法律算法化的可能與限度

Possibility and Limitation of Legal Algorithmization

法律與算法在技術性表徵上具有類似性，即產生機制的類似性、功能效用的類似性、表達方式的類似性，這些類似性為法律借鑒算法創造了可能性。法律與算法在價值性內核上具有異質性，即運行邏輯的異質性、有效性來源的異質性、認知方式的異質性，這些異質性提示了法律與算法之間的距離。關于類似性與異質性的分析為法律的算法化限度提供了一個總原則，即法律在技術性的層面可以借鑒算法的模式與成果，但在價值性內核上應與算法保持距離。這個總原則可以分解為三個限度，即科學主義限度、建構主義限度、工具主義限度。科學主義限度要求法律在借鑒科學技術的同時，警惕科學求真邏輯對法律正義命題的侵蝕；建構主義限度強調法律借鑒算法建構思維的同時，防止純粹建構理性對法治的誤導；工具主義限度則提示將法律視為達致特定目標工具的同時，要時刻關注法律的價值屬性，尤其重視法律人及其反思在控制算法思維泛濫中的重要作用。

Laws and algorithms are similar in technical representation, that is, the similarity of generation mechanism, functional utility and expression. These similarities create the possibility for laws to learn from algorithms. Law and algorithm have heterogeneity in the value core, that is, the heterogeneity of operating logic, the heterogeneity of effectiveness sources and the heterogeneity of cognitive style. These heterogeneity suggest the distance between law and algorithm. The analysis of similarity and heterogeneity provides a general principle for the algorithmic limit of law, that is, the law can learn from the model and achievements of the algorithm at the technical level, but it should keep a distance from the algorithm in the value core. This general principle can be divided into three limits: scientism, constructivism and instrumentalism. The limit of scientism requires the law to learn from science and technology and guard against the erosion of scientific truth - seeking logic on the proposition of legal justice; The limit of constructivism emphasizes that while the law draws lessons from the algorithm to construct thinking, it prevents the misleading of pure constructive rationality to the rule of law; The limit of instrumentalism suggests that while considering law as a tool to achieve specific goals, we should always pay attention to the value attribute of law, especially pay attention to the important role of legal people and their reflection in controlling the proliferation of algorithmic thinking.