Taming the complexity of the law: modelling and visualisation of dynamically interacting legal systems [RENEWAL].

在我的奖学金项目第一阶段取得成功之后，该提案的目的是从两个角度进一步理解和遏制法律体系的复杂性：一方面，我们将开发并增加我们的“Graphie”可视化界面的重要功能，使用户能够超越传统的纯“文本”范式浏览英国法律语料库。另一方面，我们将深化我们对法律文本“复杂”原因的理论研究——从普通读者定位隐藏在层次结构（建模为复杂网络）中的信息有多困难的角度来看。该项目的各个阶段包括数据收集、可视化/用户体验设计、理论建模和开发新的定量工具，以评估动态交互和不断发展的规定网络单元（例如，现行法案的条款或章节）的复杂性。“法律复杂性”问题长期存在，法律学者，从业者，甚至未经培训的用户都很难感受到，我的项目有雄心壮志，通过一种创新的，真正跨学科的方法来解决这个问题，这种方法利用了复杂性科学的工具以及最先进的技术和数据可视化框架。最终目标是证明当前立法的产生、显示和提供给最终用户的方式是非常不理想的，并且可能有替代方案。这种文化转变将取决于网络理论和复杂性科学的最新发展，它们首次标志着法律复杂性从“定性”方法到“定量”方法的转变。法律复杂性的主要来源之一是在任何特定时间有效的法律规定的绝对数量。粗略估计，目前在英国立法体系中生效的法律大约有5000万字（相比之下，人类基因组中只有大约2.5万个基因），每月有10万个正在产生或修改。另一个问题涉及议会个别法案的编写方式：原始结构以及随后的修正案使得普通读者很难理解给定条款的所有含义，因为通常需要跨网络不同“节点”的多个跳点来形成对法案内容的详尽视图。法律条文不断被创造、修改和废除，而且通常指向其他条文，无论是旧的还是位于其他地方的，这一事实使法律语料库成为一个动态的、高度内部联系的实体——令人惊讶的是，人们仍然通过几个世纪以来使用的相同工具来访问法律语料库：手动单词搜索和引用编号段落。2013年，国会法律顾问办公室发起了“好法律”倡议，“共同目标是使立法为今天和明天的用户提供良好的服务”。然而，那里提出的抑制语料库作为一个整体的复杂性的选择-例如在标准的繁文缛节或二出一进计划中-通常基于naïve“删除项目”策略，这根本不能保证实现控制法律复杂性的目标。最后，导致法律复杂性问题的另一个问题是，律师和法律从业人员通常没有接受复杂性科学和定量方法方面的必要培训，无法对其专业人员以更有效和理性的方式进行日常业务的方式产生有意义的改变。我的愿景是，先进的网络理论、数字工具和复杂性科学的分析技术，以及将法律和基础科学融合在一起的跨学科培训计划的创建，应该以前所未有的协同作用相结合，在法律的起草、可获取性和用户友好性方面制定新的可衡量标准。

Following a successful first phase of my fellowship, this proposal has the aim of further understanding and curbing the complexity of legal systems, from a dual standpoint: on one hand, we will develop and add important features to our 'Graphie' visualisation interface, which allows users to navigate the UK legal corpus beyond the traditional purely 'textual' paradigm. On the other hand, we will deepen our theoretical investigation on what makes a legal text 'complex' -- from the point of view of how difficult it is for a generic reader to locate a piece of information hidden in a hierarchical structure (modelled as a complex network). The phases of the project entail data collection, visualisation/user experience design, theoretical modelling, and development of novel quantitative tools to assess the complexity of dynamically interacting and ever-evolving units of a network of provisions (for instance, the articles or chapters of a living Act). The 'legal complexity' problem is long-standing and hard-felt by legal scholars, practitioners, and even untrained users, and my project has the ambition to address it via an innovative and truly interdisciplinary approach, which leverages tools from complexity science as well as state-of-the-art techniques and frameworks for data visualisation. The final goal is to demonstrate that the current way legislation is produced, displayed, and offered to end-users is highly suboptimal, and that an alternative is possible. This cultural switch will rest on recent developments in network theory and complexity science, which for the first time mark a transition between 'qualitative' and 'quantitative' approaches to legal complexity.One of the primary sources of legal complexity is the sheer volume of legal provisions in force at any given time. A rough estimate points to the existence of about 50,000,000 words of law currently in force in the UK legislative system (in comparison, there are only about 25,000 genes in the human genome) and 100,000 being produced or amended each month. Another issue concerns the way individual acts of parliament are written: the original structure as well as the subsequent amendments make it very difficult for a general reader to follow all the implications of a given provision, because multiple hops across different 'nodes' of the network are typically needed to form an exhaustive view of an act's content. The fact that legal provisions are continuously created, amended and repealed, and typically point towards other ones, either older or located elsewhere, makes the legal corpus a dynamical and highly intra-connected entity -- which is surprisingly still accessed via the same tools used for centuries: manual word searches and references to numbered paragraphs. In 2013, the Office of the Parliamentary Counsel launched the ''Good law" initiative "with a shared objective of making legislation work well for the users of today and tomorrow". However, the options proposed there to curb the complexity of the corpus as a whole -- for instance in the standard Red Tape or 2-out-1-in initiatives -- are generally based on a naïve 'item-removing' strategy, which is not guaranteed at all to achieve the goal of keeping the intricacy of the law under control. Finally, a further issue contributing to the problem of legal complexity is the fact that lawyers and legal practitioners do not usually have the necessary training in complexity science and quantitative methods to be able to effect a meaningful change on the way their own professionals go about their daily business in a more efficient and rational way. My vision is that advanced network theory, digital tools and analytical techniques from Complexity Science -- as well as the creation of an interdisciplinary training programme blending law and basic science together -- should be combined in an unprecedented synergy to set new measurable standards in the drafting, accessibility, and user-friendliness of the law.