data is coming from. I've found it particularly useful for relatively recent time periods.
The Lifeways of Hunter-Gatherers ("Lifeways")
The Lifeways of Hunter-Gatherers (Kelly 2013) is a 2013 edition of a 1995 book.1 It has a number of tables summarizing evidence about hunter-gatherers, as well as a lot of discussion of what theory would predict about them. This is my main source.
It is the most systematicsource I've found overall for ethnographic data on foragers: most of its tables seem to be making an effort to list comparable figures for as many different societies as possible (while citing sources), and it generally seems to base its statements on factual observations and patterns across multiple societies rather than anecdotes.
Lifeways does not much discuss archaeology; it's focused on ethnographic data.
(I've also skimmed through all the papers that cite Lifeways using Google Scholar, and pulled out relevant-looking ones, though I didn't find any more comprehensive or recent literature reviews by doing this.)
Ecological Determinants of Women's Status Among Hunter/Gatherers ("Ecological Determinants")
In my attempt to understand pre-agriculture gender relations, I started by reading the chapter on gender relations ("Men, women and foraging") in Lifeways (discussed above). Reading the chapter and looking for promising leads, I found Hayden, Deal, Cannon and Casey (1986): "Ecological Determinants of Women's Status Among Hunter/Gatherers."
I generally prefer to use more recent sources (I think social science has improved a lot since 1986), but with that aside, this looks like a pretty good source: it presents a table of 33 different societies, scored on 13 different properties such as whether a given society has "female voice in domestic decisions" and "possibility of female leaders."
I searched the papers citing this paper, and the rest of the "Men, women and foraging" chapter from Kelly's book, as well as Google results and Wikipedia pages relating to hunter-gatherer gender relations. I didn't find anything else that provided a comparable systematic look across multiple societies, using multiple concrete dimensions of gender relations.2
Appendices to "How violent were foragers vs. pre-state societies vs. early states?"
Appendix 1: criticisms of the case that very early societies had high violent death rates
I've come across a fair amount of criticism of the data implying that very early societies had high violent death rates via extensive efforts to find commentary on Better Angels.
My main critical sources are Hunter-Gatherer and Human Evolution by Richard B. Lee (largely a critique of Better Angels of our Nature) and War, Peace and Human Nature (edited by Douglas Fry, referenced by many as a source of arguments for the peacefulness of earlier societies). In addition to emphasizing the distinction between nomadic and sedentary societies, these sources make the following points:
- Modern forager societies aren't necessarily representative of pre-agriculture societies.
- With this point in mind, authors have sometimes suggested focusing more on archaeological data, which I'll discuss below.
- Of particular importance: alcohol may have played a major role in modern forager-society violence.3
- However, I note that this point could cut both ways: interaction with modern society could have lowered violence among modern foraging societies, as suggested for the "!Kung" (Ju/'hoansi) in this table. In fact, this seems more likely to me overall, given relatively lower rates of violence in modern societies.
- Bowles 2009, the main source for the forager violence figures given in Better Angels of our Nature, is error-prone, as discussed in the appendix. It appears that out of Bowles 2009's 8 listed societies, two (the two with the highest violent death rates) contain major errors: the first chapter of War, Peace and Human Nature states that one was actually a rate for foragers killed by non-foragers, and that the other combines these sorts of deaths with forager-forager violence.4 Our World in Data corrects for these issues,5 but these issues still give me pause about the general reliability of these sources.
However, what I haven't seen is an alternative set of figures (e.g., share of violent deaths) proposed. (I also note that the Ju/'hoansi figure comes from Lee himself.)
Finally, I'll comment that the violence figures for foraging societies are generally based on small numbers of violent deaths, set against small populations. This means we probably shouldn't be too confident in them. On the other hand, it could explain how foragers could have "peaceful" reputations. As Lifeways states:
<blockquote>Note that the actual number of murders is low – the San Ildefonso Agta rate of 129/100,000 is based on eleven murders (including at least two by outsiders) over a forty-three-year period, or about one murder every four years (Early and Headland 1998). Visit the Agta or the Ju/’hoansi or the Hadza most years and you too would label them a “harmless people.”</blockquote>
Appendix 2: a note on archaeological data
Throughout this post, I have mostly emphasized ethnographic data (data of societies thought to be representative of our distant past) rather than archaeological data (data from analyzing archaeological remains). This is because:
- It seems there are a lot of potential pitfalls when it comes to inferring the share of violent deaths from archaeological remains.
- It seems hard to be confident that a wound was fatal or deliberate.
- At the same time, violence might be underestimated due to incompleteness of remains.
- I worry about bias in what sorts of deaths led people to end up with recoverable remains (for example, there may have been burial sites specifically for people who died in war, or not in war - for an example of what I'm concerned about, search this page for "Crow Creek"). When I've been able to find primary sources for this data, they haven't tended to address these concerns clearly.
- I haven't been able to classify any of the archaeological data as nomadic vs. sedentary.
- Archaeological data only addresses the "share of deaths due to violence" (not "deaths from violence per 100k people per year"), which makes it hard to compare to early historical periods.
That said, the figures from archaeological analysis are pretty similar overall to the figures from ethnographic analysis, implying that people in the distant past were very violent by modern standards - at least if we are not distinguishing between sedentary and nomadic people. I reviewed some criticism of these figures from R. Brian Ferguson (chapters 7 and 11 of War, Peace and Human Nature), and overall didn't find it to change the picture much (details in Appendix 4).
Appendix 3: errors in one of the most-cited sources for violence among foragers
Here's the key table from Bowles 2009, with footnotes marking apparent errors:
|Society studied||Type||Share of deaths caused by interpersonal violence - original||Share of deaths caused by interpersonal violence - corrected|
|Ache||Nomadic||30%||Removed (figure refers to deaths caused by modern outsiders)6|
|Ayoreo||At least partly sedentary ("seasonal forager-horticulturalists")||15%||15%|
|Tiwi||Labeled as sedentary, but I believe they're nomadic8||10%||10%|
Appendix 4: Ferguson's critiques of archaeological data on violence
My rough summary of Ferguson's critiques of archaeological data on violence:
- Some of his critique is specifically about whether the sites give evidence of "war," as opposed to other forms of violence.10 For example, there is significant discussion of which structures appear to have been set up for defensive purposes. He emphasizes that evidence of organized warfare is scant in some areas, but abundant in other (often contemporary) areas, implying that "absence of evidence [for war]" could in these cases be evidence of absence.11
- He argues that about 1/3 of the 21 sites listed by Better Angels of our Nature should be disregarded (I didn't evaluate this claim as it didn't seem like it would change the overall picture), and that the rest are "unrepresentative" in that they show more violence than many other sites.12
- He describes a number of sites in Europe and the Near East. For a few, he cites "share of violent deaths" figures; for some, he simply argues that they are inconclusive or do not provide evidence of organized warfare.
I've spot-checked the sites from the last point that include "share of violent deaths" figures. Most are hard to access. At least one appears to be excluded from the Better Angels of our Nature / Our World in Data data set for good reason,13 but at least one seems to me like it probably should have been included.14 Unfortunately, Ferguson doesn't provide his own aggregative summary of the surveyed sites, with "share of violent deaths" figures different from Pinker's.
Overall, I've had a hard time making sense of the archaeological analyses. (Ferguson's case for a peaceful past seems particularly focused on potential periods of post-agriculture, state-enforced peace);15 and I haven't found any source that lists figures in aggregate for some comprehensive set of sites with clear inclusion criteria.
My overall guess, though, is that the high-violence picture given by Our World in Data is closest to the truth. This is because it seems to have multiple pre-agriculture sites (from North America), and Ferguson lists relatively few sites with lower-violence figures (and most seem post-agriculture). In any case, I certainly don't think that archaeological evidence gives us affirmative evidence that pre-agriculture violence was on the whole low. I think one could see it as inconclusive, or as suggesting higher rates of violence pre-agriculture than today.
(Note that Ferguson's piece is critiqued in depth here; I haven't dug into this critique.)
Thanks to Luke Muehlhauser for referring me to this book. ↩
The closest was Sanday 1981, whose tables I found to be harder to interpret and generally using a "lower bar" for gender equality, and which did not seem to contradict what I saw in "Ecological Determinants." ↩
From chapter 1 of War, Peace and Human Nature: "Increased aggression due to intoxication has been noted for the Hadza (chapter 14), Casiguran Agta (Headland, 1989; Griffin, 2000), Slave (Helm, 1961), Micmac (Le Clercque, 1910), Mardu (chapter 13), among others, and the fact that the nonviolent Paliyan of South India purposefully shun alcohol to avoid its negative effects on self-restraint makes the same point (chapter 15). Needless to say, alcohol was not available in the [pre-agriculture environment]." ↩
"For two of Pinker’s cases, the Ache of Paraguay and the Hiwi of Venezuela/ Colombia, all of the so-called war deaths involved frontiersmen ranchers killing the indigenous people ... All 46 deaths used by Bowles (2009) to calculate so-called warfare mortality among the Ache are listed as “shot by Paraguayan” by Hill and Hurtado (1996, Table 5.1, pages 171–173)." ↩
Discussion here. Of the two societies whose original numbers seem based on errors, (a) the Ache are omitted from the data; (b) the Hiwi are included, but with a lower figure: "The paper distinguishes clearly between deaths inflicted by Venezuelans on Hiwis and death inflicted by Hiwis on Hiwis. Table 4 (page 450) shows the statistics on both groups and makes an additional distinction between precontact and postcontact times. In the precontact times 153 deaths are tabulated. There were 11 killings of Hiwi by Hiwi in the precontact times. Out of the 160 killings in the postcontact times 6 were killings of Hiwi by Hiwi. The Source that my source quoted states: “The Hiwi mortality proﬁle is characterized by notably high rates of violence and accidental trauma” (page 449)The numbers I quoted and the share of violent deaths is not counting infanticides that were very common among the Hiwi (and are reported by Hill, Hurtado (2007))." ↩
From the intro to War, Peace and Human Nature by Douglas Fry: "For two of Pinker’s cases, the Ache of Paraguay and the Hiwi of Venezuela/ Colombia, all of the so-called war deaths involved frontiersmen ranchers killing the indigenous people , a tragic situation that has nothing to do with levels of warfare death in nomadic hunter-gatherers during the Pleistocene." From Our World in Data: "Through my contact with critics of Pinker’s book I was referred to a critique of Pinker’s data by Douglas Fry. Fry criticizes some of Pinker’s data. I checked the original sources for the criticized information, referred to the criticisms and the original sources in my comments and corrected the data when the criticism was justified." And Our World in Data indeed omits the Ache case, and replaces the Hiwi case with a much lower share of deaths from violence (7.2% vs. 17%). ↩
See previous footnote. ↩
Youngberg and Hanson 2010 names the Tiwi as one of five societies the authors were able to find that best meet their criteria for "societies that most resemble the social environment where most human psychology seems to have evolved: small bands of nomadic foragers." Lifeways also labels the Tiwi "egalitarian" in Table 7-9. ↩
Our World in Data notes another error here: "Bowles (2009) based his calculation on ten killings when in fact 9 killings are reported." As in other cases, I stick with the corrected figure from Bowles (9/10 * 5% would be 4-5%) rather than using the figure Our World in Data got from another source. ↩
For example: "Of the adults, six (2.3 percent) showed signs of violent injury, with two probable deaths, one from a cranial blow, and one from a projectile. Th ere was no temporal clustering, but they all preceded the development of farming in the area (Roksandic, Djuric, Rakocevic, & Seguin, 2006, pp. 340–345). Roksandic (2006, p. 177) concludes that these are unrelated incidents that “could have as easily happened within the community as with members of other groups." ↩
"This ambiguous and extremely limited evidence of violence against or between humans lies in stark contrast to the thousands of highly explicit images in many caves of a wide range of animals that were found and hunted by the prehistoric residents (Giedion, 1962; Clottes & Courtin, 1996). Th ey also contrast with a florescence of clear images of warfare, conflict, fighting, and warriors found in cave art aft er 8000 BC (see Nash, 2005) ... considering how many scholars have been diligently searching for signs of violence, and considering how multistranded and convincing is the later evidence of war, usually without any dramatic increases in recovery (with exceptions, such as pre/post 3500 skeletons in France), is such a stance justified? Or is it a presumptive 'warrification' of the past?" ↩
""Pinker’s (2011, p. 49) List compiles data from Keeley and Bowles to include 21 cases. One case has no killings, and it will be shown that six more of the 21 cases can be tossed out. Th e others, valid cases of multiple violent deaths, will be shown to be a very selective compilation of high-killing situations, in no way representative of “typical” war casualties of prehistoric people in general." ↩
This study. Here's Ferguson's summary: "Four-hundred-eighteen individuals, including 263 adults, were recovered from six cemeteries on the right bank of the river, 8200–5500 BC. Of the adults, six (2.3 percent) showed signs of violent injury, with two probable deaths, one from a cranial blow, and one from a projectile. Th ere was no temporal clustering, but they all preceded the development of farming in the area (Roksandic, Djuric, Rakocevic, & Seguin, 2006, pp. 340–345). Roksandic (2006, p. 177) concludes that these are unrelated incidents that “could have as easily happened within the community as with members of other groups.” In contrast, downstream from the Gorge, at Schela Cladovei (7303–7545 cal BC ), of 57 individuals, 5 have fatal injuries, from skull fractures and projectiles, and 14 others have non-lethal trauma. Seven of the total are from Schela Cladovei III, and appear to be from the same time. Roksandic et al. (2006, pp. 345–347) infer that this represents a “localized and temporarily restricted” episode of warfare (2004, p. 72). Projectiles per individual are 0.0 percent at Lepenski Vir (n = 103) and .8 percent at Vlasac (n = 118) in the Gorge, and 10.7 percent at Schela Cladovei III (n = 28) (Roksandic et al., 2006, p. 117). (Roksandic et al. use these findings to argue against the idea that the Mesolithic was characterized by endemic warfare)." ↩
E.g., "Mace-heads are oft en the earliest weapon-tool in archaeological recovery. Yet mace-heads are oft en so small or lightly constructed that they appear to be symbolic. Symbolic of what? Of military prowess, or of legitimate authority —as used today by royalty, legislatures, and courts. Maces can be weapons of war, yet my university has a mace. Yes, there is an implication of power backed by force, but that can apply to mandatory decisions. Authority to settle conflicts is, as Hobbes illuminated, the very antithesis of war." ↩