古人口学是以考古出土人骨为核心材料，研究古代人口结构、动态及其社会背景的新兴学科。本文回顾了国内外古人口学的发展历程，探讨了理论与方法的发展，分析了古人口学研究中存在的问题，并展望了未来的发展趋势。国外古人口学始于20世纪30年代，历经了从简单的人口结构统计到生命表方法的应用，再到基于贝叶斯定理的年龄估计方法的改进和更多复杂的人口学模型的应用等阶段。国内古人口学研究始于20世纪80年代，理论和方法的起点较高，90年代研究范式逐渐形成，2010年以来重视方法革新。文章总结了古人口学对年龄估计方法的改进和人口学模型的应用与改进等方法和技术的新进展，讨论了古人口学中的理论问题，包括均变论、人骨样本的代表性问题、骨学悖论、因死亡人口与总人口不同而产生的概念误用以及新石器时代人口转型理论等。目前，国内古人口学发展仍较薄弱，未来需加强方法论创新、深化多学科合作，立足中国材料提出自主理论，推动学科的本土化发展。

Paleodemography is the scientific analysis of past population dynamics, utilizing archaeological human skeletal remains for reconstruction. This review traces the methodological evolution of these techniques on a global scale, with a particular emphasis on China, highlighting significant paradigm shifts in addressing two fundamental challenges: age estimation bias and the representativeness of skeletal samples.

Paleodemography first emerged in the 1930s, when early studies were largely confined to rudimentary statistical descriptions and comparative analyses of population sex and age structures. In the 1960s, the introduction of the life table method marked a turning point, eventually establishing it as the predominant analytical approach. However, in 1982, the methodological foundations of the field were profoundly challenged by a seminal critique that exposed systematic errors in adult age estimation. Critically, the critique revealed the “age mimicry” effect, whereby estimated age distributions unintentionally replicate those of reference samples. This revelation triggered four decades of innovation: Bayesian statistics became instrumental in refining age estimation methods. These efforts culminated in the 1999 Rostock Manifesto, which advocated for probabilistic methods. Transition Analysis (TA) emerged as a pivotal technique, leveraging the progression of skeletal traits to effectively mitigate bias. Concurrently, a fundamental shift occurred in the field of demographic modeling, transitioning from the life table method to parametric hazard models. Additionally, proxy indicators (e.g., D_5-14/D₂₀₊, D_0-14/D ratios) were developed to indirectly estimate fertility and growth rates when direct evidence was unavailable.

In China, paleodemographic research began in the 1980s, focusing on analyses of Neolithic cemeteries’ sex ratios and age structures. Early interpretations often attributed observed imbalances to sociocultural factors, such as reproductive risks and gender differences in labor intensity. During the 1990s and 2000s, research expanded to include regional syntheses, exploring the correlations between demographic patterns and social complexity, agricultural intensification, and environmental stress. Life table analysis gained prominence during this period, particularly for estimating life expectancy. Since 2010, methodological sophistication has accelerated: model life tables have been adjusted to correct for infant underrepresentation; Bayesian-inspired age estimation and transition analysis have been experimentally applied; and fertility proxies have been employed to test hypotheses about agricultural transitions. Increasingly, studies have adopted integrated approaches, combining demographic data with paleopathology and settlement patterns to investigate health disparities, conflict, and population dynamics.

Recent advancements in methods and technologies have further refined age estimation techniques. including Transition Analysis software (TA2, TA3) for Bayesian-based adult aging, reducing mimicry bias and enhancing accuracy. Complementing these developments, improvements in demographic modeling have seen parametric hazard models (e.g., Gompertz-Makeham, Siler) replace traditional life tables for mortality analysis, while fertility proxies were used to estimate population growth. Semi-parametric and non-parametric models have also been applied in paleodemography.

This review also examines four pivotal theoretical issues. First, the application of the uniformitarian hypothesis to paleodemography raises questions about its validity, particularly in light of the osteological paradox— the observation that skeletal samples may not accurately reflect living populations due to factors such as selective mortality and heterogeneity in physical health. Second, the conceptual misalignment between death assemblages and living populations poses a significant challenge. Third, the finding that mean age-at-death primarily reflects fertility levels rather than mortality has reshaped interpretations. Finally, the Neolithic Demographic Transition theory, an archaeological framework derived from paleodemographic research, explains how agricultural practices drove demographic changes, such as increased fertility.

The future of paleodemography in China holds substantial promise. To fully realize this potential, it is essential to further integrate interdisciplinary research, including aDNA and isotope studies, into migration research. Leveraging China’s extensive archaeological record will enable rigorous evaluations of global theories, such as the Osteological Paradox and the Neolithic Demographic Transition. Crucially, future research would actively develop indigenous theoretical frameworks rooted in China’s unique archaeological and historical contexts, thereby advancing the discipline’s localized development. Sustained interdisciplinary collaboration remains vital for methodological innovation and robust demographic reconstruction.