Have humans always sold and purchased things? This seemingly trivial question exposes one of the most conspicuous blind spots in our understanding of cultural evolution: the emergence of what we perceive today as ‘modern’ economic behaviour. Here we test the hypothesis that consumption patterns in prehistoric Europe (around 2300–800 bce) can be explained by standard economic theory, predicting that everyday expenses are log-normally distributed and correlated to supply, demand and income. On the basis of a large database of metal objects spanning northern and southern Europe (n = 23,711), we identify metal fragments as money, address them as proxies of consumption and observe that, starting around 1500 bce, their mass values become log-normally distributed. We simulate two alternative scenarios and show that: (1) random behaviour cannot produce the distributions observed in the archaeological data and (2) modern economic behaviour provides the best-fitting model for prehistoric consumption.

人类总是买卖东西吗？这个看似微不足道的问题暴露了我们对文化进化理解中最明显的盲点之一：我们今天所认为的“现代”经济行为的出现。在这里，我们测试了一个假设，即史前欧洲（大约公元前 2300-800年）的消费模式可以用标准经济理论来解释，预测日常支出呈对数正态分布，并与供给、需求和收入相关。基于横跨北欧和南欧的金属物品大型数据库（ n = 23,711），我们将金属碎片识别为货币，将它们视为消费的代理，并观察到，从公元前1500 年左右开始，它们的质量值变为对数正态分布分布式。我们模拟了两种替代情景，并表明：（1）随机行为不能产生考古数据中观察到的分布，（2）现代经济行为为史前消费提供了最拟合的模型。