Know-how. New Challenges and Trends for the Legal Translation Field

TABLE OF CONTENTS

TABLE OF ABBREVIATIONS

1. INTRODUCTION

2. CHALLENGES AND TRENDS IN THE LEGAL TRANSLATION MARKET AND LEGAL MARKET
2.1. LEGAL TRANSLATION AS A COMPLEX TRANSLATION FIELD
2.2. MAIN PLAYERS IN THE LEGAL TRANSLATION MARKET
2.3. KNOW-HOW (KH) AS NEW WORKFLOW ELEMENTS
2.4. STATE-OF-THE-ART IN LEGAL TRANSLATION STUDIES
2.5. CHALLENGES AND TRENDS IN THE LEGAL MARKET
2.6. CHALLENGES AND TRENDS IN THE LEGAL TRANSLATION MARKET

3. STRATEGIES
3.1. DATA STRATEGY
3.2. ARTIFICIAL INTELLIGENCE (AI) AND MACHINE TRANSLATION (MT) STRATEGY
3.3. TERMS STRATEGY

4. CONCLUSION

BIBLIOGRAPHY

TABLE OF ABBREVIATIONS

AI - artificial intelligence

MT - machine translation

MAHT - machine-assisted human translation

HAMT - human-assisted machine translation

CAT Tools - computer-assisted translation tools

ABSTRACT

Know-how are continually transforming the world and global markets, including the legal market and legal translation market. Research has shown an increased interest both from academics and market players, yet the absence of a realistic overview of the current and future impact of know-how on both markets. This study aims to determine what the main contemporary challenges in the legal translation field are and how future trends will shape the future of the legal translation market. Building on results of recent conferences in Germany 2019, academic works and talks to different experts, this study further summarizes existing strategies on how to address the major changes in order to survive and thrive in the legal translation market in the future. Based on the reverse engineering method and analytic and systematic approaches, the possible interactions between complex systems such as new technologies were studied from a global perspective, so that realistic future trends were presented based on dynamic simulations of these systems.

The results confirm the hypothesis that the development of know-how implies the radical transformation, e.g. the entire business model for delivery services, which will involve outcome-based strategies and empowering human resources as enablers of technology. On this basis, it is recommended that legal translators should take strategic decisions on how to adapt to new industry realities and necessary market transformations. For that reason, three main strategies were presented addressing the most recent challenges such as artificial intelligence, machine translation and data strategy.

The results of this study mostly address German legal market and German legal translation market but can be used for further research in order to identify other strategies or strengthen the effectiveness of these strategies in other countries.

INTRODUCTION

The development of the modern world is characterized by the processes of globalization, internationalization, internet connectivity, technological breakthroughs and innovation, which have shifted the whole paradigm of human interaction in life and in business. These processes have made it possible to reduce the distance between countries and to accelerate and intensify human interaction at different levels and for different purposes in businesses, governments, economies and people's everyday lives. The role played by new translation technologies in the rapid development of globalization is not insignificant. Moreover, the English language has become established as a lingua franca in these globalized interactions.

The legal sector plays a very important role in these processes: it focuses on attempting to protect people, businesses and governments from the market players from other countries. For this reason, the legal translation field aims at achieving effective and reliable communication in different fields and at different levels.

An advantage of this development is that new technologies have completely changed the world. Such technologies have globalized and internationalized business processes, making the activity of business communication in nearly all countries faster, more accessible and more comprehensible. However, this positive impact has benefited not only businesses and business processes, but also many other individuals, organizations and activities, including those experts behind the scenes of translation: translators and translation processes.

The disadvantage of this development is that new technologies are continually changing the world and global markets as a whole, consequently altering people's lives. Moreover, since the development of new technologies is continuous, the paradigm of human interaction changes with this process, which also forces necessary changes in the translation market.

Against this general background, translators raise the main questions of this thesis: what are the main contemporary challenges in the translation field, and how can translators address these challenges to survive and thrive in the market in the future? How will future trends shape the future of the translation market, and more essentially, the profile of the translator?

The field of legal translation is the main point of interest in this thesis. This focus is preconditioned by my professional background. For several years, I have been engaged in the development of legal translation processes in large international organizations and businesses. My main duties have included co-drafting, translating and negotiating legal documents with international partners and authorities from all over the world. I have had to face challenges for which it was necessary to find solutions. Therefore, I have considered it relevant to share and to research these challenges and solutions further at the academic level, based on the results presented in recent resources and both practical and theoretical research in my field.

First, after stating the main objectives of the thesis, I will explore the major challenges facing legal businesses of different sizes and the specialization resulting from new technologies and recent developments in the legal market, the legal translation market and legal translation studies. I will then summarize these results. Second, I will provide an overview of recent trends and business strategies aimed at addressing the contemporary challenges and trends for different players in the legal translation market. My thesis consists of two main analytical parts.

First, I consulted the principal scholarly works on legal linguistics, corpus linguistics, artificial intelligence (AI), machine translation (MT), computer-assisted translation (CAT) tools and software, as well as interdisciplinary works on business processes, legal works and other related topics. Since I had access to the national library in Frankfurt, Germany, I was able to consult the most recent publications in these fields, up to end of 2019. Furthermore, I studied blogs produced by different informational companies, software manufacturing companies, freelance platforms such as Proz.com, and translator groups on social media to develop ideas for further research at the business level. Finally, I visited several relevant trade fairs and conferences in the legal sector and legal translation field, especially events addressing new technologies.

Given the complexity of this topic, it is necessary to highlight that only the most relevant and most recent resources were consulted. Nevertheless, it should be noted that it was neither feasible nor possible to cover all existing resources for all countries and language pairs.

Throughout the thesis I will use following terms:

Market players:

Law firms are companies specializing in providing legal services.

Legal businesses are all market players in the business sector in general, including law firms of any size and legal form, as well as the legal departments of other businesses.

Legal market encompasses all market players in general, including law firms and the public sector.

Legal translation market includes all legal translation businesses as market players of any size and legal form.

Legal translators include all human individual translators specializing in providing legal translation services.

Know-how (KH) includes all innovations and new technology which have or could have an impact on legal translation and the workflow, processes and quality. Such technologies can include CAT, human-assisted machine translation (HAMT), AI, blockchain technologies and various other tools.

When describing the current situation in the field of legal translation, an increased interest can be noted at the following levels: (1) the recent conferences on the impact of KH in both the legal sector and the field of legal translation, (2) the growing number of academic works on new technologies and (3) the increasing demand for legal translation skills and services.

Nevertheless, in evaluations of the current situation, there are conflicting opinions about the impact of KH on the legal market and the legal translation market. In broader society and in academic works, business texts and other sources, there are many misconceptions about the respective futures of KH, the field of legal translation and the profile of the legal translator.

There appear to be essentially two camps: all the works and opinions that focus only on the positive aspects which relate to AI and MT implementation in legal translation and are in favour of and working towards these advantages of KH, seeing the KH as a panacea and a goal of human development; and, conversely, the works and opinions that are shaped by the negative aspects involved in facing the disadvantages and the necessity of implementing of the popular tools.

This situation raises several questions. What is the impact of KH in reality? The absence of a realistic overview of the field, featuring relevant strategies based on current challenges and trends, should be considered a significant gap in the literature.

In other words, are all aspects of KH already integrated into legal businesses and legal translation businesses? Are all players aware of all KH and thus able take informed decisions on whether, how and what KH they should implement and which KH they should not apply? Expanding this reflection, are all legal translation businesses operating in the legal translation market aware of these questions and able to explain the profit of considering and addressing such questions to the client or employer?

To identify the importance of the proposed research, the topics of the recent conferences in the legal industry and the translation industry were judged objectively. The most urgent topic was determined to be technologies and their impact on these industries and players. But why is it so important to the global community and industry players?

Accelerated communication is already an essential part of successful globalization and effective business practices. The primary task in this communicative process is to provide legal translation that is simultaneously reliable, fast and of sufficient quality to justify the costs and time spent on producing such translations.

The main research problem is to show future trends and current challenges for the legal translation businesses using present or future industry know-how, so that the main gaps in the industry can be revealed for different legal translation market players.

The results of the thesis could be used to help raise awareness for both legal translation businesses and legal translators, helping these groups to take strategic decisions to adapt to new industry realities and necessary market transformations. Furthermore, for legal translators, the outcomes of the thesis could serve as a starting point for planning their continuous professional development (CPD).

The main hypotheses are as follows:

Due to the rise of KH, the markets will be changed radically, though there are trends that will allow transformation strategy to be predicted, so that creating and delivering business value will be predominant.

The profile of legal translators and business models will also be transformed, so the best transformation strategy for legal translators to ensure their survival will be to proactively assess the challenges and trends in the market realistically and then formulate appropriate short- and medium-term strategies.

Selecting the appropriate methods was crucial for the success of the thesis. To conduct this research, I have therefore combined several methods: the reverse engineering method and analytic and systematic approaches.

Analytic and systematic approaches are complementary analysis methods that facilitate the consideration of complex systems such as new technologies from a global perspective to handle such systems more realistically and investigate the possible interactions between them. In this way, future trends can be predicted based on dynamic simulations of these systems.

Future trends result from current challenges, but they cannot be predicted with sufficient accuracy, since facts from the future are inaccessible and time machines do not exist. However, there is an appropriate method, namely reverse engineering, which is widely used in business and can help determine an economically reasonable business strategy.

To better understand whether an investment in a new technology or strategy will generate business values in the long term, it is crucial to assess, as point B, the future business values of the technology or strategy, while also assessing, as point A, the current situation. This approach ensures that the route from point A to B will be feasible, while also allowing different scenarios to be elaborated for different time frames, such as for short-, medium- or long-term strategies, depending on the results of the assessment. Utilizing this framework, I will present new and relevant trends in both the legal and legal translation market. I will then critically assess these trends, evaluating their further impact, their advantages and disadvantages for specific market players and, where relevant, the time and efforts required for their implementation.

To conduct such prediction and assessment, two points are required:

(1) point A, the current status and current challenges in the markets, based on facts and research findings;
(2) and point B, the anticipated development for the future, based on current industry and academic perspectives and other recent resources on trends.

I will structure the thesis according to the method of reverse engineering. In the second and third chapters I will present points A and B accordingly for the legal translation market and legal market. Point A, the current challenges, includes the analysis of the current findings in legal translation studies and the current situation in the legal translation market and legal market. Point B, the development trends, includes the relevant short-term and medium-term strategies, which are considered as possible solutions.

2. CHALLENGES AND TRENDS IN THE LEGAL TRANSLATION MARKET AND LEGAL MARKET

The main questions this chapter seeks to answer are as follows:

2.1. What are the key differences between the legal translation field and other fields of translation?
2.2. Who are the main players in the legal translation market?
2.3. What are the current workflow elements in legal translation that are likely to be changed in the industry?
2.4. What is the status quo in the fields of translation and legal translation studies, as indicated by a brief analysis of recent academic studies?
2.5. What is the legal technology paradigm shift that is most anticipated to alter the status quo in the legal market, as indicated by an analysis of development trends at recent professional conferences and trade fairs in the respective fields of legal translation, legal tech and legal business?
2.6. What major challenges will result for the legal translation market?

2.1. LEGAL TRANSLATION AS A COMPLEX TRANSLATION FIELD

Legal translation is considered as a complex translation field for multiple reasons.

First, legal texts are formulated in languages for special purposes. Moreover, legal texts encompass many different legal genres and styles (Matulewska, 2013), which can vary in different languages. Individual legal genres have their own central set of legal terms, with differing degrees of use of legal English in different national legal systems, and such characteristics are areas of current and future research, such as in legal corpus linguistics (Gozdz-Roszkowski, 2011, pp. 232-233).

Second, translation and legal translation have an applied character and function within the frameworks of specific legal systems, which can vary widely in different countries (such as reflected in common law or civil law). Moreover, different legal systems are bound up with the specific culture and legal history of the country in which they operate and even specific legal processes: legal problems are resolved in different ways in different countries, through different institutions, based on terms of specific technical structures tailored to individual legal systems (Groot, 1991). Therefore, the choice of terminology depends not only on the languages of the legal systems in the sense of languages for special purposes, but also on the expertise of the legal translators: their ability to understand, interpret and produce legal texts depending on the legal systems, along with their expertise in both the legal systems of the source language (SL) and those of the target language (TL) (Sandrini, 1999, pp. 9-44, 45-63). Such expertise is the prerequisite for entry into the profession.

Third, analysis of the history of the development of legal translation highlights not only that in recent years legal translation studies has been recognized as a distinct field of translation and an autonomous academic discipline, but also that such studies are now considered a fast­growing research field, with interest from both industry and the public sector, especially with regard to contemporary practical and theoretical problems. For instance, some studies seek to investigate the practical problem of the parametrization of the legal translation process (Matulewska, 2013).

Fourth, there are different equivalence theories and approaches, which try to resolve the main equivalence difficulty that arises from the different aspects outlined above, namely that different concepts and legal norms do not straightforwardly correspond to each other because particular legal concepts are strictly bound to a specific legal system, culture and society at a specific point of time (Tessuto, 2008). Many researchers and practitioners share the opinion that an absolute equivalence between the concepts and norms of different legal systems is unachievable (Sarcevic, 2006), which impacts the whole translation process, with the consequence that legal translators should try to devise the best possible partial equivalence (Groot, 1991) or sufficient equivalence (Matulewska, 2013, pp. 217-237).

Fifth, a legal translator must not only be able to interpret the type, style and genre of the source text (ST) (Matulewska, 2013, pp. 32-38), as indicated above, but also to focus on the requirements of the message recipient or target audience, which necessitates further analysis of the purpose of the target text (TT) for the target audience, the author and/or the assignor. Only after this initial interpretation and analysis is completed the actual translation process can begin (Matulewska, 2013, pp. 18-20).

When considering earlier TL-oriented equivalents approaches and later studies which differentiate the further types of recipient, Sarcevic's (2006, p.15) recent approach should be highlighted, as it focuses on achieving an equivalence between the legal effects of both SL and TL, not only on determining the respective meanings of the terms.

This leads to the sixth criterion for a legal translator and his or her profile (competences, properties) to play the central role in the legal translation process: the requirement that such a translator possess substantial legal knowledge and translational skills, since the legal translator must have not only the competence required to understand the meaning of the ST, but also the competence necessary to transfer the intended meaning and legal effect to the TL (Matulewska, 2013, pp. 46-48).

Considering new challenges in the market and changes in the legal translation workflow, the legal translator and his or her profile has become even more important.

Therefore, it is essential to determine new milestones within the legal translation process, major market players and those challenges that impact the legal translation process and market players.

2.2. MAIN PLAYERS IN THE LEGAL TRANSLATION MARKET

According to the estimations of Florian Faes (2019, p. 24), the language industry intelligence platform which provides business insights on demand, talent transfers, technology and other elements of the translation and language technology markets, the demand for digital translation will rise from 23.2 billion USD in 2019 to 28.2 billion USD by 2022, with the largest customer segment forecast to be the public sector, followed by the technology sector. This demand is currently served by several types of providers, including super agencies, leaders and challengers, as well as smaller market players such as freelance translators, small agencies, internal language service providers (LSPs) and public organizations (Faes, 2019, pp. 25-26).

At present, the five super agencies TransPerfect, Lionbridge, SDL, RWS and Welocalize have a 10% share in the market in total. These agencies can provide services worldwide 24/7 and can meet the requirements for long-term tender and the complex needs of large companies (Faes, 2019, pp. 25-26).

The second group, leaders, has a 12% share in the market: their combined turnover is 25 million USD per year. This group consists of the German company Kern AG, the Danish business Language Wire and the French Acolad group. Their characteristics are industry or regional specializations (Faes, 2019, p. 25-26).

The third group, challengers, has a 5-7% share in the market: their combined turnover is 5-25 million USD per year. This large group includes providers with subject specialization such as German EVS Translations, legal and finance translation specialists, and small agencies acting as subcontractors for larger agencies (Faes, 2019, p. 26).

The fourth group is smaller LSPs, smaller market players with a combined turnover of nearly 8 million USD per year, which constitute different proportions of the market in different countries. For instance, there are 500 LSPs based in Germany, while over 2,500 LSPs currently operate in the UK (Faes, 2019, p. 26). However, according to the information registered in Xing and LinkedIn profiles, there are more than 600,000 individual profiles and approximately 26,000 company profiles in total (Faes, 2019, p. 26). Moreover, this group also includes internal LSPs working for large companies and public sector organizations.

However, regardless of the size and specialization of different providers, new challenges will impact all market players in the translation and legal translation market, as estimated by Slator (Faes, 2019, p. 27). To succeed, these providers will need to transform the way they deliver their services, which entails both harmoniously integrating new processes into the workflow and empowering HR as enablers of technology (KPMG, 2019, pp. 3-4).

2.3. KNOW-HOW (KH) AS NEW WORKFLOW ELEMENTS

The survey 3Q18 Global Pulse conducted by KPMG reveals such areas as driving down operating costs and exploiting opportunities for digital transformation as the top two business challenges currently facing companies worldwide, followed by securing talent, digital security, and regulatory compliance (KPMG, 2019, pp. 2-3). Apart from this group, there are challenges specific to companies operating in the legal market and the legal translation market. Moreover, legal translation has an applied character that renders it dependent on the changes in the market it serves.

As the first step, I will present the challenges for the work and business processes, or put in business terms, the business delivery model. The last decade brought new challenges for the market and for research: the need to integrate new technologies into the workflow of translation to digitalize, accelerate and automatize the many steps in the translation process. Moreover, these changes affect not only the implementation of new technologies for translation, such as CAT tools and MT, but also the nature of the tools required for project management, data digitalization, data management and data protection in compliance with the General Data Protection Regulation (GDPR), speech recognition technology and new legal tech applications. In the case of legal translators, this process leads to greater requirements for translation speed, price and quality in the market, and for language service composition, including post-editing services or eliminating simple translations that can be performed faster by machines (Czopik, 2019, pp. 30-40).

On the one hand, new technologies are being developed to accelerate and facilitate the process, keeping in mind that the legal data involved is increasing in amount and complexity. Therefore, new translation technologies can be considered as outstanding future opportunities.

On the other hand, as these new technologies become more complex and refined, it raises many questions and concerns about whether and how it remains possible for translation service providers and individual translators to keep pace with these recent developments to succeed in the market. In this sense, new technologies generate new challenges that entail radical changes in business models and competence profiles.

However, several experts on the translation market, including software manufacturers such as Loctimize GmbH and Lionbridge Technology, consider as possible solutions to the potential challenges posed by new translation technologies proactive assessments of their own workflow to create a hybridized workflow model that combines both human and AI strengths, while at the same time eliminating weaknesses (Marciao, 2019, p. 41).

This proposal requires the assessment of all those elements in the workflow that could be improved, such as by being digitalized, automated, covered by machine translation, accelerated through creation of qualitative of translation memories (TM) or made more efficient by other means.

Among all approaches considered, I have highlighted the concept of reverse engineering. While this method's focus is typically restricted to quality assessment (Hoppe, 2019, pp. 229­239), it can be extrapolated to the entire translation workflow. By encouraging translation and legal specialists to cooperate more closely, the elements which can be further optimized are revealed.

To determine the optimal workflow elements for the average legal translation project or process, the following steps should be considered:

1. How to ensure a flawless and compatible digital environment; for instance, by using process-compatible formats and systems, new software for accounting, communication tools or other similar approaches (Hoppe, 2019, pp. 229-239).
2. How to provide a compatible legal translator for the project(s), based on the evaluation of the needs of the specific legal business (assignor, target audience) and the profile of the legal translator (e.g. education, experience, IT skills and other skills).
3. How to assess which steps and elements of the digital environment can be optimized or transformed using existing technologies.
4. How to assess the economical reasonability and terms of such transformation, so that it can be implemented on a strategic basis, based on the informed decisions of decision­makers in the companies involved, for both legal business and translation business.

For instance, based on the cases from translation practice presented at the Conference on Translating and Interpreting 4.0 (held in Bonn from November 22-24, 2019) and recorded in the proceedings of this conference, there is remarkable potential for the integration of the kinds of hybridized models outlined above, even in such areas as certified translation (Said, 2019, pp. 50-59).

I extrapolated the following formula, which was initially applied to the implementation of the translation memory system for the assessment of the whole workflow, necessitating that the workflow be adjusted to ensure optimal use: ‘Current efforts to operate the step without optimization + relevancy of consequences (if not optimized) / Efforts to implement = value for the business' (Hoppe, 2019, p. 233).

The concept of “optimal use” may be significant in this sense, since the implementation of technologies is typically considered in terms of solutions to specific business problems, and furthermore, is viewed as a strategically informed decision, involving investments. Moreover, the need to take such decisions affects all market players, regardless of the size of any particular translation or legal business. The concept of optimal use introduced above can be replaced in this context by the business term return on investments (ROI).

Why is the concept of ROI so important, particularly in the current market situation? I will explain its importance based on the internal information gathered during one of my interpreting assignments for the Russian Ministry of Industry and Trade for my talk with Geoffrey Willis (2019), the International Senior Apparel Industry Executive of Trigon Select, a company with expertise in performance evaluation and intelligence systems for better performance in apparel. Geoffrey Willis is a well-known specialist with over 40 years of experience and outstanding management and training skills. In his opinion, the most important step is not just to incorporate any high-tech measure into the workflow and then wait for the whole process to be accelerated. Rather, the idea is to not only execute a step faster, but to execute it more efficiently (Willis, 2019). This approach can also be applied to the current legal translation process. If the company decides to implement an MT tool that will translate material rapidly, but with reductions in quality that make the tool insufficient for its specified purpose, then the tool is not being optimally used or implemented. By using the reverse engineering approach, a company can determine the specific quality it needs to meet its business requirements, only deciding which tools would be optimal for this purpose after the initial assessment has been conducted.

However, given the current speed of development of high-tech tools and the market, this rapid change can impede the assessment of which elements in the workflow can be improved. To facilitate the required technical due diligence, I will present the main challenges and innovations in the fields of translation and legal translation in the next section of this chapter, beginning with the state-of-the-art of the field of legal translation studies and then continuing with the legal market and legal translation market.

2.4. STATE-OF-THE-ART IN LEGAL TRANSLATION STUDIES

What is the state-of-the-art in legal translation studies, based on an analysis of recent studies and academic works concerning this topic?

When analysing the development of legal translation, it can be noted that in recent years legal translation studies has not only been recognized as a distinct field of translation and an autonomous academic discipline, but also established as a rapidly growing research field that incorporates diverse approaches from different disciplines, such as computer science, law and economics, resulting in a multiplication of the types and areas of current research (Ruano, 2019, p. 130). Moreover, as the second marker of the field's current status, Rosario Martin Ruano (2019, pp. 137-138) highlights the shifts from qualitative to quantitative data and from prescriptive to empirical, descriptive, and data-driven computer-assisted approaches, combined with the prevalence of corpus-based research methods. This methodological change has resulted in subjective knowledge being replaced by empirical data, which have many applications, such as in legal training or as best practices for making informed translation decisions. Critically assessing current developments in descriptive legal translation, Rosario Martin Ruano (2019, pp. 137-138) points out that such empirical findings have an applied potential that has not yet been fully explored. Ruano (2019, pp. 137-138) holds that these new approaches should not be allowed to develop into prescriptive practice, but should rather be understood as dynamic snapshots of the changing reality of legal language in motion. Legal reality is not fixed, but rapidly changing due to the impact of the new challenges of globalization and internationalization, the development of automatization and other new technologies, and the use of Legal English as the global legal lingua franca (Ruano, 2019, pp. 145-146). Moreover, the legal translation and legal translation studies fields should be capable of adapting to such technological developments and constant changes due to their applied character.

When exploring the recent opinions of some authors on this subject, a new paradigm can be discerned. This outlook highlights legal language and legal translation as the tools of new global legal language and communication (Gozdz-Roszkowski, 2011, pp. 232-233), resulting in the automatization of legal translation practices, hybridized legal languages and new models for legal translation, which should be constantly re-detected and updated. Such updating processes require a more intense exchange of empirical data between legal practice and legal translation practice, which should be further integrated with the results of legal translation studies (Ruano, 2019, pp. 145-146). However, implementing such methods entails radical changes in sociological approaches to legal translation and the empowerment of the professional profiles of legal translators in general as important market players.

Consequently, despite the projected progress in research and practice in legal translation studies, based on the latest comparisons from different countries, there is a need in practice for overriding transformations. For example, by reconsidering unified definitions of basic concepts such as legal translation, judicial translation or sworn translation (Albi & Ramos, 2015), by fostering the creation of institutes of sworn translators where required or by adopting measures designed to improve the professional image of legal translators as important performers in legal communication.

Studies that focus on the observation of legal translation practices can also contribute to replacing subjective knowledge with empirical data. In this regard, Ruano (2019, pp. 148-149) indicates that the target audience, context, culture and practice of legal translation are important factors.

To sum up, there is an increasing interest in legal translation studies, which is producing varying approaches and results, highlighting the need for a more intense exchange of empirical data between the fields of legal practice and legal translation practice, combined with the active integration of relevant results from legal translation studies.

2.5. CHALLENGES AND TRENDS IN THE LEGAL MARKET

As has been noted both in academic studies and at a recent conference for the translation industry in Germany in 2019, the entire translation market needs to undergo radical change due to digitalization and new technology. In turn, radical adjustments are required in the profile of legal translators and the business model of translation businesses and legal businesses to ensure an adequate response to these changes. However, it is often overlooked in current debates that translation has an applied character, which means this activity cannot exist without the industry or the market which it serves. In this sense, the translation market must prioritize adjusting to the changes in the market it serves, namely the changes currently occurring in the legal market. Therefore, in this chapter, I will provide a brief outline of those challenges in the legal market that have an impact on the development of the legal translation market. In the next chapter, I will examine the main challenges in the legal translation market.

The first concept that ideally describes the status quo in the legal market is legal tech. To envision the current developments and future trends in the legal market, I attended professional conferences on legal tech, including the legal innovation conference Legal Revolution 2019 in Frankfurt in Germany, and conversed with experts from different law firms, courts and other legal market players. I then analysed the results and conducted follow-up research using recent academic literature and corporate webpages. In this chapter, I will present these results regarding legal tech. The results are mostly based on the legal market in Germany.

Legal tech is a fixed concept, an abbreviation of the term “legal services and technology”, which means the implementation of new digital technologies to facilitate and automatize legal services, whenever possible. In general, any legal business proceeds through three phases of legal tech: Legal Tech 1.0 includes the basic digitalization steps (e.g. document management systems, electronic accounting tools, legal electronic databases); Legal Tech 2.0 encompasses automatization standardization in the workflow (e.g. in the creation of legal documents, such as contracts from building blocks); and, finally, Legal Tech 3.0 focuses on the implementation of AI (Wagner, 2018, p. 7). While Legal Tech 3.0 is the current trend, not all phases of legal tech are implemented equally in the legal industry.

Legal Tech 3.0 began in Germany in 2016 and was mainly aimed at generating feasible economic profits though cost saving and the simplification of processes. However, the progress in this phase was limited in accordance with the so-called Legal Services Act in Germany (German: Rechtsdienstleistungsgesetz), and it is thus important to highlight that the German legal market is considered to be rather conservative in this sense (Bauer & Mayer, 2019, p. 140). This topic is currently broadly discussed in legal society: the limitations imposed by this law have begun to blur because there are several permitted services, along with numerous grey zones in the German legal market, such as online automated accounting systems (Wagner, 2018, pp. 45-46).

When analysing the impact of legal tech on the development of the German legal market, several recent trends that are essential for the legal translation market (at least in Germany) can be highlighted.

Challenges First, there are at least four main challenges that companies aim to resolve using legal tech in the German legal market (including the public sector) in the coming years (Bauer & Mayer, 2019, p. 140):

1. Many legal businesses and public sector organizations are being forced to bear the increasing cost pressure, since those companies and public authorities to whom the legal services are provided now require transparency and increasing efficiency for legal operations costs, for legal departments operating as an internal legal service providers, for legal advisors functioning as an outsourced legal service provider, and for legal service providers working in the public sector.
2. Large law firms tend to implement new technologies on their own initiative. They believe it is necessary to improve their image as innovative law firms, to offer customers a cost advantage and to become more competitive in the legal market.
3. Law firms also seek to expand their portfolios and profits through new IT-based products, which they develop themselves with the intention of (re-)selling these new products. Alternatively, such firms invest in IT start-ups with the same goal. These IT start-ups develop new legal tech applications that can be sold as individual products or implemented in the legal operations of law firms to increase the efficiency of or automatize these processes. A further alternative is the outsourcing of products and services to other law firms. Finally, these IT- start-ups offer the IT-based legal tech solutions just outlined to the customers of the legal firms, creating at the same time additional competitive pressure on the legal market.
4. In addition, since legal markets are directly experiencing the implementation of legal tech and its impact, new legal questions are raised, such as those regarding data protection or legal security in cloud-based solutions. Taking into consideration this lag in the development that involves businesses and end-users exploiting the advantages of AI and other new technical solutions to solve their problems, and considering that legal services are of high importance for society since they provide legal security, the legal market will move in several grey legal zones. Therefore, rather than being conservative, the public sector must accelerate legal tech processes and change national laws where required. The first steps towards the full implementation of legal tech are currently being designed and executed to ensure that within the next five years at least the phase Legal Tech 1.0 (digitalization) will be mandatory (Doumanidis, 2019, pp. 131-138), so that further steps must follow soon.

Future trends In light of the above analysis, the following future trends can be considered a solution to the main challenges in the German legal market outlined here (if those trends that are most relevant for the development of the legal translation market in Germany as future trends are selected, as will be more thoroughly discussed in subsequent chapters) (Bauer & Mayer, 2019, pp. 141-142):

1. Economically reasonable optimization of the workflow

The continual implementation of all steps of legal tech (industrialization, standardization, automatization and AI) across the entire legal market will cause radical changes in traditional workflow and force economically reasonable optimization of individual legal processes in the workflow, such as document and knowledge management systems, leading to the achievement of higher ROI (Hartung, Bues, & Halbleib, 2018).

Several solutions for changes in the workflow are currently employed in the legal advice sector, whereas legal tech is implemented as the substitute for only two major functional steps: the ascertainment and analysis of legally relevant facts of the case, and the legal assessment of the given facts of the case. The legal tech tools presently utilized include chat bots and query masks (e.g. Modria); e-discovery tools (e.g. e-dicovery.at); IT-supported creation of contracts (e.g. lawlift.de), customized IT solutions tailored to the needs of specific businesses and smart contracts (Breidenbach & Glatz, 2018). Other legal businesses will need to follow up on these changes.

Several solutions for business models are offered below (Bauer & Mayer, 2019, pp. 144­145):

a) Law firms found or co-found legal tech start-ups with IT companies or allocate larger proportions of their budget to legal tech initiatives. In general, the increasingly intense cooperation of IT-companies with law firms and the rapid spread of legal tech applications in the legal market is already predicted by the results of recent studies. Furthermore, many legal products will be merged with digital products and sold as their integrated components.
b) In large law firms, IT applications will replace the initial steps in the legal advice process, which will affect mostly entry-level lawyers.
c) Medium-sized and small law firms will need to specialize in certain services by implementing only the specific IT tools required for those services to differentiate themselves and be more competitive in the legal market.
d) AI-based products are expected to become dominant in the legal market only in a mid-term horizon. Nevertheless, such products will cause disruptive changes in supply chains and business models in both the legal market and legal translation market. Thus, it is advisable for market players to use the remaining time available for transformation wisely.

2. Big data, (international) networking and blockchain.

The amount of digitalized legal data continues to grow as a result of several factors. First, the digitalization of the communication format in the public sector, including public authorities of the judicial branch and electronically accessible laws and legal texts of the legislative branch and the executive branch. Second, large businesses such as Freshfield are creating new support and legal tech centres and platforms. In general, the future trend is considered a (digital) data- driven decision-making process (Hartung et al., 2018), where all participants will be able to work in a virtual team, even while located in different countries or deciding on legal matters. Moreover, the spread of virtual courts, smart contracts using blockchain systems and online shared court documents is already predicted as part of the new networking strategy (Breidenbach & Glatz, 2019). However, in the framework of this thesis, I will only highlight big data strategy as the most relevant factor for the legal translation market because networking, blockchain and other more complex strategies should be addressed separately, based on the specific needs of the legal business.

These trends will result in an increasing demand for legal translation, because new legal translators will be recruited to translate growing legal data, at least as a short-term solution.

To sum up, legal tech is anticipated to be both the most important challenge and the most effective solution for the rapidly changing legal market in the coming years, judging from the initial movement towards digitalization and the increasing investments in AI, to take the most recent trends, although these processes will not occur uniformly for all market players. Such trends as the optimization of workflow and big data, networking and blockchain will probably lead to changes in the supply chains of legal services and legal translation services, although most probably only in a medium-term perspective. In general, the changes currently occurring will lead to an increasing demand for legal translation, at least as a short-term solution. 2. ce translation has an applied character, in the previous chapter I highlighted the main challenges and trends in the legal market which will most probably impact developments in the legal translation market. In this chapter, I will describe this impact in further detail, based on the results of recent conferences and academic books on legal tech and legal translation.

Based on the empirical data collected by Holger Knoblauch (2019, pp. 144-145), who held discussions with legal tech lawyers in various German law firms, and my observations from the recent Legal Revolution 2019 conference in Frankfurt, Germany, I will briefly summarize the results below:

1.6. CHALLENGES AND TRENDS IN THE LEGAL TRANSLATION MARKET

1. Sin Adapting translation technologies to legal tech changes. The subject of legal tech is attracting greater attention, with the result that more lawyers are required to implement legal tech in their law firms as part of their billable hours (Knoblauch, 2019, p. 145). Consequently, the legal translation market is facing a growing need to adapt the technological solutions currently used in the legal translation market to newly developed legal tech solutions, at least in the medium-term future and at least for the most important clients, since incorporating new technologies is a worthwhile investment from the medium-term perspective.
2. Basic translation skillset and legal English. By contrast, in my opinion, based on my conversations with the leading experts at the international Legal Revolution conference in December 2019, another hidden trend can be discerned: greater numbers of German lawyers are being required to improve their legal English skills and sometimes obtain a basic translation skillset in their law firms as part of their billable hours, on the assumption that German lawyers will need basic understanding of the increasing number of documents composed in legal English. In the future, more legal texts will be created with the assistance of sample templates and text building blocks with software support, so that only a one-time translation will be needed for such blocks, while lawyers will have a post-editing or pre-editing role. Alternatively, if such MT is of sufficiently high quality, lawyers will be able to complete the translation from the blocks, and a basic translation skillset will be adequate for the task. In this case, the demand for pure legal translators will decrease, and fewer classical assignments will be offered to legal translation businesses (Knoblauch, 2019, pp. 146-147).
3. Multilingualism. New legal tech products will be multilingual by default, either through multilingual programming or integration with MT tools (Knoblauch, 2019, p. 145), resulting in greater numbers of (at least one-time) legal translation assignments.
4. Localization and internalization. Multilingualism also requires implementing new legal tech products in other markets and creating more jobs for legal translators as experts in the localization and internalization of existing and legal IT-based products (Knoblauch, 2019, p. 145).
5. Classical legal translations produced by humans will disappear in several niches, where the legal steps will either be replaced by or integrated into new AI- and IT-based products (Knoblauch, 2019, p. 145). However, there are several areas and language pairs which will be less affected in the short term, such as rare areas of law, highly complex individual cases, mandates and cases with a strong personal connection, such as criminal law cases (Knoblauch, 2019, p.147). Nevertheless, there are several strategies for dealing with these challenges and developing possible solutions. For instance, Holger Knoblauch (2019, p. 145) suggests as a solution from the medium-term perspective creating new or additional specializations in areas and language pairs which will be less affected (e.g. criminal law translation), or establishing new skilled positions, such as post-editing specialist, to support the development of customized solutions for law firms though such methods as post-editing by MT (the implementation of individual workflow blocks). Another strategy, in the short term, would be for those parties immediately affected by recent developments in the legal market to acquire information about law firms that intend to implement legal tech and apply for a position at these organizations, because such firms will certainly hire new legal translators to cover the rising demand resulting from the increasing amount of legal data (Knoblauch, 2019, p. 145).
6. Another relevant impact of legal tech that will change the current definite shape of the legal translation market from the medium-term perspective is the continued rise of super agencies and leaders. This trend will likely persist, since law firms and large companies from other industries will most probably cooperate only with those LSPs that possess a sufficient level of qualification and technical KH and can serve as outsourced project managers capable of covering all necessary languages and translation steps (Knoblauch, 2019, p.145). Moreover, Slator confirms this view and explicitly highlights as the future trend those market players who will benefit most from the aforementioned challenges. First, super-agencies which already invest in new technologies; second, cloud-based technology platforms which offer possibilities to cooperate in a shared format work with direct client; third, new high-tech start-ups which will benefit because they can structure the supply chain independently from existing clients or investments (Faes, 2019, pp. 27-28).

To sum up, the changes in the legal market due to legal tech will have direct but different impacts on the development of legal translation market players, because these two markets are closely connected, with the consequence that the legal translation market will adapt translation technologies to its needs. This adaptation involves the transformation of the market shape and professional profile of legal translators: depending on the specialization and the size of the legal translation business, the necessary change could entail the extinction of the job, on the one side, or increased workload, on the other side. Therefore, different adaptation strategies will need to be implemented by all legal translation market players to ensure that they successfully adapt to these changes in sufficient time.

3. STRATEGIES

3.1. DATA STRATEGY

The big data challenge can be understood as the growing amount of digitalized legal data that should be managed properly to ensure it is (1) received, (2) stored and (3) structured such that the data can be easily accessible for (4) further processing (e.g. by AI). Despite the challenges faced during implementation, the digitalization process has generally made the data more readily explorable. Researchers and practitioners show the manifold spectrum of the application. In this chapter, I will develop a deeper analysis that highlights the current challenges and trends in the German legal market and legal translation market regarding the management of legal data.

The first step is the switch to digital data, overall digitalization, which is the milestone that many players in the market still fail to reach, based on the results of the Legal Revolution conference 2019. By consulting with the justice panel, I collected insider information from serving judges (administrative court, higher regional court, Federal Court of Justice). The first notable trend in the German judicial branch is the obligatory change of communication formats into digital form, obligatory digitalization. E-files (German: e-Akte, digitalized cases) will be mandatory from January 1, 2022 in all public authorities of all branches, with the necessary infrastructure for digitalization being developed as required, according to several national laws in Germany (Doumanidis, 2019, p. 135). In the legislative branch, the laws and legal texts are to be digitalized and made electronically accessible to the public. However, many laws and legal texts are already provided with the professional translation into English and stored on publicly accessible official platforms. During the period 2017-2019, an advantage was gained over other branches due to the rapid digital development of the administrative branch occasioned by the influx of migrants and the acute need to resolve the numerous asylum cases rapidly, which necessitated the digitalization of all such cases.

Nevertheless, this digitalization process is not flawless, and it will take several years to complete. For instance, all lawyers admitted in Germany now have a special electronic attorney mailbox, beA for short, for official communication with the judiciary, with authorities and with each other. However, this mailbox cannot be uninterruptedly operated, and in cases of technical failures, the courts are often obliged to send the documents by fax. It is important to highlight that court (or sworn) translators are involved in these digitalization plans: from July 22, 2019, such translators will have the opportunity to access legal files in preparation for translation, if these documents are available in digital format. Moreover, they can use an electronic court and administrative mailbox or DE-mail to send the translated documents to the courts, although it is currently necessary for translators to first send an inquiry to the court (until December 31, 2021) to determine whether the court can accept an electronic translation. Furthermore, a court translator is now obliged to prepare an electronic signature card and signature device (Doumanidis, 2019, pp. 131-138).

The second step is the storage of legal data. As this step is complex, I have highlighted the major recent challenges below.

The first challenge is devising an appropriate storage system, which will allow not only the storing of digitized legal data, but also a user-friendly workflow, managing the creation, use and storage of documents in multiple formats. Such software is widely used in law firms, such as electronic document management software, and in established IT-companies and other start-ups specializing in developing customized IT products to meet the market needs (Bong & Glock, 2019).

The question is what criteria should be selected for such a storage system. Judging by the offers at a recent conference on legal tech (Bong & Glock, 2019) and in recent books (Breidenbach & Glatz, 2018), the criteria should be defined as follows: First, the digitalized legal data should be stored in a user-friendly format that allows lawyers to re-use it. Second, the format should be designed so that it can serve later as the basis for AI implementation. Third, concerning the multilingual legal environment, the digitalized legal data should be easily retrievable for legal translators.

The second challenge that has often been articulated, both in a conference on legal tech (Bong & Glock, 2019) and in relevant academic works, is ensuring the security of legal data, an essential requirement for such highly sensitive data, including protection against cyber­attacks and raising awareness about data security risks on site. These problems should be solved together while the future digital environment is developed (Hartung et al., 2018, p. 239).

The third challenge is establishing shared community online access to sensitive legal data. This topic in online litigation has been broadly discussed in the book Online Courts and the Future of Justice by Richard Susskind, published in December 2019, and at some recent conferences. The results of the test phase of online litigation in several countries are mentioned in Susskind's work (Susskind, 2019) and can be directly observed in action in China (Netcourt, 2019). The point is to provide online litigation and create greater access to more rights, making the pursuit of justice easier. However, this process requires legal translators to become an essential part of online litigation, enabling participants cooperate on the creation and editing of legal documents online, similarly to the Google Documents platform. What form such cooperation will take, and what challenges and opportunities it will generate for legal translators, will become clearer after the conducting of several test phases in different countries.

The fourth challenge is the cloud-based online access to sensitive legal data, which is legally problematic in most cases. A survey conducted by the Clifford Chance company has confirmed that 37% of companies currently invest in clouds or tend to use them to store legal data. The Clifford Chance company have created innovative platforms for legal services, using the power of collaboration and thereby accelerating the progress of their projects (delivery and applied solutions) (Clifford Chance LLP, 2012).

The third step is the restructuring (or design) work on the stored legal data required for various purposes.

Digitalized legal data are stored mostly to enable later re-use, according to recent trends. The next stage will involve the implementation of AI (components) or automatization and other related processes. This challenge will be discussed separately in the next chapter, together with the MT challenges.

Furthermore, digitally stored legal data can often be used to create additional electronic resources, such as digital databases. First, to further accelerate legal work, legal texts stored in digital databases can be processed with search engines, enabling lawyers to use software to analyse and build legal texts from the building blocks that form such databases. Second, digital databases in the format of legal corpora can also be useful for legal translators using computer- assisted methods to improve translation quality, to accelerate the speed of translation workflow or to realize other goals (e.g. for educational purposes).

Apart from its role in developing term strategy and creating terminology memories for CAT tools, which is discussed separately in the next chapter, stored legal data can also be used to create legal corpora (or digital databases for translation purposes) to enable the systematic and empirical exploration of the patterns of law and language.

Multipurpose corpora In recent years, legal corpus linguistics has made significant advances, stimulated by the development of new technologies aimed at facilitating the growth of computer-supported legal linguistics and computer-assisted legal linguistics (Hamann & Vogel, 2017, pp. 103-104). Computer-supported legal linguistics uses two approaches based on algorithms and software: corpus-based and corpus-driven (Felder, Müller, & Vogel, 2012, pp. 124-125). In the corpus­based approach, the hypothesis is tested based on the corpus. Conversely, in the corpus-driven approach, new hypotheses are developed from the existing corpus (Tognini-Bonelli, 2001). Presently, most corpus linguists use both methods. Computer-assisted legal linguistics utilizes the very recently developed approach of analysing quantitively retrieved data based on algorithms to supplement qualitative reasoning (Vogel, Hamann, & Gauer, 2018).

I will not go into much further detail or list all the recent innovations in corpus linguistics and corpus pragmatics, since these topics lie beyond the scope of this thesis. In this framework, I will only highlight the major trends in and applications of legal text corpora, considered as an essential part of the big data strategy and an integral aspect of the future profile of the legal translator. Therefore, among all the academic studies and expert opinions analysed, I will only consider the resources which are most relevant for the legal market and legal translation market and essential for deeper study of these fields, taking such resources into account where required to formulate business decisions regarding implementing new technologies.

I have studied the results of the first international conference on “The Fabric of law and language: Discovering Patterns through Legal Corpus Linguistics”, which took place in March 2016 in Germany and was convened by Friedemann Vogel and the international research group Computer Assisted Legal Linguistics (CAL2). These results were published in JLL 2017 (Vogel et al., 2017). This conference was selected for the following reasons: first, it was marked out by its rather unusual, but highly fruitful collaboration of linguists and lawyers (Vogel et al., 2017, pp. 94-95); second, it revealed recent challenges facing both the legal market and legal translation market regarding the development of big data and corpora.

I have summarized the outcomes and allocated them in 4 groups as the main advantages and challenges for law firms and legal translators.

First, the corpus design is a debatable issue. After analysing the literature on the corpus for legal translation, it was evident that the first breakthrough is that the situation has radically changed: earlier, there was a scarce number of corpora, especially of those specifically created for legal translation, with only projects and tests available (Vogel et al., 2017, pp. 94-95). However, there is now a growing number of diverse corpora, and nearly every corpus has a different design (reflected in different compositions, objectives and structures), due to the vast number of distinct languages and legal systems (Felder et al., 2012, pp. 414-451, 314-354). The first challenge is determining how to create a meaningful and methodologically balanced design of legal corpora, so that pattern comparisons between different languages and different legal systems can be rendered more universal or unified (Vogel et al., 2017, pp. 94-95).

Though the corpus tools are technically well developed in general, and these tools seem to deal successfully with the growing amount of legal data (at least up to the specific level at which they can achieve sophisticated statistical results), there are some weak points to criticize. For instance, there are difficulties involved in determining the ideal size of the corpus to ensure the results are reliable (Marin, 2017, pp. 18-45), establishing if Google can be used as a universal database for legal interpretation (Mouritsen, 2017, pp. 73-73), or deciding if the ideal future trend would be for all databases or corpora to be universalized or if it would be better to create domain-specific corpora (Vogel et al., 2017, p. 93). Nevertheless, it ought to be noted that this field has only recently emerged, so there are more collaborations, projects and ideas on empirical linguistics and law to come, which should both necessitate and facilitate the development of some sort of meta-theory concerning the creation of such legal corpora (Vogel et al., 2017, p. 93).

In this framework, the main questions will be as follows: what is the purpose of the legal corpus (for translation), and who will create and invest in the development of this corpus for a specific law firm, legal department or (outsourced or internal) legal translator or LSP? In many cases, it will be a top-down decision not only to create legal corpora, but also to invest in the right infrastructure, so that these questions will be answered in the process of developing such corpora. Not only practice can benefit from such corpora-infrastructure, but also researchers in legal translation studies and interdisciplinary studies (Felder et al., 2012, pp. 13-15, 314), the need of which will be highlighted further. A good example of such a strategy is the EUR-Lex collection of documents: its initial purpose was to make legislation public, providing greater access to authentic and translated legal documents, but the eventual outcome was a large number of unintended applications in law and linguistics. Based on this example, a law firm can start with a decision to build a corpus even without some specifically defined purpose in the beginning, starting only with the initial aim to create useful infrastructure for both lawyers and translators (Vogel et al., 2017, pp. 98-99); such infrastructure can change the focus of legal database work from “just retrieving information” to a more linguistically sophisticated level. For instance, by including a KWIC, a keyword-in-context-display with collocations, such as in EUR-Lex (Vogel et al., 2017, p. 100).

This technical benefit may already be a solid reason why it would be more profitable for law firms to cooperate with medium-sized and large LSPs, since these organizations have more knowledge and experience and can (co-)invest in the creation of legal corpora (Faes, 2019, p. 25).

Second, the legal corpus is considered as revolutionary for educational purposes (Breeze, 2017, p. 16). The new computer tools, corpora and ways of processing language information enable the enriching of teaching approaches for legal language, legal translation and law in more sophisticated ways, which is especially important since legal language is a language used for specific purposes. I will not enumerate all the possibilities of new technologies for teaching purposes in this thesis. However, as they are an essential part of the future profile of the legal translator, I will highlight the relevant new technological trends in this framework. On a basic level, using corpora can raise awareness about specific high- frequency terminology and typical formulaic language; on an advanced level, using corpora can further develop knowledge of the structure of typical legal genres (Breeze, 2017, p. 16).

In broader sense, different legal corpora can be used for training on the correct usage of terms in different legal contexts and genres, both for lawyers and for legal translators who are learning a legal foreign language for the purposes of competent communication and/or high- quality translation. After all, a legal corpus can be used for training in methods of understanding or interpreting the legal contexts in which a term occurs. For instance, a seminar on Law and Corpus Linguistics was created in the US at the BYU Law School, with a focus on the applied character of such studies, so that corpus linguistics has become part of the curriculum not only for linguists or translators, but also for lawyers (Mouritsen, 2017, pp. 80-81). Similarly, there are now examples of courses for translators that form part of master's programmes on corpus­based translation studies (Lefer, 2019-2020).

Third, since there is potential for using corpora in courts, there is an acute need to further develop reliable corpora. As Lee and Mouritsen note, ‘The need is acute when the interpretive task involves questions of law' (Lee & Mouritsen, 2017, p. 92). Legal linguist and law professor Larry Solan (2017) shows in his studies the possible further applications of corpora in both legal linguistics and law, analysing recent U.S. court cases in which judges applied a corpus-based approach to legal interpretation when taking decisions. Such use of corpus data is rather novel in the US and other counties, so it can be considered as a shift of the paradigm (Solan, 2017). Though he concludes that corpus analysis will not help to solve all possible interpretive issues, Solan (2017, p. 64) and other linguists argue that it can be a very useful contribution to the objectivity and predictability of decisions, providing empirical grounds for and functioning as a diagnostic contribution to the traditional interpretive approach, while also filling in gaps in the assessment of ordinary meaning, thereby advancing the theory of interpretation and the methodology of corpus linguistics (Lee & Mouritsen, 2017, p. 92).

However, as Mouritsen (2017, pp. 73-74), another prominent legal linguist and law professor, shows in his studies, there are currently several debates regarding, among other issues, what corpora should be used (i.e. general or specific) and what methodological problems the use of Google for legal purposes may present, because such corpus analysis is still a brand-new field of application. Nevertheless, the fact that this approach enables greater reliance on corpus-based results rather than intuition is considered by many legal and linguistic experts as a significant breakthrough, and corpus analysis has been even represented in the media as a ‘revolution' and ‘welcome development' that will ‘proceed on a basis of concrete facts about how we use language, rather than on a welter of idiosyncratic assumptions, as has too often been the case' (Zimmer, 2011).

Moreover, at least at one law faculty now provides essential training in corpus linguistics. Since this course has an applied character, it will enable several potentially critical issues to be addressed by the next generation of lawyers, at least in the U.S. legal system (Mouritsen, 2017, pp. 80-81).

Fourth, there is the challenge of communication and collaboration between linguists and lawyers. As Lee and Mouritsen observe, ‘Moving forward, judges, lawyers, and linguists will need to collaborate to settle on some best practices in this emerging field' (Lee & Mouritsen, 2017, p. 91). More progress can be made in the new field of legal corpora, which would benefit not only legal translation theory and practice, but also legal theory. Moreover, such research could be more meaningful for law firms if there is greater collaboration between linguists and lawyers, and more attention from the legal community (Vogel et al., 2017, p. 94­96). Although there has been some educational impact (e.g. the courses in the US mentioned above), this outcome alone is insufficient, given the existing technological possibilities within the field. The researchers urge expanding collaboration and establishing ,an international network for research on computer-assisted legal linguistics' (Hamann & Vogel, 2017, pp. 101­102).

Finally, such collaboration should involve researchers in the legal translation studies field. As aligned parallel corpora, multilingual corpora are an integral part of the internal or outsourced translation workflow if based on the terms specific to the law firm, so that after creating a high-quality translation, both the ST and TT can be aligned and used for accelerating further translation projects using an MT system. However, technological weak spots can still be revealed as potential growth points, such as in automatic term recognition and analysis. In this framework, corpus linguist Maria José Marin (2017) compares in his studies from 2012 to 2017 various existing computer algorithms that are used for automatic term recognition based on the corpus-driven application, exploring the technological state-of-the-art in this field (Marin & Camino, 2012).

In his recent December 2019 article, Rosario Martin Ruano (2019) critically assesses the recent findings of empirical data-driven approaches in the field of legal translation. Ruano (2019) notes that although these findings are often claimed to assist legal translators in making better and more informed decisions, they should be evaluated more critically, considering their failings. He analyses the recent development vectors of legal translation studies, highlighting two present and future trends that were stated by Biel and Engberg (2013): first, the predominance of corpus-based research methods in legal translation studies; second, the shift from qualitative to quantitative methods and from prescriptive to descriptive approaches. However, Ruano concludes that while such empirical findings using corpora can be considered as a contribution, ‘as in the case of historical-descriptive translation studies, the aim of large­scale computer-assisted, empirical studies grounded on the exploration and analysis of big data was to discover the properties and regular patterns of translation behaviour as derived from real texts' (Ruano, 2019, p. 134) for practical applications (e.g. in training contexts). He observes that the limitations of the research methods should also be considered and studied. Accordingly, Ruano both shares and advocates for the opinion that was spotlighted as the fourth point in this chapter, that this new innovative research methodology ‘might cross boundaries across research approaches and disciplines in order to shed light on the many facets of this social practice' going beyond the limitations of current legal translation practices (Ruano, 2019, p. 134). Such limitations can be considered as indicating the need for legal translators to further enhance their own research procedures, while also extrapolating research methods that have been validated in other fields to the legal translation studies context. For instance, while corpus­based linguistic analysis is used in other fields, it is still not widely applicable in legal translation studies, although it is now considered a necessity, and its results have been empirically confirmed by recent researchers, such as Felder and Vogel (Felder et al., 2012, pp. 346-347), as noted above.

To sum up, data strategy serves as an adaptation to the increasing amount of digital legal data, and it includes different steps of data managing: (1) digitalization as receiving, (2) storage and (3) structure for (4) processing (e.g. by AI). In both the German legal market and legal translation market, this strategy is developing in a very similar direction. Each major step of the process currently involves different challenges for both markets. For instance, the storage process is affected by problematic legal issues such as the need to preserve the security of legal data in the context of shared community online access and cloud-based online access.

An optimal data strategy is to create corpora for legal services and/or legal translation services, which have various applications. The integration of the corpora into the processes of legal practice and legal translation seems to be inevitable, which means the KH and new competences required to create and manage these corpora will become an essential part of the future profile of the legal translator. Moreover, researchers have highlighted the growing emergence of the implementation of the corpus-assisted legal linguistic approach in the legal market and have urged for more intense collaboration between lawyers, legal linguists and legal translators (Mouritsen, 2017, pp. 80-81). This trend is evidenced by number of debates in the media, such as Harvard Law Review and Yale Law Journal Forum, concerning recent revolutionary cases in U.S. courts in which the presiding judges used corpora to interpret the law (Mouritsen, 2017, p. 82). Despite criticisms of this field, judges and lawyers still have to interpret the law, so machines will not replace humans: on the contrary, practitioners will have more tools in the format of the legal corpora, while the results of interpretive practice will become more predictable and objective, allowing society to benefit.

3.2. ARTIFICIAL INTELLIGENCE (AI) AND MACHINE TRANSLATION (MT) STRATEGY

The topic of the implementation of AI is generating great excitement in both legal society and the legal translation field.

Therefore, in this chapter, I will examine this major development trend further. I will first briefly present the AI-related challenges and opportunities for the current legal market. I then continue with an assessment of the impact of AI-driven changes on the legal translation field, comparing the latest information from conferences, academic works and experts' opinions.

This topic is highly relevant on a short-term basis, given the boom in investments in AI (in all business sectors) (Perrault et al., 2019) and the growing number of practical cases of implementing of AI in the market. For instance, CMS, Freshfield are creating new support and legal tech centres, for which new translators are being recruited to translate the legal data, and there are many similar cases, some of which were presented at the 2019 Legal Revolution conference in Frankfurt. However, according to the experts' opinions during panel discussions and presentations, the speed of implementing of AI can vary (e.g. it is a long-term prospect for the legislative branch, the executive and the judiciary branch), since successful implementation is dependent on many factors, primarily the availability of digitalized data. I held a discussion with Olga Prikhodko (2019), the leading expert in the successful legal AI project pravoved.ru being run by Skolkovo, the Russian innovation centre, and she confirmed this opinion, based on the results of this project, which were presented at the Legal Evolution conference.

I took a step further by participating in a Coursera course to learn the basics of AI from the leading AI experts from the Google Brain team and the Baidu AI Group and to comprehend how to transfer the strategies behind AI development into legal and legal translation businesses. Within the framework of this study, the same conclusions were found in this course, providing the insight that it is possible for any business to become a well-established AI-based company, or at least to implement AI wisely, by following recommendations of these experts outlined in their Playbook (Ng, 2018).

Based on these Playbook recommendations, I will present the simplified general phased implementation strategy of AI. However, given that this Playbook was created mainly for larger companies, it is important to remember that every company should develop its own AI strategy, using the following steps, while keeping in mind its long-term basis and business value (Ng, 2018). The first phase is to execute pilot projects.

The first step is to understand whether the AI is the right tool for the company and its processes and will create business value. In this sense, the AI implementation is not the goal in and of itself, and the idea (and the general hype suggesting) that AI can or should do everything is not realistic. Such AI tools provide technical solutions for specific business problems, such as accelerating repetitive processes. Therefore, clearly defined objectives and technical due diligence are required. For instance, it might be asked whether the legal data is repetitive enough or exhibits such patterns, and whether the law firms of any other authority have enough legal data, so that starting an AI project is a reasonable decision here.

The second step is to determine the right organization-specific approach by initiating the measures the company can already take, executing small pilot projects first, rather than starting with the attempt to build a strategy. At the conference Legal Revolution 2019, AI experts presented several cases of purposeless purchases of large amounts of legal data from third party companies, without prior testing of the company's own legal data in small pilot projects (Prikhodko, 2019).

The second phase is to create the right in-house team and provide AI training so that other team members are included and can cross-functionally support pilot projects.

The third phase is to develop the company's own strategy. In this phase, there are many challenges. First, the data should be labelled and structured properly. These mechanical tasks are mostly delegated to paralegals, based on the information given at the Legal Evolution conference. However, such tasks can serve as starting points for additional professional activity for the in-house legal translators, against the background of the further implementation of MT elements. In my view, therefore, international projects should include a translation strategy and an expert with translation skills.

The fourth phase is to develop internal and external communication methods, which includes a reasonable HR strategy, to not only attract the necessary talents but also retain them, offering additional educational resources if needed. Finally, leading AI experts highlight the need to maintain clear internal communication, focusing on clarifying the concerns employees may have about being replaced by robots and machines.

As the final phase, before the company seeks to scale intelligent automation technologies, based on the presentation by KMPG drawing on the 3Q18 Global Pulse survey report and their further corporate resources, important key performance indicators (KPIs) should be included as follow-up on the changes (e.g. treating cost effectiveness and risk sensitiveness as the most important KPIs). Moreover, KMPG provides a target operating model for companies' strategic planning (KPMG, 2019, pp. 2-5). This model can help address business challenges, while serving as a basis for building a solid strategy (KPMG, 2019, p.5).

Consider, for instance, the practical case of Engelhard Arzneimittel: the legal department of this company invested one year in creating the concept of the AI, taking an additional year to digitalize all data with OSR software and to structure it. For 2020, they are planning the test phase to reveal and then correct any potential weak points. The company's initial strategy was to adjust existing software to facilitate the building of contracts from pre-formulated and legally safe blocks, and to enable the system to send reminders of important legal deadlines. Based on this case, it can be assumed that to implement the AI for international contracts and the projects that aim at building texts from pre-formulated and pre-translated blocks, another year will be needed to collect, translate and correct all blocks required for the further use of the MT tools (Schumann, 2019).

The EY Panel results from the Legal Evolution conference confirm these findings (Schulz, Sohn, & Horn, 2019). All the steps outlined imply investments on a long-term basis and a solid company-customized strategy. However, the recommendation of the experts from this panel is to use the opportunity to accelerate the transformation of the field and to be a part of such projects as experts, because there are currently very few AI experts working in the area of legal translation, and even fewer such experts with knowledge of MT and translation skills.

To answer the question of what impact AI has on MT currently, I will briefly resume my account of the history of the development of MT methods.

In the last three years, the quality of MT has achieved a breakthrough, and like AI, neural machine translation (NMT) has generated great excitement. However, when recalling the major development milestones of MT, which are depicted in Figure 1 below, the beginning of each milestone was identified as an exciting breakthrough. Earlier MT applied a rule-based approach, in which the computer scanned the sentences of the ST, then split them into abstract pieces, a kind of computer interlanguage, and then built the TT, using these pieces according to the rules of the TL. While this approach is considered as a fundamental milestone of the current state-of-the-art, researchers have detected that the involvement of only these rules and structure is insufficient for accurate translation, so other elements must also be included (Gaus, 2019, pp. 168-170).

Therefore, the next approach developed switched off the use of large numbers of other elements in the format of corpora and employed statistical methods. This method, statistical MT, showed success starting with the version developed in 2007, mainly because of technical breakthroughs regarding three criteria: powerful processors, low prices for data storage, and large amounts of available data to build corpora (Gaus, 2019, p. 170). Though this approach was more successful, its performance was still not consistently good enough.

The latest approach, NMT, appeared in 2014, starting with attempts to train machines to obtain higher and more consistent performance. The idea behind NMT was to utilize digital (artificial) neural networks as a sort of artificial memory, using the deep learning method in MT. In this approach, the computers are still dependent on large corpora (both ST and TT) and statistical methods, to ensure that the machines will have enough data to decode the ST and to predict the words for the TT, while also learning from this process. By 2016, this approach was applied in all the most advanced MT systems globally and utilized for translation services in the largest IT companies, such as Google (Bojar et al.).

Abbildung in dieser Leseprobe nicht enthalten

Figure 1. A brief history of machine translation (MT) (SYSTRAN, 2020)

However, considering the state-of-the-art of translation services using the NMT systems, several challenges remain.

First, not all language pairs and translation fields show the same progress, as revealed in recent studies, such as those comparing human and machine translations. For instance, a recent study of Arabic and German by Daniel Falk (2019), indicated the challenges involved in the translation of long Arabic sentences, taking into account varying dialects and intercultural differences. In another recent study, Gaus (2019) compares the quality of human translation (HT) and NMT translated legal texts, with a focus on the GDPR texts, finding inconsistent terminology and mistranslations in the MT-generated versions. However, Gaus notes that MT can be used more efficiently for translating repetitive and formulaic legal texts (2019, p.175­176). A further experiment was conducted by Peter Schmitt (2019) from November 2018 to July 2019, where different technical texts were translated with DeepL Pro and Google Translate and then compared. He concludes that MT can achieve a remarkable level of quality, comparable with human translation, but indicates that the comparability of the quality depends on the purpose and complexity of the texts involved, as well as on the expertise of translator, in the sense that even human professional translators can make errors and depend on context to determine meaning. Moreover, human professional translators do not always have maximum text-specific competence (i.e. perfect textual understanding), so in this case MT may provide better translations, competing with the quality achieved by human translators (Schmitt, 2019, pp. 184-188).

Given that machine learning reveals its limitations only by decoding and understanding language, critical scholars and practitioners argue that any translation of creative content or content that involves creative or interpretive mental work and liability should involve a human translator (Gaus, 2019, pp. 175-176). For instance, as confirmed by the recent case of MotionPoint's Global Growth team, ‘grievously mistranslated content becomes a serious liability and credibility problem for a brand' (Hutchins, 2017). Moreover, additional liabilities can result from inaccurate machine translations, as when Google's search algorithms penalize mistranslated MT-generated web content (Hutchins, 2017). In this context, the importance of having the necessary expertise for understanding this problem can be highlighted. In the case of legal texts, the question of liability for errors is even more sensitive. In Mustu's study, the comparative quality of TT produced by MT and HT with the aid of the DeepL system was assessed. This research considered DeepL as an assisting tool for a legal translator but not vice versa, because there are terminological, stylistic, structural and language pair challenges which the machine cannot resolve alone (Mustu, 2019).

Second, despite some very promising initial progress, NMT still experiences the ‘early stages of maturity', according to Alon Lavie, a Machine Translation Group Manager for Amazon Pittsburgh, who adds that ‘ MT is still far from being a solved problem' (Marking, 2017).

In this framework, therefore, hybrid approaches to MT should be not be neglected. While this approach appeared even earlier than NMT, combining traditional statistical and rule-based methods (Costa-jussa et al., 2016), the recent developments in machine learning have also been incorporated into this approach by researchers (Hunsicker, Yu, & Federmann, 2012). The hybrid architecture of MT considers challenges and combines the advantages of both statistical and rule-based approaches, enabling a more accurate translation to be created (Yeong, Tan, Gan, & Mohammad, 2018).

In the context of hybrid approaches, another method should be highlighted that is usually not included in any graphs of the development of MT, although, in my opinion, it now should be. This approach is sometimes called the hybrid human-machine translation approach and includes human participation in several phases of workflow translation, such as human post­editing (Gutiérrez-Artacho, Olvera-Lobo, & Rivera-Trigueros, 2018). In general, one of the recent future trends may be more sophisticated hybridization of human and machine translation workflow, which combines the advantages of all approaches to increase productivity and quality, utilizing human translation as an essential element (Gutiérrez-Artacho et al., 2018).

While this integration of MT and HT will not replace human translators, it will change the whole translation process (workflow) and the profile of the translator.

Therefore, it is recommended to realistically evaluate both the pros and cons of the implementation of MT in each case (Porsiel, 2017). Put another way, the integration of MT and AI into the regular workflow should be individually adjusted and tailored to the specific company's needs and based on objective results and recent resources (e.g. guidebooks) (Porsiel, 2017). This approach should not be considered generally as an all-purpose medium or, conversely, as an immature decision or overly complex investment.

Within the framework of legal translation, it is essential to assess risks and liability issues connected with the implementation of AI and MT, particularly in relation to the format or location in which the sensitive legal data are stored (e.g. cloud-based or transferred in different phases to and from legal translators and/or machines, servers or the cloud).

What are the future trends and specific future impact of AI and MT on the profile of the legal translator and the legal translation field?

First, automatization is causing radical transformation in the job market, displacing 400 million people, but also creating new jobs. According to the results of McKinsey's study (2017), by 2030, nearly half of all professional activities in the world will be automated and one third will be displaced by new technologies, with varying proportions across researched countries (e.g. 24% in Germany) (McKinsey Global Institute, 2017, p. 8). For 75-375 million people, this transformation will entail either switching occupations or upgrading skills (or both) (McKinsey Global Institute, 2017, p. 8).

Second, regardless of the size of the legal translation businesses affected, automatization processes, AI and MT will change the legal translation business (Hoppe, 2019; Zielinski, 2019): prices are falling, while the requirements for productivity and quality are rising. The roles of simple language mediators and translators of medium quality will have to be replaced by MT, since they will not meet the new requirements. However, translators with expertise in the legal field will be in demand (Faes, 2019, pp. 27-28).

Third, experts seem not to have a unanimous opinion on the direction of developments in the field, although translation and MT are still considered as hard problems for AI. Therefore, specialists in the legal translation field will need to be involved in the implementation processes of AI and MT in the legal businesses (Faes, 2019, pp. 27-28).

To sum up, AI and MT can be highlighted as the second major trend.

AI can and should be implemented within every business to ensure it is up to date with the market and responsive to future trends, and there are proven strategies to implement AI in economically reasonable ways, following recommendations by leading experts and the models provided by successful cases. Nevertheless, AI is not yet technologically mature and remains a long-term investment accompanied by challenges and risks. Therefore, an optimal AI strategy should be customized to the needs of specific companies and include a reasonable HR strategy.

The same business and technical due diligence should be conducted to formulate optimal MT strategies. Although such a strategy is dependent on the development of AI, it will require a hybridity of forms of cooperation between human legal translators and AI to combine the advantages of both approaches. An important finding here is that the architecture of AI and MT workflow should be dynamically accessible to error corrections by human specialists to ensure effective optimization.

A key future trend that can be predicted is that classical human legal translation produced by human translators will eventually disappear in most areas, at least in the long term; however, new competences, such as AI skills, will be required as the field adjusts to these changes.

3.3. TERMS STRATEGY

The final strategy is term strategy. As was discussed at the beginning of this thesis, choice of terminology and the notion of equivalence are probably the most critical challenges in legal translation. Moreover, as explained in the preceding chapters on AI and MT challenges, these are still the areas with the most critical errors in need of addressing, despite the breakthroughs in the development of AI and MT. It should now be clear that term strategy is interconnected with the two previously outlined strategies.

I have differentiated several challenges which are of great practical importance for further research, according to the phases of the workflow of legal translation: extraction of terms from corpora and tools for term mining; creation of systems for storing, managing and retrieving of terms; automatic quality assessment methods of terms; integration with CAT tools and other tools and systems.

The first challenge is the terminology extraction process, or terms mining, which is considered as time consuming, but of high importance, especially in the legal translation field. In general, there are different online tools and software on the market. However, in the study by Micha (2018, pp. 13-17), the best practices are considered to be hybrid approaches to term mining that use both statistical and linguistic methods, combining the advantages of each. Statistical methods are more suitable for big data, while linguistic methods are better suited for language-specific features. Terms mining is an intermediate step in a workflow; the step that follows is evaluating the extracted terms based on threshold values and storing the results in a terminology database. As Micha (2018, p.17) points out, there are no established threshold values in this process, as such values should instead be determined by the specific company in a set of terminology guidelines, which should be dynamically accessible. Moreover, since his study aims at helping companies to take informed decisions, a further important decision criterion is the possibility of integration with external information sources, corpora and other digital (and online) tools. Based on the promising results of these studies, there is space left for decision-making processes conducted by the legal company and legal translator, such as determining what automatic term recognition tools are more efficient and what KPIs should be used to test these tools (Marin, 2017, pp. 18-45). Furthermore, it is important to highlight that the terminology extraction process is not a fully automated process and requires legal expertise from a legal translator (or a specialist in charge of this task).

The second challenge is the development of the corpus tools. Technically well-developed corpus tools allow sophisticated statistical results to be achieved, although there are gaps to be filled through further research, such as determining what the ideal size of the corpus should be to ensure the results for the legal term mining process are reliable (Marin, 2017, p. 20).

The third challenge is the implementation of an optimal terminology database, glossary and dictionary in the workflow. There are several solutions currently on the market, such as the products offered by UniLex, which are considered as industry leaders, are suitable for legal translation and are compatible with the most popular CAT tools, such as SDL Studio, memoQ and external resources (Acolada GmbH).

The fourth challenge is the reasonable integration of all steps in the workflow of legal translation to provide terminological consistency, which should be accessible and dynamically regulatable (Faes, 2019, p. 28) to improve quality and create suitable architecture for better deep learning processes. This challenge has a direct impact on the workload, a factor that should be included in the calculation of billable hours, which will alter the ways legal translators are remunerated. Depending on the company's needs and the legal company or LSP, these pre-editing and post-editing activities, term management and other tasks could be handled by different specialists: a trained paralegal, lawyer, legal translator or other trained specialist, always remunerated accordingly.

Though these time and cost efforts are necessary to produce a sufficiently high translation quality, they must not be excessive and should clearly meet the specified requirements that guide the operations of the AI and MT. Thus, as explained above, ROI should be considered and an economically reasonable strategy for the implementation of new workflow phases should be developed (Heard, 2017).

To sum up, current KH in the market can be considered primarily as useful tools which can accelerate the legal translation workflow, though they have essential limitations (e.g. KPIs must be estimated by the specific company). Moreover, to achieve high quality consistently is even more challenging for these tools, given the legal liability of legal translation and its high complexity: even if only partial equivalence can be achieved, it requires the legal expertise of a legal translator in a specific legal field. New tools for terms mining are appearing, and the development of corpus tools and other electronic and online tools is often based on new AI, MT or corpus technologies. However, such tools still require human assistance to be effective, so the terms strategy should include not only technical due diligence and ROI calculations, but also a reasonable HR strategy, which is an essential component.

4. CONCLUSION

Both the legal market and legal translation market are experiencing digital transformation and having an impact on each other. The results from recent conferences and studies prove that different market players have started to effectively communicate and harmoniously integrate new technologies as a part of their corporate strategies. However, this development implies the transformation of the entire business model for delivery services, which will involve outcome­based strategies and empowering human resources as enablers of technology. According to the principles of the reverse engineering method, to be able to predict future trends, two points in the timeline are required: the current situation, defined as point A, and the possible outcome in the future, defined as point B. Only from this perspective can a suitable strategy be formulated.

Therefore, in the second part of the thesis, I focused on the current challenges of new technologies, based on recent findings regarding the current market for legal services and legal translation services, as well as the legal translation studies field. I have presented existing major future trends that will impact the further development of the legal translation field. In the third chapter, I examined these trends, based on different expert opinions and the results of relevant studies, books and conferences, with an emphasis on using the newest possible resources (up to 2020), including less readily accessible resources, such as conversations with experts from market-leading companies.

Furthermore, I differentiated key development vectors for the future:

- technical due diligence, which includes technical assessment of workflow and technical solutions, data and cloud strategy, terms and TM (CAT tools) strategy and automatization strategy (NMT strategy and AI strategy);
- business due diligence, which includes HR due diligence as an important element of the whole digital transformation strategy, including translators with significant expertise in TM, AI, MT and law (given the requirements of this thesis, I have focused on only the most relevant strategies for the legal translation market); and finally,
- legal due diligence, which includes such issues as data protection, the accessibility of data storage, on-time digitalization, and others, which I have addressed briefly. I will briefly sum up these key findings of the thesis below.

Development of legal translation studies Legal translation studies have changed significantly, experiencing exponential growth in recent years. These developments are fully justified, given the impending new challenges, such as AI implementation. New multidimensional research perspectives have had a practical impact on the legal translation field, adding value for the development of the law, legal practices and the market. Nevertheless, many authors have called for the mutual empowerment of the theory and practice of legal translation and for the further enhancement of the professional profiles of legal translators in general.

Among the significant results of legal translation studies, recent descriptive research can be emphasized. Such research can contribute to legal translator training, facilitating both entry into the profession and professional development. Since best practices, models and procedures are necessary to develop strategic competence, but are not always easy transferable methods, the competence to effectively apply different strategies to solve communication problems (e.g. translation) is considered a core skill for specialized translation.

Development of legal market and development of legal translation market The legal market and legal translation market are interconnected. As the current phase of Legal Tech 3.0, AI is and will remain the trend for the coming years, although not equally for all market players (at least for German markets). This trend is probably most relevant from the medium- or long-term perspective, since legal tech is not yet homogeneously implemented in the market. Therefore, many market players will follow all three phases of legal tech, including digitalization. Moreover, it is of high importance that national laws be designed to accelerate the development of legal tech (e.g. in courts). The legal market is not only directly experiencing the implementation of legal tech and its impact, but also facing new legal issues connected to this process, such as data protection and legal security for cloud-based solutions, requiring law firms to help shape individually tailored legal solutions based on their own experience.

The future trends that will provide solutions to the current challenges are economically reasonable optimization of the workflow, big data, networking and blockchain. These trends will most probably have different impacts and results in changing workflow, business models and services supply chains, but will not homogeneously affect all market players.

These trends in the legal market will result in trends of adaptation in the legal translation market:

- From a short-term perspective, an increasing demand for legal translation will result.
- From a middle-term perspective, the trend will be adaptation of translation technologies to legal tech changes to provide multilingualism, localization and internalization for the new and updated IT-based legal products that will be developed on the legal market.
- From a long-term perspective, adaptation will include changes in the market shape and professional profile: classical legal translation produced solely by humans will disappear in several niches. Large LSPs with more qualification and technical KH will benefit from these changes to workflow, business models and services supply chains. Human legal translators will adapt their competences, either adding technological competences or new specializations.
- As a negative scenario, organizations that fail to adapt to these challenges by not taking into consideration trends as the vector for development will find it very challenging to survive, since even the conservative public sector will be mandatorily digitalized from 2022.

Data strategy The rapidly increasing amount of digital legal data is considered as a challenge that can be addressed though an optimal data strategy, which includes different steps of data managing: (1) digitalization as receiving, (2) storage and (3) structure for (4) processing (e.g. by AI). In general, in both the German legal market and legal translation market, this strategy is presently developing in a very similar direction, but each major step currently poses different challenges for both markets. For instance, in the storage process there are problematic legal issues such as the security of legal data in shared community online access and cloud-based online access.

An optimal data strategy aims at AI implementation as the long-term goal; therefore, it should consider creating user-friendly professional digital databases for legal services, such as smart contracts and/or corpora for legal translation services. There are attempts to resolve legal questions with the legal corpora, at least in the US, although further development will require, first, more intense collaboration between legal linguists and lawyers, and second, in the case of international proceedings and the further development of the online courts mentioned above, more intense collaboration with legal translators.

In general, researchers and practitioners show the manifold spectrum of the application of digitalized legal data using new technologies. These future trends should be considered as part of creating the digital environment, infrastructure for digital and online courts and digital online legal community in which more collaborative legal communication and work will be technically possible and legally secure. The trend for the medium-term future will involve the transfer of legal data to the cloud, the resolution of cyber-security issues, or shared community online access and cloud-based online access to sensitive legal data to create international online courts. These trends will ensure sufficient involvement from legal translators in both the short- and the medium-term future.

Furthermore, since the implementation of all these steps implies both technical and legal competence, as well as basic translation skills, legal translators or lawyers with the appropriate expertise will need to be trained or recruited. Despite the pessimistic perspectives of several niches in the legal market and legal translation market, and despite criticism of new AI-based approaches, there are gaps to be filled by legal experts, since brand-new fields are being created (e.g. through the use of legal corpora). This trend is evidenced by the inclusion of corpus-based studies as a brand-new part of the curriculum of BYU Law School and as part of recent master's programs in translation.

MT and AI strategy Both AI and MT are over-hyped but fundamental components of competitive advantage and the principal business challenge for the future. Thus, the implementation of MT and AI in legal translation is no longer a question of “if’, but rather of “how” and “when”: every company should build its own AI strategy. An optimal implementation in the workflow is only beneficial from the long-term perspective, as successful practical cases confirm. Moreover, such strategy considers the needs of a specific business: investments in AI aim at feasible business value for the specific processes of individual organizations during business due diligence.

However, the first challenge is to be more realistic about what AI and MT can achieve and what they cannot: AI and MT have performance limitations, which can be revealed during technical due diligence. Moreover, implementing AI and MT should involve all departments, so that changes in the workflow can be customized to the needs of each specific company and optimized to be more efficient.

One possible future trend may be the hybridization of the human and machine translation workflow to combine the advantages of both methods and to increase overall productivity and quality, with human translation remaining an essential element of the process. Such integration will not replace human translators; on the contrary, it will change the whole translation process (workflow) and the profile of the translator. Both the AI and the MT strategy should include a reasonable HR strategy: hiring or training new specialists with the required legal, translation, and AI and MT expertise. HR strategy is becoming an increasingly important part of the whole digital transformation strategy, while life-long learning remains an essential element.

Terms strategy The choice of terminology is essential to achieve legal translations of high quality, though partial equivalence can be achieved at best, even when using sophisticated MT and AI. Due to the legal liability of legal translation, developing the terms strategy is very challenging, and it therefore requires expertise in a legal filed and translation skills as the necessary prerequisites to enter the profession.

There is new KH in the market, which can accelerate the legal translation workflow: new tools for terms mining and newly developed corpus tools, along with other electronic and online tools. However, for most of these tools, relevant KPIs and an ideal TM or corpus have not yet been established, so each company should customize the tools they utilize to facilitate efficient integration, economically reasonable implementation and consistency of terms, while making terms accessible and dynamically regulatable to enable the AI to learn more effectively.

I have gathered the main conclusions into theses, which are equally important for legal translators and law firms, but especially for legal market and legal translation market players.

1. The main conclusion of the thesis is that KH is only a development trend, similar to point B, toward which all progress in the industry is currently tending. Although the future cannot be predicted exactly, it is clear that the future of law and legal translation will at least include KH. Therefore, development can and should be actively and proactively influenced by driving both law and legal translation in the direction of the growing business value of all markets.
2. KH is a long-term development. The second conclusion is that KH is not a goal in and of itself, and it is first necessary to determine the economic goal for KH to be implemented and the business value that can be created; it is impossible to implement every aspect of a strategy at once, and there is no sound economic reasoning behind such an approach.
3. Development strategies can vary significantly and may result in a radical transformation. An optimal business strategy should consider all the trends discussed above as the basis for the optimal implementation of AI, automatization, MT and higher terminological quality. Moreover, an optimal business strategy should critically assess how the company will optimize the future workflow of legal services and legal translation services to create business value. Each company and translator must identify their own strengths, determine their economic benefits and ROI, and formulate their strategies, but on a long-term basis, since the introduction and implementation of new technologies usually takes several years. New technology should accelerate the development and progress of a legal (translation) company and a legal translator, but not waste their resources. For some niches, a development strategy can result in a radical transformation, such as a legal translator acquiring additional competences.
4. The introduction of new technologies is not only an outcome, but also a process; this development is currently occurring at all levels of society and will most probably radically change all processes in the markets. In this context, an interdisciplinary approach is relevant: the intersection of translation, jurisprudence, language and technology, combined with a closer incorporation of the translator into the company's processes, is recommended.
5. Shifting the paradigm in a new world entails the gradual implementation of a new holistic approach in both the translation industry and the legal translation industry. This approach will require perception of the processes as parts of the whole unit and interconnection of the parts of these processes (e.g. parts of the workflow in the translation process or project).
6. The role of a legal translator should in many respects be proactive: it is probable that not all legal companies yet realize or know why it is essential to better incorporate legal translators into their organizational structures incorporate legal translators or how the processes of legal translation can be more efficiently improved.
7. New development trends and KH are neither inherently evil nor good; this is an accomplished fact, and scenarios for further development can also be chosen by market players, based not on emotions and hype, but on a solid analysis of the market and the company's own needs, strengths and purposes (point A), so that an effective strategy (point B) can be built, as a result of studying the relevant trends and growth vectors for a future profile.

BIBLIOGRAPHY

Acolada GmbH. UniLex™ - Electronic Dictionaries. Retrieved from https://acolada.de/en/products/unilex-electronic-dictionaries/

Albi, A. B., & Ramos, F. P. (2015). Legal translation in context: Professional issues and prospects (2 rev. print). New Trends in Translation Studies: vol. 4. Oxford: Peter Lang.

Bauer, W., & Mayer, F. (Eds.). (2019). Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ): BDÜ Fachverlag.

Biel, L., & Engberg, J. (2013). Research models and methods in legal translation. Linguistica Antverpiensia, New Series, 12, 1,8. Retrieved from https://lans- tts.uantwerpen.be/index.php/LANS-TTS/article/view/316/220

Bojar, O., Chatterjee, R., Federmann, C., Graham, Y., Haddow, B., Huck, M., Zampieri, M. Findings of the 2016 Conference on Machine Translation. In O. Bojar, C. Buck, R. Chatterjee, C. Federmann, L. Guillou, B. Haddow, M. Turchi (Eds.), Proceedings of the First Conference on Machine Translation: Volume 2, Shared Task Papers (pp. 131-198). Stroudsburg, PA, USA: Association for Computational Linguistics. https://doi.org/10.18653/v1/W16-2301

Bong, A., & Glock, P. (2019, December). Legal@2022: KPMG Law, Legal Process & Technology. LEGAL ©EVOLUTION 2019. LEGAL ©EVOLUTION 2019, Frankfurt am Main, Germany. Retrieved from https://www.legal- revolution.com/de/programm/vortraege-und-panels

Breeze, R. (2017). Corpora and Computation in Teaching Law and Language. International Journal of Language & Law, 6, 1-17. https://doi.org/10.14762/jll.2017.001

Breidenbach, S., & Glatz, F. (Eds.). (2018). Rechtshandbuch Legal Tech (1. Auflage). München: C.H. Beck.

Breidenbach, S., & Glatz, F. (Eds.). (2019). Rechtshandbuch Legal Tech (2. Auflage). München: C.H.Beck.

Clifford Chance LLP (May 2012). Cross-border M&A: Perspectives on a changing world. Retrieved from https://eiuperspectives.economist.com/sites/default/files/Cross- border%20M&A.pdf

Costa-jussa, M. R., Rapp, R., Lambert, P., Eberle, K., Banchs, R. E., & Babych, B. (2016). Hybrid approaches to machine translation. Theory and applications of natural language processing, 2192-0338. Switzerland: Springer.

Czopik, J. (2019). Die Zukunft ist jetzt. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 30-40). BDÜ Fachverlag.

Doumanidis, E. (2019). Die Justiz wird digital -für alle? Der elektronische Rechtsverkehr für Übersetzerinnen und Dolmetscherinnen. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 131-138). BDÜ Fachverlag.

Faes, F. (2019). Die Übersetzungsbranche im Umbruch: Wie Wettbewerbsdruck und KI den Markt aufmischen. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 24-28). BDÜ Fachverlag.

Falk, D. (2019). Möglichkeiten und Grenzen der neuronalen maschinellen Übersetzung am Beispiel des Sprachenpaars Arabisch-Deutsch. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 204-206). BDÜ Fachverlag.

Felder, E., Müller, M., & Vogel, F. (2012). Korpuspragmatik: Thematische Korpora als Basis diskurslinguistischer Analysen / herausgegeben von Ekkehard Felder, Marcus Müller, Friedemann Vogel. Linguistik, Impulse & Tendenzen, 1612-8702: Vol. 44. Berlin, Boston: De Gruyter.

Gaus, L. L. (2019). Close but no Cigar: Adventures in Machine Translation. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 168-176). BDÜ Fachverlag.

Gozdz-Roszkowski, S. (2011). Patterns of linguistic variation in American legal English: A corpus-based study. Lodz studies in language: Vol. 22. Frankfurt am Main, New York: Peter Lang.

Groot, G.-R. de (1991). Recht, Rechtssprache und Rechtssystem - Betrachtungen über die Problematik der Übersetzung juristischer Texte. Terminologie & Traduction, 3, pp. 279­316.

Gutiérrez-Artacho, J., Olvera-Lobo, M.-D., & Rivera-Trigueros, I. (2018). Human Post­editing in Hybrid Machine Translation Systems: Automatic and Manual Analysis and Evaluation. In A. Rocha, H. Adeli, L. P. Reis, & S. Costanzo (Eds.), Advances in Intelligent Systems and Computing. Trends and Advances in Information Systems and Technologies (Vol. 745, pp. 254-263). Cham: Springer International Publishing. https://doi.org/10.1007/978-3-319-77703-0_26

Hamann, H., & Vogel, F. (2017). The Fabric of Language and Law. Towards an International Research Network for Computer Assisted Legal Linguistics (CAL2). JLL. (6), 101-109. https://doi.org/10.14762/jll.2017.101

Hartung, M., Bues, M.-M., & Halbleib, G. (Eds.). (2018). Legal Tech: Die Digitalisierung des Rechtsmarkts. München: C.H. Beck; Vahlen.

Heard, R. (2017). FIT position paper on machine translation. Babel, 63(1), 130-135. https://doi.org/10.1075/babel.63.1.09hea

Hoppe, B. M. (2019). Proaktive Qualitätssicherung beim Übersetzen.. .auch am Ende der Nahrungskette im Dokumentationsprozess. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 229-239). BDÜ Fachverlag.

Hunsicker, S., Yu, C., & Federmann, C. (Eds.) (2012). Machine Learning for Hybrid Machine Translation. Montréal. Retrieved from https://www.aclweb.org/anthology/W12- 3138.pdf

Hutchins, C. (2017, February 8). Can Neural Machine Translation Compete with Human Translation? (Part 1). Retrieved from https://www.motionpoint.com/blog/can-neural- machine-translation-compete-with-human-translation-part-one/

Knoblauch, H. (2019). "Legal Tech" - wie die Digitalisierung die Rechtsbranche und die Rahmenbedingungen für Rechtsübersetzer verändert. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 139-149). BDÜ Fachverlag.

KPMG (01/2019). 3Q 2018 KPMG Global Insights Pulse Survey Report 13: Realign your business model for automation and talent challenges. Retrieved from https://advisory.kpmg.us/content/dam/advisory/en/pdfs/global-insights-pulse-3q18- report.pdf

Lee, T. R., & Mouritsen, S. C. (2017). Judging Ordinary Meaning. SSRNElectronic Journal. Advance online publication. https://doi.org/10.2139/ssrn.2937468

Lefer, M.-A. (2019-2020). Corpus-based translation studies: Courses description - LTRAD2302. Study programme. Retrieved from https://uclouvain.be/en-cours-2019- LTRAD2302.html

Marciao, J. (2019). NMT und Sie. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (p. 41). BDÜ Fachverlag.

Marin, M., & Camino, R. (2012). Structure and design of the British Law Report Corpus (BLRC): a legal corpus of judicial decisions from the UK. Journal of English Studies, 10, 131-145. https://doi.org/10.18172/jes.184

Marin, M. J. (2017). Legalese as Seen Through the Lens of Corpus Linguistics. An Introduction to Software Tools for Terminological Analysis. JLL. (6), 18-45. https://doi.org/10.14762/jll.2017.018

Marking, M. (2017, January 31). What's Amazon Up to in Machine Translation? Retrieved from https://slator.com/technology/whats-amazon-up-to-in-machine-translation/

Matulewska, A. (2013). Legilinguistic translatology: A parametric approach to legal translation /Aleksandra Matulewska. Linguistic insights: studies in language and communication, 1424-8689: volume 171. Bern: Peter Lang.

McKinsey Global Institute (12/2017). JOBS LOST, JOBS GAINED: WORKFORCE TRANSITIONSIN A TIME OF AUTOMATION. Retrieved from https://www.mckinsey.com/~/media/mckinsey/featured%20insights/Future%20of%20Org anizations/What%20the%20future%20of%20work%20will%20mean%20for%20jobs%20 skills%20and%20wages/MGI-Jobs-Lost-Jobs-Gained-Report-December-6-2017.ashx

Micha, K. (2018). Terminologieextraktion als Kombinationstatistischer und linguistischerInformationsquellen. Edition, Fachzeitschrift für Terminologie, 2, 13-17. Retrieved from http://dttev.org/images/edition/ausgaben/edition-2018-2-e-version.pdf

Mouritsen, S. C. (2017). Corpus Linguistics in Legal Interpretation. An Evolving Interpretative Framework. International Journal of Language & Law, 6, 67-89. https://doi.org/10.14762/jll.2017.067

Mustu, P. (2019). Englische Rechtsübersetzungen: Was DeepL & Co. im Zeitalter von 4.0 (noch) nicht können. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 270-276). BDÜ Fachverlag.

Netcourt (2019). Netcourt. Retrieved from https://www.netcourt.gov.cn/portal/main/domain/index.htm

Ng, A. (2018, December 13). AI Transformation Playbook: AI Transformation Playbook. How to lead your company into the AI era. Retrieved from https://landing.ai/ai- transformation-playbook/

Perrault, R., Shoham, Y., Brynjolfsson, E., Clark, J., Etchemendy, J., Grosz, B., Niebles, J. C. (12/2019). The AI Index 2019 Annual Report: AI Index Steering Committee.

Porsiel, J. (Ed.). (2017). Maschinelle Übersetzung: Grundlagen für den professionellen Einsatz. Berlin: Bundesverband der Dolmetscher und Übersetzer e.V. (BDÜ).

Prikhodko, O. (Ed.) (2019). How to make AI work for your business and not waste money. Frankfurt, Germany: Pravoved, Russian Legal Tech Company. Retrieved from https://www.legal- revolution.com/images/pdf/slides2019/How_to_make_AI_work_for_your_business.pdf

Ruano, R. M. (2019). Beyond Descriptive Legal Translation Studies. In I. Simoniuvs & M. Kristiansen (Eds.), Forum für Fachsprachen-Forschung: Band 149. Legal translation: Current issues and challenges in research, methods and applications. Berlin: Frank & Timme, Verlag für wissenschaftliche Literatur.

Said, F. (2019). Sinnvolle Digitalisierung von Kundenprozessn bei beglaubigten Übersetzungen für Provatkunden. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 50-59). BDÜ Fachverlag.

Sandrini, P. (1999). Übersetzen von Rechtstexten: Fachkommunikation im Spannungsfeld zwischen Rechtsordnung und Sprache. Forum für Fachsprachen-Forschung: Bd. 54. Tübingen: G. Narr.

Sarcevic, S. (2006). New approach to legal translation. The Hague, London: Kluwer Law International.

Schmitt, P. A. (2019). Google Übersetzer und DeepL: Lächerlich, bedrohlich oder hilfreich? In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 180-201). BDÜ Fachverlag.

Schulz, J., Sohn, E., & Horn, A.-K. (e. a.) (2019). EY Law Panel Rechtsabteilung 4.0. LEGAL ®EVOLUTION 2019. LEGAL ®EVOLUTION 2019, Frankfurt am Main, Germany. Retrieved from https://www.legal-revolution.com/de/programm/vortraege-und- panels

Schumann, A. (2019). Modernes Vertrags- und Dokumentenmanagement als gemeinsame Plattformfür Rechtsabteilungen und operative Bereiche. Legal Department Contract Management Specialist. LEGAL ®EVOLUTION 2019. LEGAL ®EVOLUTION 2019, Frankfurt am Main, Germany. Retrieved from https://www.legal- revolution.com/de/programm/vortraege-und-panels

Solan, L. M. (2017, August). Patterns in Language and Law. International Journal of Language & Law, 6, 46. Retrieved from SSRN: https://ssrn.com/abstract=3026373

Susskind, R. (2019). Online courts and the future of justice (First edition). New York: Oxford University Press.

SYSTRAN (2020). S-curves in the history of machine translation. Retrieved from https://www.systransoft.com/understand-neural-machine-translation/

Tessuto, G. (2008). Legal Concepts and Terminography: Analysis and Application. In V. K. Bhatia, C. N. Candlin, & P. E. Allori (Eds.), Linguistic insights, 1424-8689: v. 64. Language, culture and the law: The formulation of legal concepts across systems and cultures. Bern, Oxford: Peter Lang.

Tognini-Bonelli, E. (2001). Corpus linguistics at work. Studies in corpus linguistics: v. 6. Amsterdam, Philadelphia Pa.: J. Benjamins.

Vogel, F., Hamann, H., & Gauer, I. (2018). Computer-Assisted Legal Linguistics: Corpus Analysis as a New Tool for Legal Studies. Law & Social Inquiry, 43 (04), 1340-1363. https://doi.org/10.1111/lsi.12305

Vogel et al. (2017). “Begin at the beginning”. Lawyers and Linguists Together in Wonderland. JLL. (6), 90-100. https://doi.org/10.14762/jll.2017.090

Wagner, J. (2018). Legal Tech und Legal Robots: Der Wandel im Rechtsmarkt durch neue Technologien und künstliche Intelligenz / Jens Wagner. Wiesbaden, Germany: Springer Gabler.

Willis, G. (2019, October 28). Interview by T. Istomina. Düsseldorf, Germany.

Yeong, Y.-L., Tan, T.-P., Gan, K. H., & Mohammad, S. K. (2018). Hybrid Machine Translation with Multi-Source Encoder-Decoder Long Short-Term Memory in English- Malay Translation. International Journal on Advanced Science, Engineering and Information Technology, $ (4-2), 1446. https://doi.org/10.18517/ijaseit8.4-2.6816

Zielinski, D. (2019). "Intelligenz" beim Übersetzen. In W. Bauer & F. Mayer (Eds.), Übersetzen und Dolmetschen 4.0: Neue Wege im Digitalen Zeitalter: Translating and Interpreting 4.0: New Ways in the Digitial Age. Tagungsband der 3. Internationalen Fachkonferenz des Bundesverbandes der Dolmetscher und Übersetzer e.V. (BDÜ) (pp. 42-43). BDÜ Fachverlag.

Zimmer, B. (2011, March 4). The Corpus in the Court: ‘Like Lexis on Steroids'. The Atlantic. Retrieved from theatlantic.com/national/archive/2011/03/the-corpus-in-the-court-like- lexis-on-steroids/72054.

[...]