The role of CAT tools in Patent Translation

Tesi di laurea in Laboratorio CAT
Università del Salento, A.A. 2018/2019

The role of CAT tools in Patent Translation - Serena Cuscianna


The concept of patenting dates back to ancient times, though there is no clear proof of who was the first inventor to gain intellectual property. According to British intellectual property expert Robin Jacob, the first kind of patent appeared in the 600 B.C.: it was about a new kind of loaf of bread (Kwong, 2014). However, several academics, such as the professor and architect Massimo Ricci, agree with the idea that the world’s first industrial patent was granted to the Italian architect Filippo Brunelleschi in 1421 for an ingenious crane system for shipping and transporting marble from the quarries in Carrara (Tuscany, Italy). About thirty years later, Henry IV granted the earliest known English utility patent to Flemish-born John of Utynam. He obtained the 20-year monopoly on producing stained glass with a method that until that moment was unknown in England. English patent system is considered to be the world’s oldest continuously operating system of patents, but not the oldest one ever existed. Also in this case, Italians hold the record: in 1474, the Venetian Act was signed as the first patent law. From that moment on, patent legislation evolved and updated across the world to the current status (Kwong, 2014).

Since the earliest times, the patent has represented both a technical and legal text and its intrinsic language constitutes a jargon requiring the utmost attention during the translation process. The latter is an even more ancient process than the concept of “patent”. The shared idea is that the Bible was the first major translation in the Western world. However, though it is evident that the translation need was born as early as the first human interactions, scholars and historians indicate a Sumerian poem (the Epic of Gilgamesh, which was translated into Asian languages, and dates back to around the second millennium BC) as the first translated document. In later times, the need for translation became greater, especially after the Industrial Revolution when the economy rapidly developed. With the advent of industrial printing, new machinery speeded up the production of business-related texts and materials with a consequent gain in the time available to invest in research and translation of material to enter foreign markets (Kwong, 2014).

Since the 18th century, companies have benefited from official translation services, but the dawn of modern practice came with the development and the progressive introduction of Information Technology (IT). After World War II, the first machine translation experiments were performed, but the results obtained were excessively expensive and not practical. For instance, in 1945, the Georgetown University and the International Business Machines (IBM) jointly developed the first public demonstration of machine translation, also known as the “the Georgetown- IBM system”. It was the «first non-numerical application of the newly invented ‘electronic brains’ that most people had heard of» (Hutchins, 2006, p. 1). The demonstration raised great expectations but the results from MT research in the next 10 years were disappointing. As a consequence, an investigation committee, known as the “Automatic Language Processing Advisory Committee” (ALPAC), was set up to consider the prospects for MT. The committee «concluded that current MT systems were inadequate and uneconomic and that there was no immediate or predictable prospect of useful machine translation» (Hutchins, 2006, p. 28). Consequently, a different path has been chosen. Today, computer-aided translation seems to be the best compromise in patent translation: translation is carried out by a human translator, who uses software tools (proprietary or open-source, installed on her/his computer or remotely accessible) to speed up the translation process.

This dissertation aims to illustrate how patent translation is facilitated by the use of specific software, known as computer-aided translation tools, resulting in a higher quality product.

Taking into account the structure of this dissertation, the first chapter will explore the concept of “patent”. In addition to the European legislation, with in-depth references to the Italian and English system, a general overview will be provided by explaining what a patent is, when something is considered patentable, the lifecycle of a patent and the process management. Furthermore, attention will be paid to presenting the types of patents that can be obtained on the basis of their subject matter, each one with specific characteristics depending on Countries.

Overall, a patent maintains an almost standard structure. The linguistic register has its specificity, too. The second chapter of this dissertation will focus on how to write a patent: mandatory elements will be listed and discussed, such as the cover page or the claims. In fact, in patents, the writing style «is virtually a formula, which the patent attorney has constructed from a number of preferred phrases and expressions that are linked together by a finite set of possible connectors or flags» (Zerling, 2010, p. 3). Since patents are legal documents with highly technical terminology, it is advisable to hire a professional service to handle the drafting, revision, and translation of a patent as well as a good attorney who will assist the inventor in a possible lawsuit. The issue of the involved actors will be one of the topics discussed in the second chapter.

The third chapter will focus on computer-aided translation; it will provide a definition of CAT and its core components (e.g., termbases, translation memories, etc.). Today, there is a wide variety of both commercial (said, shareware or payware) and freeware translation software. This dissertation aims to draw attention to a specific type of specialized translation that is affected by a profound cognitive gap. With the aim of carrying out a study that is as accurate as possible, the analysis has been limited to a single CAT tool: SDL Trados Studio. This CAT tool was chosen because, as stated by different authors such as Samuelsson-Brown (2010, p. 78), SDL Trados Studio is the translation tool most widely used by translators who want to translate and review projects, use terminology and get the most out of machine translation. Besides, Trados Studio has been the object of several comparative studies with other CAT tools and has proved to be the one with the most comprehensive features (Kurniawati, Rahajeng, Kristanto, & Kastuhandani, 2016, p. 103).

Patent translation differs from other forms of translation, in so far as it requires highly specialized terminology and highly specialized experts. Patent translation is in between legal translation and technical translation, and standard university courses rarely include classes on legal translation or, even more rarely, on patent translation. My personal interest in this niche branch of translation stems from a three-month internship at the translation company Global Voices LTD (, specialized in patent translation. Through this experience, I have had the chance to approach this type of translation which was new to me, and personally intervene in the translation of dozens of utility patents on several subjects: engineering, life science, medical devices, and chemical products.

Based on the translation experience gained during the internship, the third and fourth chapters will explain the role of CAT tools in patent translation, by making particular reference to the use of translation memories. Taking advantage of one of the projects completed during the internship, I will analyze the performance of the translation memories used with SDL Trados Studio software.

The study has been carried out in full compliance with the directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC, through the use of anonymized and appropriately decontextualized data, to ensure compliance with the period of secrecy of the patents examined, and the Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation).

Read the next chapter “The patent: overview“.