Archaeologists have long recognized that spatial relationships are an important influence on and driver of all manner of social processes at scales from the local to the continental. Recent research in the realm of complex networks focused on community detection in human and animal networks suggests that there may be certain critical scales at which spatial interactions can be partitioned, allowing researchers to draw potential boundaries for interaction that provide insights into a variety of social phenomena. Thus far, this research has been focused on short time scales and has not explored the legacies of historic relationships on the evolution of network communities and boundaries over the long-term. In this study, we examine networks based on material cultural similarity drawing on a large settlement and material culture database from the U.S. Southwest/Mexican Northwest (ca. 1000–1450 CE) divided into a series of short temporal intervals. With these temporally sequenced networks we: 1) demonstrate the utility of network community detection for partitioning interactions in geographic space, 2) identify key transitions in the geographic scales of network communities, and 3) illustrate the role of previous network configurations in the evolution of network communities and their spatial boundaries through time.

考古学家早就认识到，空间关系对从地方到大陆的各种社会进程具有重要影响和驱动力。最近在复杂网络领域的研究侧重于人类和动物网络中的社区检测，表明可能存在某些可以划分空间交互的关键尺度，使研究人员能够为交互绘制潜在边界，从而提供对各种社会现象的洞察力. 到目前为止，这项研究一直集中在短期尺度上，并没有探索历史关系对网络社区和边界演化的长期影响。在这项研究中，我们利用美国的一个大型定居点和物质文化数据库，检查基于物质文化相似性的网络 西南/墨西哥西北部（约公元 1000-1450 年）分为一系列短时间间隔。通过这些按时间排序的网络，我们：1) 证明了网络社区检测在地理空间中划分交互的实用性，2) 识别网络社区地理尺度中的关键转变，以及 3) 说明了先前网络配置在网络演化中的作用随着时间的推移，网络社区及其空间边界。