“算法冒犯我”:用户与算法的情感实践及其自主性

"Algorithms Offend Me":Emotional User-algorithm Practices and User Agency

将算法视为可交互的传播主体,有助于挖掘更深层次的算法与人的复杂关系。通过对18名算法推荐类应用程序用户的深度访谈,本文发现算法与用户之间存在一种特殊的情感实践,本文将其称为“算法冒犯”。算法冒犯呈现出多维度的内涵:既是用户直接的负面情绪表达,也构成了算法在互动关系中对用户主权的“越界”和“僭越”。同时,这也是“算法批判”的集体叙事在个人层面的彰显,具体体现为画像失真、语境失当和隐私侵犯。而用户自主性则是用户应对算法冒犯的主要机制,表现在温和抵抗、自我调适、算法素养三个方面。通过凸显情感实践维度的价值,本文期望推动算法研究进一步回归日常生活与主体经验。

Regarding algorithms as communicative subjects is helpful to mine the deeper entanglement between the algorithms and people. Through in-depth interviews with 18 users of algorithm recommendation applications, this paper finds that there is a kind of special emotional practices between the algorithms and users, which is named as "algorithmic offence". Algorithmic offence presents multi-dimensional meanings: it is not only the direct negative emotional expression, but also constitutes the "deviation" and "arrogation" to the user’s agency. At the same time, it is also the manifestation of the collective narrative of "algorithmic criticism" at the individual level, which is embodied in profile distortion, context inappropriateness and privacy infringement. User agency is the main mechanism to deal with algorithmic offence, which is manifested in three aspects: mild resistance, self-adaptation and algorithmic literacy. By highlighting the value of emotional practice dimension, this paper hopes to promote algorithm studies to further return to daily life and subject experience.