• 中国科学学与科技政策研究会
  • 中国科学院科技政策与管理科学研究所
  • 清华大学科学技术与社会研究中心
ISSN 1003-2053 CN 11-1805/G3

科学学研究 ›› 2022, Vol. 40 ›› Issue (3): 401-409.

• 科学学理论与方法 • 上一篇    下一篇

“算法”与国家治理创新——数据、算法与权力的知识生产与逻辑呈现

宋锴业   

  1. 中山大学政治与公共事务管理学院
  • 收稿日期:2021-01-31 修回日期:2021-03-15 出版日期:2022-03-15 发布日期:2022-03-15
  • 通讯作者: 宋锴业
  • 基金资助:
    国家自然科学基金面上项目“抱怨群体的用户画像及其在复杂性社会治理中的应用”;国家社会科学基金重点项目“基于数据赋能的基层社会治理现代化研究”

"Algorithm" and National Governance Innovation: Knowledge Production and Logic Presentation of Data, Algorithm and Power

  • Received:2021-01-31 Revised:2021-03-15 Online:2022-03-15 Published:2022-03-15

摘要: 数据、算法在当前国家治理实践中发挥着巨大作用,为了揭示算法的知识本质及其权力逻辑,根据科学技术研究(STS)和批判数据研究(CDS)对西方算法理论进行梳理的基础上,结合多个警务部门的大量深度访谈资料和数据,从中得出经验见解来打开算法的知识黑箱。研究表明,作为一种构建知识和权力的技术机制,算法一方面可以构建非结构性数据间的相关关系,以预测分布式群体的行为;另一方面,促进对复杂社会需求的高度响应和协调。除了塑造社会秩序和回应社会需求,文章进一步揭示了隐藏在算法光鲜技术科学背后的风险根源。通过对引发算法技术争议的焦点——算法透明性进行分析,在此基础上剖析算法的风险是如何在“技术”和“社会政治结构”两套科学/叙事知识中产生,并随着算法在各领域的广泛应用而引起“平等与正义”的科学争议。由此,可以为算法这一新的研究范式提供连贯的分析与统一理论,系统呈现算法给社会科学研究带来的挑战。

Abstract: data and algorithm play an important role in the current practice of national governance. In order to reveal the knowledge essence and power logic of algorithm, this paper combs the Western algorithm theory according to the scientific and technological research (STS) and critical data research (CDS), and combines with a large number of in-depth interviews and data from many police departments, so as to open up the application of algorithm the "black box" of knowledge. Research shows that, as a technical mechanism to build knowledge and power, on the one hand, the algorithm can build the correlation between nonstructural data to predict the behavior of distributed groups; on the other hand, it can promote a high degree of response and coordination to complex social needs. In addition to shaping social order and responding to social needs, this paper further reveals the hidden risk root behind the brilliant technology science of algorithm. Through the analysis of algorithm transparency, which is the focus of algorithm technology controversy, this paper analyzes how the risk of algorithm is generated in the two sets of scientific / narrative knowledge of "technology" and "social and political structure", and causes the scientific controversy of "equality and justice" with the wide application of algorithm in various fields. Thus, it can provide a coherent analysis and unified theory for the new research paradigm of algorithm, and systematically present the challenges of algorithm to social science research.