[author]Zuo Weimin
[content]
The lineage orientation of legal research in the era of big data: self disciplinary law?
*Author Zuo Weimin
Professor at the Law School of Sichuan University and member of the Academic Advisory Committee of the China Institute of Socio-Legal Studies at Shanghai Jiao Tong University
Abstract: In the context of the big data era, based on the development of empirical research, it is necessary to examine the development of the legal system, especially whether empirical research should move towards "self disciplinary law" with the help of big data. With the help of network technology, the "self scientificization" of legal research cannot be ignored. The application of knowledge and technology from natural sciences to the analysis of legal phenomena can to some extent change the problem of previous legal research being too abstract and subjective. However, while maintaining enthusiasm for natural science and technology, researchers should also be careful to avoid falling into the trap of technologism, as ideas and humanistic care for the legal world have always been the greatest pursuit of legal research.
Keywords: big data; Legal empirical research; Social Science Law; Self disciplinary law; crossover
Introduction: Jurisprudence, Social Science Law, and Empirical Legal Research
There have been discussions on research methods and schools of law for a long time. Since the beginning of this century, when Su Li categorized the legal research paradigms since the founding of the People's Republic of China into three major categories: political law, interpretive law, and social science law, most of the debates in the legal community regarding the approaches and methods adopted by researchers have revolved around this. With the continuous maturity of academic research and the gradual changes in demand in the academic market, political and legal studies have basically exited the historical stage shortly after the reform and opening up, so current related discussions mainly focus on the latter two. Interpretive jurisprudence (often referred to as legal doctrine), according to Su Li's summary, is characterized by a high focus on specific legal systems and technical issues, and an emphasis on studying legal issues in real legal life. Due to its alignment with specific historical needs and advantages such as lower costs of knowledge production and transmission, the study of legal doctrine quickly dominated the majority of legal research in China after the 1980s, and its influence continues to this day. However, influenced by deductive logic, the prevalence of legalism has also led to the idealization and "meta theorization" of China's legal knowledge system to some extent.
The emergence of social science law is somewhat related to the above-mentioned characteristics of legal doctrine. Due to their dissatisfaction with deduction or inference and their desire to explore specific judicial processes, some scholars have begun to introduce knowledge and methods from other social sciences into legal research. Objectively speaking, social science law has left a significant mark in the development of China's legal research paradigm. Influenced by it, experience rather than norms, field investigations rather than logical deduction have gradually gained more and more attention, deepening the understanding of Chinese legal practice by scholars, decision-makers, and even the general public. The view that 'law is a local knowledge' has also returned to the public eye in a completely new way.
With the continuous expansion of the influence of social science law, further exploration has begun to revolve around the diversification of social science law. One of the debates is about the positioning of legal empirical research related to social science law. Essentially, legal empirical research is an empirical research paradigm centered on data analysis, which involves collecting, organizing, analyzing, and applying data to grasp and interpret real legal practices, in order to provide policy recommendations or basis for legal modifications. In the view of social science law, the rise of empirical research methods is simply the result of its own division of labor and continuous refinement of research methods. Some scholars directly propose, "Isn't empirical legal research just... social science law?" Is this really the case? This is the first question that this article aims to address. Furthermore, in the current era of deepening judicial informatization and openness, judicial big data has been utilized by empirical researchers. What are the similarities and differences between this big data legal research and traditional small data research? This is the second question to be answered in this article. Subsequently, with the continuous expansion and deepening of the application scope and degree of natural science ideas and methods such as statistical models and machine learning in legal empirical research, can we consider a new research paradigm called "self scientific jurisprudence" to be forming? If this is indeed the case, how should the future development of "self science and technology" be carried out? This is the third and most important question to be answered in this article.
1. Revisiting an old question: The relationship between empirical research in law and social science law
The relationship between legal empirical research and social science law affects whether the former has an independent academic status. In the early stages of the emergence of empirical research in contemporary Chinese law, it was not mature enough, mostly limited to simple enumeration and analysis of numbers and samples, and a professional academic community had not yet formed, far inferior to the flourishing of social science law. At that time, the academic ecology was indeed prone to the judgment that empirical law was only a subsidiary of social science law. In fact, legal empirical research often recognized itself as belonging to the broad paradigm of social science law. But as empirical research has increasingly demonstrated its unique academic pursuit and purpose, thinking mode and research methods, especially in the current deepening integration of natural sciences and empirical research, legal empirical research has become quite distinct from social science law and has become an independent research paradigm.
1.1 Starting with the term 'Empirical'
According to the Black's Law Dictionary, the term 'Empirical' means' of, relating to, or based on experience, experiment, or observation '. In the broadest sense, any legal research based on practical experience can be referred to as' Empirical Legal Research '. Scholars have used it to broadly refer to quantitative studies of partial weight analysis and qualitative studies of qualitative analysis. However, the connotation of the term "legal empirical research" is not without controversy, and some scholars specifically refer to quantitative research centered on data analysis. In my opinion, what is truly worth paying attention to is how to grasp this concept in operation. What needs to be explained first is what is meant by "experience"? In the context of legal research, it simply means that the starting point and destination of the research should be determined as the basic operational state of the law on an objective level, with the ultimate pursuit of solving practical problems. Experience and norms should be considered as fundamental concepts in legal research. Since 'experience' is the core of understanding 'Empirical Legal Research', another key question that follows closely is how to acquire this experience? Social science law is experiential, but this does not mean that the study of legal experience is equivalent to social science law. In the 1990s, under the promotion of scholars mainly represented by Su Li, the knowledge of social sciences, including sociology, began to be accepted by the legal community, and scholars showed a strong interest in the real operation of law. However, social science law has not solely promoted the formation and development of legal experience research in China, because shortly after the rise of social science law, scholars have been practicing another form of empirical research - legal empirical research that focuses on quantitative analysis. For example, Professor Bai Jianjun in the field of criminal law has been engaged in empirical research since the late 1990s, which can be said to have pioneered legal empirical research in the field of criminal law.
It can be seen that the development of social science law and empirical research in China has different paths. The commonality between the two is that they both promote the transformation of Chinese legal research from "metaphysical" to "metaphysical". It is necessary to clarify that although the field of "social sciences" in social science law is relatively broad, the most influential approach in legal research may still be the qualitative methods used in sociology, despite the increasing emphasis on quantitative research in Western sociology. That is why social science law experts conduct a detailed analysis of individual cases in a 'sparrow dissection' style. In contrast, legal empirical research pursues more representativeness of data, because in this research paradigm, only through the collection and application of large-scale data can the actual state of the operation of the legal system be more fully reflected. For the study of individual cases, even if it is extremely subtle, it is difficult to comprehensively and accurately grasp the true context of legal activities.
A noteworthy new trend is that with the continuous development of emerging technologies such as big data and artificial intelligence, the Chinese legal community, which has always been good at drawing on new things and concepts, has quickly introduced relevant theories and methods into legal research, and its influence continues to expand. A typical example is that "computational algorithms" have become a hot topic in the current academic community. Regarding the formation and development of computational law, some scholars have pointed out that it can be classified as empirical law and is a product of the entry of legal empirical research into the era of big data. Logically speaking, computational law should be a research method based on mathematical and quantitative analysis, which is different from social science law research that focuses on case analysis.
Based on the above analysis, the author believes that it is necessary to clarify the logical relationship of the existing conceptual system again. Specifically, Empirical Legal Research in law should be listed as one of the two fundamental paradigms of legal research, alongside the study of legal values/norms that focus on rational reasoning. Social science law and empirical legal research both belong to the field of legal experience research, with the former emphasizing qualitative research and the latter emphasizing quantitative analysis. The two cannot generate a simple relationship of inclusion and being included, as shown in the following figure.
1.2 Is it 'real' without 'proof' or is it a combination of 'real' and 'proof'?
It should be acknowledged that there are significant differences between social science law and legal empirical research. Clarifying this difference is beneficial for us to have a more comprehensive understanding of the overall framework of current legal research. Simply put, the above differences are mainly reflected in the handling of the relationship between "reality" and "evidence" in the research characteristics. In the view of social science law, even though it can include empirical research that emphasizes qualitative research and empirical research that emphasizes quantitative research, it is also acknowledged that qualitative research is the main focus, while quantitative research has always been on the periphery. For example, Hou Meng believes that the domestic social science and law community, whether it is legal economics or legal sociology, excels in individual cases. Of course, this is not sufficient to prove the rationality of the division between social science law and empirical research. Because the key lies in how social science law scholars apply case studies?
Similar to other legal research paradigms, social science law also carries the historical mission of enhancing China's level of rule of law. According to the research habit of social science law, this historical mission is mainly accomplished through in-depth analysis of certain cases with significant historical or practical significance that arise in the judicial process. How can social science and law serve this purpose? When Su Li elaborated on the core issues of social science law, he pointed out that the focus of social science law is not on a specific legal concept or provision, but on the mutual influence and constraints between laws or specific rules and various factors in social life, emphasizing the discovery of the "behind" or "inside" truths of individual cases. It can be seen that even within the field of social science and law, there are different voices regarding the function of the discipline.
In terms of the ability of social science law to reveal profound principles or universal principles behind individual cases, this research method has made significant contributions to promoting the process of rule of law in China. However, the previous viewpoint has already hinted at the fact that social science and law mainly provide "reasoning". But it seems difficult to fully explain what reason is, because compared to legal principles, it may be more abstract. Where does the principle of social science and law come from? As mentioned earlier, social science and law attach great importance to the importance of individual cases in research. Without typical cases, research is almost impossible to carry out. However, individual cases are only individual cases, and all they can provide is the case itself. The analysis based on this depends on the researcher's own theoretical expansion and extension, which may be the manifestation of the principles emphasized by social science law. In my opinion, the so-called expansion and extension are actually the manifestation of the researcher's imagination. But it is obvious that both academic intuition and imagination are essentially abstract. This also determines that the role of principles in social science and law is often to bring readers a sense of intellectual "enlightenment" pleasure.
It should be said that the most prominent feature of social science law is the emphasis on "reality" rather than "proof". Social science law is both "empirical" and "practical", but due to the abstract principles based on imagination and intuition, it is sometimes not "proven" or "discussed but not proven". There are two main reasons for this lack of evidence: firstly, researchers are unwilling to provide evidence. Chen Xingliang pointed out that compared with legal doctrine, social science law studies law more in a holistic sense, or in other words, it originally studied holistic law. I believe that even compared to legal empirical research, this viewpoint still holds true. For social science law, although the entry point is usually subtle, researchers' problem awareness is often not limited to individual or several legal provisions, but rather the entire legal system. This makes social science law often lack a relatively concentrated research focus while possessing a grand perspective. A typical manifestation is that in the study of social science and law, "big words" such as power, politics, and institutions appear frequently, but researchers are often vague about what kind of power, what kind of politics, what kind of system, and so on. Therefore, the unwillingness to "prove" here refers to the fact that researchers often do not empirically verify research from a micro perspective in order to maintain the uniqueness of social science and law. The second is that the research itself is difficult to 'verify'. The principles of social science and law are not only abstract, but also highly personalized. When the research conclusion relies almost entirely on the researcher's personal theoretical extension and expansion, the correctness of the conclusion is guaranteed at least within the context set by the researcher. However, even for the same case, different researchers may have different interpretations due to differences in knowledge background, research interests, etc., which may lead to different or even completely opposite conclusions. But in any case, as long as it does not exceed the logic arranged by the researcher, this conclusion is always correct, and research cannot be mutually falsified. It can even be argued that some studies in social science and law only distinguish between convincing and unconvincing.
Unlike social science law, legal empirical research emphasizes the combination of "reality" and "evidence". At the level of "reality", there is not much difference between the two, both of which are empirical. The key difference lies in the attitude and specific development towards "evidence", mainly manifested in the importance of data in their respective studies, and whether it is possible to verify data and argue or refute an academic hypothesis based on the data. In my opinion, data in social science and law is often just a basis for supporting theoretical claims. Su Li believes that statistical analysis cannot replace textual reasoning. Mathematics is often just a decoration, and it is more important to select examples and explain the reasoning for readers. In contrast, although the imagination of researchers is also very important in legal empirical research - in fact, imagination is essential for any research - it has not yet reached the point where it can determine the level of research excellence. The reason is simple, the reasoning given by legal empirical research is based on the deep utilization of data and samples (mainly data usage techniques).
In terms of specific manifestations, empirical research in law is relative to social science law. On the one hand, legal empirical researchers are willing to 'prove'. In legal empirical research, researchers tend to strengthen the representativeness of research conclusions through rich data. As for the research object, the starting point and foothold of legal empirical research are single or several specific legal norms, and after examining the actual operation of these norms - for those with poor effects - certain improvement suggestions are given. As for how to explain the operational effects of legal norms at the micro level, researchers tend to rely more on data interpretation instead of theoretical deduction, and this explanation process is the expression of "proof" in legal empirical research. On the other hand, legal empirical research can also be 'proven'. Unlike the highly personalized construction of social science legal theory, although there are various ways to use samples in legal empirical research, the basic data is given, the analysis mechanism is common, and the conclusions drawn will be almost the same. In short, legal empirical research can not only have a distinction between convincing and unconvincing, but the research methods themselves also have a distinction between right and wrong. Besides professional researchers, legal empirical research can also be falsified or proven for general readers and judicial practitioners.
It needs to be reiterated that there is no superiority or inferiority between the two research paradigms. In fact, the two can and often do form a cooperative relationship. Specifically, social science law can extract "legal principles" from trivial "facts" based on its grand theoretical background and meticulous empirical analysis, and point out the significant characteristics and shortcomings of a certain aspect of China's legal system as a whole; For empirical research in law, the analysis of social science law can serve as a source of problem awareness. Its task is to "prove" the analysis of social science law and intervene in the analysis of specific institutional norms. This is equivalent to presenting the conclusions of social science and law in the form of data, and it is also a test of the correctness of the conclusions of social science and law.
Huang Zongzhi believes that in terms of methodology, research in the Chinese field has actually been dominated by the opposition or complete separation between theory and experience. Although social science law and legal empirical research are essentially experiential, the former's experience sometimes seems to encounter some obstacles in the process of transitioning to pragmatism. The important reason is that research in social science and law is difficult to falsify. In contrast, this phenomenon is not easily present in legal empirical research, as its mathematical and chemical characteristics can be evaluated and can also be discovered and corrected when errors occur. Therefore, in China, which is continuously transforming into a modern rule of law country, although we need intellectual satisfaction and pleasure, we also need strong experiential support with practical significance for reform decisions. In other words, we certainly need to be 'exciting', but we also need to be 'useful'.
2. Research on Legal Big Data from the Perspective of Methodology
Legal big data research is a new form of empirical research that has developed in recent years. It originated from the centralized and unified online access of judicial documents, which has given rise to some new legal practices. Prior to this, the data that could be presented on a certain scale in legal research were mainly work reports or various legal yearbooks, statistical yearbooks, etc. issued by central and local judicial organs and statistical departments. Compared to such data, the biggest breakthrough in the online publication of judicial documents is reflected in the official release of structured and digitized judicial documents on such a large scale for the first time. In this case, the number of documents publicly available on the Judicial Document Network and the number of documents obtained by researchers through various means largely determine the scale of a study, which directly raises the following question: Does the emergence of judicial big data mean that there is an objective evaluation standard for the value of empirical research, that is, whether the larger the number of documents obtained, the stronger the representativeness and the higher the research quality?
In theory, if we can obtain big data from the entire sample, the quality of the research itself and the reliability and validity of the research conclusions can be guaranteed, because "when the sample is equal to the population, the sampling error is zero." However, this scenario is only "desirable." In reality, the data on the Judgment Document Network is often incomplete and constantly changing, and may even be limited due to changes in public strategies. Regarding this, some scholars have bluntly pointed out that the current research on legal big data is just a fictional "big cake", and for scholars who are interested in engaging in related research, it can only be a "pie in the sky" at this stage. The practical situation is indeed like this: although the online publication of judicial documents has brought unprecedented opportunities for legal big data research, the research that can truly fully and effectively utilize this resource is still limited.
Since there are challenges in obtaining and using judicial documents, does this mean that true legal big data research does not exist? I think it depends on how we understand the concept of big data. Bai Jianjun believes that big data does not lie in the absolute size of the sample, but rather in its comprehensiveness. The term 'all' here refers to the fact that the amount of data used by the research institute is' all 'in terms of problem awareness. This reminds us that on the one hand, big data is not necessarily 'big'. Taking my research on the selection mechanism of local judges as an example, although the sample size is less than 6000 judges, it is still sufficient to provide us with an observation window to discover some deep-rooted habits and power logic in the selection process, as well as the shortcomings of the judge selection mechanism. Obviously, from the perspective of data volume, a sample of less than 6000 cases is difficult to call big data, but in terms of the research question, it is a "complete" data. On the other hand, large-scale data is not necessarily good data. Cheng Jinhua pointed out that although the Judgment Document Network can provide us with massive information, it may not necessarily be able to obtain scientific information. In addition to the lack of judicial documents, the insufficient structure of the documents themselves also affects the reliability of the sample data to a certain extent. From this, it can be seen that there is no necessary or direct connection between the large amount of data and the true meaning of big data research. The research using big data methods may not necessarily guarantee good quality, and the key issue lies in whether the selected data can meet the research needs.
However, while advocating for a "cooling down" of researchers' understanding of big data research methods, it must be acknowledged that the current sample size of research has already exceeded any previous period and is irreversibly developing towards a "massive scale". This naturally raises another question: the larger the amount of data, the less likely it is to rely solely on manual analysis. Does this mean that traditional empirical research will no longer be valued or even have no room for survival?
The traditional empirical research method refers to the researcher's manual calculation based organization, classification, and analysis of limited data. To some extent, it can also be simply assumed that traditional empirical research is often just small data research. It has not lost its value. Firstly, the choice of research method should be determined by the research task. One of the fundamental purposes of current legal research in China is still to objectively understand legal practice. Although the significant increase in data volume may indeed deepen or expand the process of understanding and practice, as mentioned earlier, big data legal research also faces a series of problems including data distortion, which will to some extent hinder the deepening of this process. We should pay attention to whether the amount of data meets the problem awareness of the research, as small data may also generate research value for big data or "whole" data. Secondly, in terms of the proactivity of data utilization, traditional quantitative research is far stronger than big data research. On the surface, big data technology seems to have unparalleled data utilization capabilities, but on the other hand, big data methods are passive because they can only wait for a specific amount of data to be generated before initiating subsequent processing; And researchers themselves cannot intervene in the process of data generation - even the predictive functions that big data takes pride in need to be built on this foundation.
In contrast, traditional empirical research often has a higher degree of data utilization initiative and can "customize" data according to the needs of researchers. Of course, "customization" here does not mean collecting data for "validation" based on established conclusions, but rather refers to researchers actively obtaining high-quality information data with a purpose and direction for specific problems. Moreover, the important difference between big data legal research and traditional empirical research lies in their different answers to the following key questions, namely whether causality or correlation exists. In short, big data legal research focuses on analyzing whether there is a statistically significant correlation between variables and between variables and conclusions (although some researchers also advocate causal analysis based on big data, the author has reservations about this). Traditional empirical research on small data, on the other hand, places greater emphasis on exploring various social and legal phenomena and their underlying causes from a genetic perspective.
It should be pointed out that mining and displaying the correlation between things seems to be an inherent advantage of big data technology. Regarding this, some scholars abroad believe that with the help of relevant relationships, by exploring "what" rather than "why", we can better understand the world, and all of this is based on the premise of humanity entering the era of big data. But from another perspective, when the sample size is geometrically enlarged, finding the correlation between specific objects is not originally a very difficult task. In certain circumstances, the correlation may even be "created" due to the enlargement of the sample size. Scholars have pointed out that in databases, due to the large amount of data, it is usually easy to obtain statistically significant regression coefficients, but this does not mean that there is a causal relationship between two variables. For example, big data analysis is likely to tell us that from 2006 to 2011, the murder rate in the United States was closely related to the market share of Microsoft's IE browser, as both showed a sharp downward trend, but there was actually no causal relationship between the two.
Wang Tiansi believes that while analyzing the relevant relationships before factors interact with each other is important, deepening causal analysis of the interaction process based on factor analysis still holds a more important position in the deepening of human understanding. In the field of law, this conclusion also applies, because at least for current Chinese legal research, exploring the causes of various legal issues and providing corresponding solutions clearly has an unquestionable position - we need to ask why. Due to the relative passivity in data utilization, big data legal research originated from big data and is also limited by it. Especially when big data is assumed to have full sample characteristics to avoid confidence issues caused by insufficient samples, researchers may give up further theoretical analysis efforts.
In small data research, the pursuit of causal relationships is clearly more likely to receive attention. It can be said that summarizing and analyzing the interaction process between causes and outcomes is an important purpose and charm of traditional empirical research. The reasons why traditional empirical research seeks causal relationship analysis are multifaceted. On the one hand, traditional empirical research is a research paradigm about explanation. "When we want to explain an event, we naturally expect to find the cause of the event. Explanation can only be expressed in a causal manner, which makes it difficult to explain using only correlation." It follows a "problem to data" approach, and the conclusions drawn are often direct explanations of the problem. On the other hand, traditional empirical research also requires precision in data utilization to analyze causal relationships. The vast amount of data in the era of big data to some extent means the mixing of data, which means that big data research methods can only search for clues of correlation in these complex data. In contrast, the samples selected by traditional empirical research are refined and small-scale data that have been carefully screened under the guidance of specific and subtle problem awareness, and are more likely to explore causal relationships.
Although there are differences between causal relationships and correlation relationships, the purpose of the author is not to provide a subjective judgment on which is better or worse. On the contrary, the author always believes that in the process of approaching Chinese legal practice and solving Chinese legal problems, whether it is causal or related relationships, we cannot neglect them. As some scholars have pointed out, although causality is the fundamental basis for human rational behavior and activities, we cannot deny it ourselves. However, the correlation that big data technology can more easily highlight has indeed promoted our deep reflection on traditional causal concepts at the practical level. But with the increasing popularity of the concept of big data, correlation analysis based on big data has the potential to obscure causal relationships. A typical manifestation is that some big data studies, after assuming or discovering correlations between variables, directly draw positive or negative conclusions about a certain issue, while intentionally or unintentionally ignoring the analysis of social background, that is, believing that "as long as there is enough data, numbers themselves can explain everything." However, the correlations between variables obtained through thoughtless big data analysis may be false correlations, and sometimes they may just be coincidences. In essence, no matter how large the data is, as long as there is no rational design, control, and analysis, the conclusion may be wrong.
3. Flash in the pan or future mainstream: prospects for the development of "self disciplinary law"
Giddens once said, "Modernity has thrown us off the track of all types of social order in an unprecedented way, thus forming its way of life. In terms of extension and connotation, the changes that modernity has been involved in are more profound and far-reaching than most of the changing characteristics of past eras." The impact of this modernity is particularly evident in the field of legal research: in the first decade of this century, big data was almost completely unfamiliar to the legal community, but now the integration of big data technology and legal research seems to have become an irreversible new trend, and has produced some research results. At the same time, knowledge and methods from natural sciences or computer science fields such as machine learning, semantic recognition, and social science experiments have flooded into legal research and gradually formed corresponding academic groups.
At present, the integration of law and other disciplines seems to be expanding from social sciences to natural sciences. The product of this integration seems to correspond with social science law and can be called "self science law", which is a legal research paradigm that uses natural science thinking methods and techniques, especially statistics, data science, etc., to study legal issues and phenomena. Undoubtedly, "self disciplinary law" is a new trend that has emerged in the field of legal research and is far from fully developed. The research in this article is also exploratory and framework oriented. There are two questions worth considering: firstly, will "self disciplinary law" be a flash in the pan, or may it become an important direction for future legal research? Secondly, if the research approach of "self disciplinary law" wants to enter the mainstream, what is the direction of efforts? In response to this, the following text focuses on two aspects of analysis.
3.1 Definition of "Self Science and Law"
Scholars have explored issues related to "self science and technology". As some scholars have pointed out when discussing the current development status of artificial intelligence law, the inconsistency of field names is an urgent problem to be solved, because there is currently no particularly inclusive concept that can be unanimously recognized by research subjects - at least there are diversified terms such as big data and artificial intelligence law, computational law, etc. In my opinion, adopting any of the above concepts cannot comprehensively cover the latest development trends in related fields. Artificial intelligence law is not only the application of artificial intelligence technology in legal research, but also involves professional skills and knowledge in multiple fields. Similarly, concepts such as computational law are also worth considering in terms of their comprehensiveness. Given the overall trend of continuous integration and linkage between legal research as a typical social science and natural science that we are facing, we should view this change in the field of legal research from a global perspective. In this regard, the "self disciplinary law" proposed by the author is not just a research method, but also encompasses the trend of interdisciplinary integration mentioned above.
It should be pointed out that if the emergence of social science law is to make legal research "experiential", then "self science law" is based on this to make legal research "scientific". How to make it scientific? Scholars have pointed out that quantitative law, computational law, and other fields are all manifestations of the close relationship between law and technology. They attempt to reduce subjectivity and enhance scientificity in the legal system through objective standards, which have a positive impact on the scientific development of the legal system. It is worth noting that this scientific exploration is relatively abundant in the judicial field. For example, Li Xueyao and Liu Zhuang used experimental methods to discover that the reasoning behind legal judgments by legal professionals can be referred to as a "pretentious technique". For example, Tang Yingmao and others used experimental methods to analyze the impact of live court broadcasts on trial activities. It should be acknowledged that the above attempts have played a certain role in promoting the scientific development of legal research.
Some commentators point out that in the era of big data, what big data needs to enter is a legal system that can express itself as mathematically as possible, rather than relying mainly on traditional analytical methods. The reason why natural sciences have advantages over social sciences in terms of objectivity and accuracy is that, if properly grasped, some problems, concepts, and principles can be simplified into numbers or formulas, and then different variables can be embedded and scientifically analyzed to obtain corresponding conclusions. On the contrary, traditional legal research methods mainly rely on textual descriptions, which are often highly personalized based on subjective understanding. The same problem may lead to completely opposite conclusions from different scholars. This makes it difficult for descriptions of legal issues to flow smoothly among researchers, and even difficult to reach consensus on certain fundamental issues. Of course, this is also related to the differences in research nature between natural sciences and social sciences: the former can reproduce specific experimental environments multiple times when conditions permit to verify a certain conclusion; For the latter, it is difficult to reproduce specific environments.
In the author's opinion, the so-called "self disciplinary law" first includes a change in thinking concepts. The concept of this new paradigm of legal research is not simply about using statistical tools to solve problems in traditional legal research. It strives to create legal knowledge with scientific thinking attributes and establish a legal research system under the binary structure of logic and experimentation. It should be pointed out that "self disciplinary law" does not mean to completely digitize and formalize legal research - this is neither possible nor necessary, but rather to transform the appropriate part of legal issues - legal experience phenomena - into recognizable and even objectively calculable symbols, which can be substituted into the established models for computer or manual calculations, in order to eliminate the influence of subjective cognition of researchers in this process as much as possible. Ji Weidong believes that in the current era of geometrically increasing data, using mathematical models, statistical knowledge, and algorithms to find the optimal or suboptimal solution in complexity, in order to provide objective scientific basis for decision-making, is the essence of political arithmetic. Ji Weidong's viewpoint to some extent involves the core essence of "self disciplinary law", which is to transform the logical deduction and conceptual deduction that operate in the subjective world of researchers into a visual process similar to solving arithmetic problems. Domestic scholars have summarized this process as "social computing" after physical computing and biological computing, and believe that it has the potential to become a new focus of scientific computing research and development.
Secondly, "self disciplinary law" emphasizes the falsification thinking in legal research, which is also one of the key functions of "self disciplinary law". The problem consciousness in current legal research mainly comes from two aspects: one is from the study, where a phenomenon in legal practice should be judged as a "problem" in legal research based on logical deduction or foreign experience; The second is to move towards practice and gain inspiration through methods such as data collection and field investigations. On the surface, problems discovered through firsthand and experiential approaches usually have a high level of reliability. However, as this experience is formed by the researcher's personal investigation, it is inevitably limited by their abilities and conditions, and often relies on the researcher's subjective judgment, which inevitably leads to the existence of "pseudo problems". In this regard, the value of "self disciplinary law" lies in the elimination of falsehood and preservation of truth in theoretical problems, because with the help of methods from disciplines such as mathematics, statistics, and computer science, "self disciplinary law" can eliminate the subjective influence of individual researchers as much as possible; At the same time, appropriate data can also be used to determine whether a legal issue is just a local, accidental phenomenon or a systemic flaw, thereby helping researchers to falsify it. Self science methodology not only tends to collect and utilize more representative broad area data, but also focuses on scientific predictions for the future, that is, evaluating the possible consequences and achievements of research conclusions or reform plans through specific statistical models, machine analysis, and other means. In the author's opinion, this forward-looking research is clearly meaningful, as the limited resources for reform and the pursuit of stability in the operation of the legal system inevitably require decision-makers to only choose the plan that is most likely to achieve twice the result with half the effort or at least "balance of payments" for experimentation. In this regard, 'self scientific research' is equivalent to a preliminary but precise screening, with the ability to select the most practical reform plan for decision-makers. In other words, we can regard "self science law" as a paradigm revolution in research, which will change the fundamental research characteristics of social sciences driven by theoretical assumptions in the past, and add data-driven elements to form a new paradigm of "data-driven+theoretical assumption driven", ultimately achieving the transformation of legal research from "soft science" to "hard science".
The development of "self disciplinary law" may have a certain impact on traditional legal analysis methods, and legal researchers who were originally skilled in intuitive thinking and conceptual deduction can pay attention to this trend. From the perspective of the legal system, science, especially digital technology, has the characteristics of strong innovation and fast iteration, which will conflict with the relative stability of the former. But what we should see more is the opportunity brought by the transformation of legal research brought by "self disciplinary law". Specifically, the value of "self science and technology" is reflected in at least the following two aspects.
Firstly, "self disciplinary law" may solve the inherent problems of semantic ambiguity and obvious subjectivity in legal analysis. Some commentators point out that the mathematization of legal expression can liberate law as much as possible from excessive value judgments, rhetorical dependencies, and arbitrary variable handling, thereby promoting a more scientific realization of the rule of law. Although the author does not fully agree with the concept of "mathematization of legal expression", he also acknowledges that overly subjective expression and analysis are not helpful for the development of legal research, and may even deepen the "cracks" within the legal research community. Because the essence of some legal debates is only differences in formulation or expression, and the "solutions" proposed by scholars are mostly limited to adjustments in expression. The proposal of "self disciplinary law" is not intended to create a new paradigm for legal research in form, but to promote the improvement of the "standardization" of legal research and reduce the blockage of academic exchanges caused by unnecessary individualized expression.
Secondly, the development of "self disciplinary law" can enhance the "practicality" of theoretical research. The new liberal arts is a key path for the construction and development of disciplines in the digital society era. In my opinion, although there are multiple interpretations of the "new" aspect of the new liberal arts, the most fundamental aspect in terms of legal research is the emphasis on the "new" ability of legal research to solve legal problems. However, in the current overall situation, the actual contribution that academic research in law can make to this is limited. This is because theoretical research may be logically consistent, but often cannot be transformed into executable operational plans. For decision-makers, they prefer to see answers about 'what to do' rather than statements about 'what' or 'why'. On the question of "what to do", traditional legal research can only provide rough opinions based on the subjective thinking of researchers in most cases, but it is often difficult to answer questions such as how to determine the scope of reform, how to carry out reform, and the possible consequences of reform that require a neutral and rigorous evaluation to provide answers. It can be considered that the current legal theory lacks the research ability based on "numbers", which is precisely what "self disciplinary law" excels at, that is, strengthening the persuasiveness of theoretical research results by enhancing the degree of digitalization in legal research, and ultimately improving the ability of legal theory research to solve practical legal problems.
3.2 Between Technology and Thought: The Future of 'Self Science and Law'
The field of natural science relies on information technology, and legal researchers often have a vague understanding of natural sciences, including computer science, without knowing the reasons behind them. This intellectual ambiguity may give rise to far fetched so-called "natural science products". In the author's opinion, "self disciplinary law" needs to carefully handle the relationship between "technology" and "ideology". Only by striking a balance between the two can researchers make "self disciplinary law" a "new force" in the field of legal research.
3.2.1Technology: Enhancing the substantive scientificity of "self science and technology"
Due to past education, most legal professionals are unable to have sufficient dialogue and communication with technical personnel, which has caused many problems. The limited participation of legal professionals in the design of judicial artificial intelligence ultimately limits the effectiveness of judicial artificial intelligence products. At the deeper level of education, there is also an inherent perception of the disconnect between law and natural sciences, which is seen as two separate territories. Some scholars resist the integration of natural science elements into legal research. In terms of the future development of "self science and law", the urgent task is to enhance the integration of knowledge and technology between the research community and natural sciences, especially to grasp the ability to conduct legal research through mature technologies of data science. In other words, researchers in the field of "self science and technology" need to truly master and utilize the corresponding information technology means in order to produce academic achievements with practical value. In fact, the above phenomenon is related to the long-term separation of arts and sciences. The current promotion of interdisciplinary studies, including new liberal arts, actually has a sense of remedy, that is, legal education itself needs to be scientific.
What are the bottleneck issues that urgently need to be addressed in the current research of "self science and technology"? Scholars have pointed out that current legal models are effective in handling simple and clear legal relationships, but they perform poorly in complex matters involving various factors such as judicial forecasting. They believe that there are three reasons for this: firstly, the irregular structure of legal data; Secondly, the conclusions drawn from model production lack interpretability; The third is the conflict between the openness of judicial judgment reasons and the closedness of computer systems and language. In other words, at present, "self science law" can only produce products with simple variable relationships and cannot cope with more comprehensive legal phenomena. I generally agree with this view, but at the same time, I believe that self disciplinary law is not only a matter of data, but mainly a problem of the thinking paradigm of law itself. Of course, there is also a problem of how to quantify the value judgment of law. The difficulty of text mining that people value is actually only a technical issue. As for the insufficient compatibility between the above legal relationships and natural science models, it can be attributed to the expression of legal language.
It is worth noting that the conversion of text on paper is only one aspect of the problem of legal language expression. The real test of "self scientific law" is the automation of relatively abstract legal value evaluation standards. Logically speaking, after converting legal phenomena into machine language, it is necessary to provide a computer or legal model with a reference rule that can automatically identify under what circumstances and what judgments should be made based on the standards determined by the reference rule. The author believes that in pure manual analysis, the selection and application of legal standards rely on the subjective value analysis of judges. It is extremely challenging to replace human labor with machines or models to complete this task, and even impossible in the factual determination or legal judgment of many difficult cases: because it is uncertain whether "silicon-based" intelligence can grasp or even surpass "carbon based" humans. Some scholars believe that "no matter how comprehensive big data is, it can only record human behavior and cannot accurately describe human thinking." In my opinion, forcing computers to describe human thinking is not very realistic, but it does not mean that computers are completely unable to simulate human thinking, especially under the premise of clear facts. It is desirable to grasp and apply legal rules with clear boundaries. Due to the fact that simulation relies on the results of machine learning based on a large amount of empirical data, successful simulation is possible for two reasons: firstly, the current legal evaluation standards already have a basic framework, which means that even if legal phenomena are complex, the existing framework can still be applied on a large scale, and the special problems outside the framework are relatively few. Therefore, this to some extent narrows down the scope of machine learning tasks. Secondly, machine learning is based on pre-existing data materials, and the analysis and learning of such materials do not involve value judgments. What we need to do is to "feed" the machine as many samples as possible, so that it can continuously grasp various simple and complex situations, thereby improving its fit. In summary, the next research focus of "Self Science Law" should be to explore and solidify the cognitive functions of computers, and improve the ability of intelligent systems to understand legal data, conduct logical thinking, and engage in self-learning by simulating the cognitive thinking of legal professionals.
Obviously, the traditional natural sciences, humanities, and social sciences have become disconnected from the pace of the times in their independent research on "people," "things," and "objects." Breaking down the barriers between humanities and natural sciences and achieving their integrated development has become an inevitable trend. The urgent task of "Self discipline Law" is to strengthen the application ability of existing researchers in natural sciences and information technology. However, in the long run, we need to continuously expand the influence of "Self discipline Law" in the field of legal research, which relies on deep exchanges between disciplines. For example, the Chinese Computer Society established the Computational Law Industry Branch in October 2021 and held the first Computational Law Seminar. It is undeniable that this measure of building communication platforms for scholars from different disciplines is a good way to promote the deep integration of law and natural sciences.
3.2.2 Ideology: Avoiding the trap of technology supremacy
Technology is not omnipotent. For the development of "self disciplinary law", we must soberly recognize that technology can never replace human thinking. Overemphasizing the role and value of technology can easily fall into a trap of technologism and ultimately lead to the end of the development of this discipline. On the surface, it is only after humans entered the era of big data that the emergence of "self science and technology" was promoted. However, further reflection reveals that the needs of legal professionals are the reason for the formation of this new research paradigm. Specifically, with the increasing complexity of social conditions and legal phenomena, the cost of dealing with legal issues is beyond the capacity of human labor. Therefore, it is necessary to rely on natural science and technology to describe and analyze legal phenomena, in order to make up for the shortcomings that human labor cannot, and ultimately make the logic of problem-solving clearer.
It should be pointed out that although computer systems can simulate human thinking for "bionic" operations, simulation is only simulation and cannot and should not be equated with human creative thinking. Shen Weixing believes that the refinement of legal practice needs is one of the important conditions for ensuring the scientific and standardized application of computing technology in the legal field. In terms of precision alone, "self science and technology" can make efforts, and the ability to perform large-scale calculations, induction, and summarization based on correlation analysis is beyond the reach of humans. However, once we introduce the variable of whether it meets the practical needs of legal development, we will find that pure precision is not the only requirement. In other words, accuracy does not necessarily mean correctness. Taking judicial rulings as an example, at least in the Chinese context, the judgment results not only need to comply with the "law of the system", but also need to examine the "law of ideas" and "law of practice", and it is obviously unrealistic to completely rely on computer systems to complete the latter. Therefore, in the development of "self disciplinary law", the problem of emphasizing correlation analysis cannot be overcome. A feasible strategy to avoid this trap of technologism may be to maintain the creative thinking of legal professionals and expand their role in the application of technology.
Researchers in the field of "self disciplinary law" need to remember that humanistic care for the real legal world is always the first pursuit of legal research. The products produced by "self science and technology" should be "warm" rather than cold data. We should continuously develop and utilize the advantages of "self science and technology", but at the same time, we should be aware that information and technology are only tools, not goals, and the logic behind them echoes the answer to the question of "big data or small data". In recent years, "self disciplinary law" has attracted the attention of legal professionals with its many advantages and has emerged in the field of legal research. However, this advantage is also highly confusing - beneath the advantage lies the ubiquitous trap of technological supremacy. In previous research, the author pointed out that concern for human nature is the primary prerequisite for legislative products to be complied with by judicial personnel. In the future, both legislation and theoretical research should shift towards a subjective development path. Otherwise, both practice and theory may eventually become "flowers in the mirror, moon in the water". Isn't it the same for 'self science and technology'? Even if the accuracy of the legal calculation system continues to improve and the computing power is greatly enhanced, the final result given is only a "should be" equation. How to transform from "should be" to "could be" still requires attention to the role of traditional logical thinking among legal professionals, emphasizing the key values of subjectivity and humanity.
The development of "self disciplinary law" still needs to pay attention to many issues, such as "Algorithm Black Boxes" and ethical crises in experiments. It can be said that in dealing with any problem, even a slight mistake in "self disciplinary law" may lead to the development of the discipline going astray. In my opinion, the key to handling these issues well is still to respect the dominant position of legal professionals' thinking in legal calculations. The aforementioned issues are actually different manifestations of the trap of technologism, because once a mindset of technological omnipotence is formed, we are prone to overlook the flaws in technology itself, and only thinking full of human warmth can illuminate the continuous traps.
Conclusion
From a historical perspective, the pattern of legal research in China is undoubtedly becoming increasingly complex, and traditional and singular concepts are no longer sufficient to summarize the endless research paradigms. In the current era of constantly emerging new terms and technologies, researchers who wander in between are prone to losing their sense of direction, which also leads to continuous disputes over disciplinary affiliation. It is undeniable that this debate has limited significance for achieving the fundamental goal of legal services for the rule of law. Although this article reintroduces concepts such as social science law and empirical research, and proposes a tentative and framework based new category of "self disciplinary law", it does not intend to intervene in the ongoing academic debate. Because the "theoretical red line" of this article has always been clear: there may be varying degrees of differences between disciplines and research paradigms, but for researchers, the fundamental mission is how to transform theoretical research results into legal productivity with practical significance. In the process of transitioning to a modern rule of law country, our identities are no different - we are all legal professionals who need to forge ahead and contribute our modest efforts to achieve the goal of a rule of law country. What is the future of "self science and technology"? This depends on whether there is a group of creative interdisciplinary researchers who continue to dedicate themselves and work hard. Let's wait and see.