Interview
David Reich – Bronze Age shock, the Neanderthal puzzle, & the sudden spread of farming
- Ancient DNA field data involving approximately 16,000 individuals over the last 18,000 years indicates a massive intensification of natural selection over the past 5,000 years (Bronze Age) compared to the prior 5,000 years, driven by environmental shocks from farming, urbanization, and population density.
- Specific genetic shifts detected include a rapid increase in TIK2 tuberculosis risk variants from 6,000–8,000 years ago followed by a decline in the last 3,000 years, ABO blood system B variant rising to 10% at the expense of A during the Bronze Age, and peak depigmentation occurring between 4,000 and 2,000 years ago.
- Selection signals for intelligence and years of schooling are predicted to have peaked between 5,000 and 2,000 years ago with no significant evidence of selection in the last 2,000 years, contrasting with hunter-gatherer predictors that were approximately three standard deviations below the modern mean.
- Immune and metabolic traits experienced unusually strong selection during the Bronze Age due to higher population densities and contact with domesticated animals, while selection for obesity and type 2 diabetes traits has historically acted to reduce body fat under the "thrifty genes" hypothesis.
- Detecting natural selection requires sample sizes of tens of thousands to observe frequency shifts of a few percent or ten percent, with migration events causing detectable fluctuations that obscure selection signals, making isolation periods of a few hundred years optimal for detection.
- The human genome is described as reacting more strongly to events 5,000 years ago than to the initial transition to agriculture 11,000–12,000 years ago, with selection coefficients of 1% or more being effective even in smaller populations once a threshold size is reached.
- Approximately 40% to 80% of European DNA originated from Yamnaya steppe pastoralist migration 4,500 years ago, a process driven by cultural replacement rather than genetic mutation, with the "wrenching" of the Bronze Age considered qualitatively greater than the farming transition.
- The cognitive revolution 50,000 years ago is hypothesized as a polygenic adaptation involving a common ancestral population from 70,000 years ago, rather than a single mutation, with the genetic capacity for farming existing long before the 12,000-year Holocene climate stability triggered agriculture.
- A proposed model suggests modern humans contributed 5% of their DNA to Neanderthals 300,000 to 200,000 years ago, potentially explaining why Neanderthal mitochondrial DNA and Y chromosomes are modern human-derived, replacing the standard divergence model with a "collective intelligence" or cultural transmission framework.
- Current research trends anticipate the use of genomic data from 10 million positions to predict trait types, with selection statistics above a value of five providing calibrated estimates of real mutations, and future sequencing expected to reveal relationships between archaic and modern humans in regions beyond Europe and the Middle East.
- Statistical power in the dataset suggests 479 independent positions with 99% confidence and roughly 3,800 positions with 50% confidence are under selection, while latent genetic variation within the 8 billion population provides the "stretchiness" for rapid adaptation to new environments without a mutation limit.
- Challenges in detection include short timescales of 200–300 years being insufficient for observable natural selection effects, and the difficulty of distinguishing cultural phenomena from genetic mutations in population replacement events.