Where the terms come from, or "click me gently"

Virtual reality has become such an integral part of our lives and language that we don't even stop to think about how to correctly say — "browser" or "viewer", "support" or "settings", "icon" or "avatar", "press" the mouse or simply "click" it, "Web" or "web", "Internet" or "internet", etc. What really is the difference when it comes to terminology?

Before the Internet was a household name, it was prohibited to write technical documentation for procedural technicalities - almost everything was regulated by government standards. In the new era, the language of user manuals is more focused towards meeting the needs of all of humanity. Why do corporations continue to spend so much money on establishing and controlling branding?

Let's return to the "press or click" example. Personal devices have edged out classic computers, the touch interface has already changed the scope of everything, and this is just the beginning: the video-, sound-, neurological-, and even more sophisticated means of exchanging data and commands with computers, gadgets, intelligent assistants, navigators, robots, and other machinery are rapidly building up their artificial intelligence and sensor functionality. And to cope with all the new ways of managing one’s virtual environment, you'll need new gestures, motions, voice commands, and other language tools.

All the variety of actions will require verbal notations, which will be short, clear, distinct from each other, generally accepted and not void of living language. Nowadays, we talk to our devices, drawing our finger over their screens, tapping them from time to time, occasionally pushing mechanical buttons that are almost a thing of the past, even a simple finger snap can issue commands with some devices. That's why all the short, vivid words and phrases used in the digital realm need to be clearly discussed for the future, so that they do not end up void of designated meaning.

The other reason comes purely from a product marketing perspective. When one company device differs from another company device via the form of a button only, or with a number of taps or strokes for unlocking, device manufacturers start patenting everything desperately, including implementing "a unique way to recognize the device owner by the sound spectrum of their finger snap."

In this regard, the translation community is critically important - as terminology translators would normally be eager to name technological actions sometimes in an egotistical manner by working on another "masterpiece of terminologist's thought" - but instead, they have to strive to achieve the right approach that will provide a lasting meaningful purpose. This comprises the development of the verbal notations system, which must be simple and understandable for the mass consumer market, and which does not fail due to misguided ambiguity of actions, and which must be focused on describing capabilities of modern sophisticated technology, and - moreover – be compatible with the terminology of previous generation technology that is still in service.

Understanding the complexity of the problem and gaining a lot of experience with the terms of very demanding customers, Logrus Global LLC provides a full range of terminology development and management services. A glossary is not simply a dogmatic set of rules. In most cases, it is required to adapt the well-known template solutions to the customer's needs, and you have to be able to ask the customer the right questions that fit the technology being implemented.

Depending on the size of the project, such work can vary from the primary extraction of dictionary terminology from the documents in a single language, identifying discrepancies and errors in previously created bilingual glossaries and consistent integration of new terms into the existing glossary, to the comprehensive support of multilingual terminology databases and manuals throughout their life cycle - from creation - through upgrades - to removal and replacement with new terms. We will explain in the following paragraphs what types of work methods are required for managing terminology and ensuring translation consistency.

Retrieving terms from source texts

To create a glossary of any document, it is necessary to extract the terms from the client's text first. If the text is small, this action can be performed manually, however, if its volume is measured in author's sheets (for reference: the author's sheet contains 40,000 characters with spaces or 20 conventional pages), it is better to automate this process.

Automated dictionary compilation programs extract terms from text using statistical and linguistic methods that take into account the morphological, lexical and grammatical features of the terms, and, of course, their frequency of use. Over the course of the last decade, Logrus Global LLC has created and developed its own terminology extraction tools.

Of course, even the most intelligent programs do not produce a ready-made glossary. For example, words that do not have to be in the glossary are removed by means of an individual stop sheet for each subject area. The frequency with which the phrase appears in the text is not a decisive sign of the term. The programs do not distinguish the colloquial vocabulary from the special vocabulary, and this is a serious common problem for languages and cultures, where the professional jargon of the documented product's developers suffers from extra figurative and metaphorical use. Finally, key terms can be met on rare occasions, or not be met at all! For example, when Microsoft invented the Microsoft Fluent Interface (or Ribbon), the term "Ribbon" was so top-secret (to prevent competitors from assessing methodology), that it was removed from the translated texts almost until the product release.

In light of these reasons (and various others), the final development of the automatically assembled glossary is inescapable. It can be conducted partially by hand at the stage of terminology analysis and demands the involvement of not just an editor, but a terminologist itself, who can build up and keep in mind the entire glossary "tree" and who has a special nose for terms and internal contradictions amongst the entire glossary.

Analysis and translation of terms

In terminology analysis, each selected term needs to be matched with a definition and examples of its use. Therefore, the optimum equivalent in the target language will be chosen. The classic example being the terminology accountability methodology used that were relevant in the days of the MS-DOS operating system. Personal computers that were compatible with IBM PCs could use three kinds of RAM (Random Access Memory), which were called conventional memory, extended memory, and expanded memory in English manuals (in fact, there was more types of memory, but let’s not complicate the story too much).

The MS-DOS operating system could only use the Conventional Memory, however, technically, the same memory beyond the "address space" of MS-DOS was called Extended Memory, and the memory that was stored on individual expansion memory cards (and soon after - on the same exact boards as conventional memory) was called Expanded Memory. The odd thing is that these two English words - "extended" and "expanded" - are synonymous, therefore, they have one and the same meaning in some languages also.

There also can be other situations when it comes to the use of specific terminology: one term can be translated differently depending on the area of use. For example, the word "frame" will be translated as "framework" in building and construction, or as " body" in machine tool building, "data unit" in 3D graphics and animations, "false charge". Therefore, during the terminological analysis phase, we always involve experts in the subject area.

Throughout the iterative process of terminological analysis, it is nearly impossible to think of translations for all the original terms at once. However, when some terms are translated, one has to compare new translations with the previous ones and verify. This process determines whether they are in agreement or not, whether there are latent or obvious contradictions, whether a linguistic impasse has been reached at some point choosing the variant of a translation, and which may have subsequently caused seemingly unresolvable issues in choosing translations for the remaining terms in the queue.

A separate additional problem is the so-called "nested" terms: terms consisting of a single word are included in the double-length terms and the last ones - in triple-length in their turn. In such constructs, it is very easy to get into a mess and confuse future users. For example, in MS-DOS, all the double-length terms contained a single-length term - Memory, which could not be translated differently than "memory," and thus, it was necessary to come up with non-matching translation variants for the second of the two words. And this is just in the most simplified case.

An analysis of the terms in the source language and their translation results in a bilingual or even multilingual terminology dictionary - the Glossary.

The Glossary is built - the work is just beginning

The Glossary is created in the first approximation. From that point on it then becomes a dual-purpose tool.

On the one hand (on the side of terms in the source language), the Glossary limits the fantasy of developers and becomes a required document for the customer's technical writers working on the creation of manuals, and user interface developers. Neither of them should invent new terms without serious need, and, if so necessary, they must follow a certain procedure for approving a new term for use.

On the other hand (on the side of translation of terminology), the Glossary becomes a mandatory guide for documentation translators and the user interface itself.

In order for the Glossary to perform both of its functions well, it needs to be integrated both with content management systems on the side of the localizable product developers, and with the translation automation systems of the translation service providers selected by the customer.

No matter how good the new Glossary is, it's never complete. Simply because it is built on a limited set of texts and new texts are created continuously, the original products are considered as "alive", the original terminology is considered as "floating". The Glossary needs to be maintained throughout its life cycle, allowing it to be accessed on both sides, by product developers and by hired translators.

In bad cases, the entire terminology management systems have to be used to streamline the process of creating new terms and approving their relative translations. In the past, the larger clients held a full complement of terminologists and linguists for each language that needed to be translated. This has become an unattainable luxury in the modern day, therefore, customers are either plunged into chaos - when terminology and word generation are no longer monitored, and the Glossary exists just under its own limited momentum, or they simply outsource terminology management to external suppliers.

Outsourcing of terminology services is actually a very progressive step… if it is outsourced to the right professional providers. However, sometimes the client’s inner Scrooge often leads them to outsource linguistic services to the graduates of the language chairs in third world countries, where a non-native language is studied, instead of applying directly to well-qualified native speakers. The interpreters (which are a toss-up of either amazing or going into a tail-spin speakers) then come to know the "qualified opinion of a reviewer-terminologist" hired by the employer in the cheaper location. In other words, it’s anyone’s guess what the finished product may look like.

Packaging the Glossary in a Terminology Database

It is not enough to simply create a Glossary — it must be connected with the translation environment utilized by the customer's translators. Then, the Glossary has to be converted to the format of one or another database.

In the most simple cases, you do not need the database in principle —the customer produces the Glossaries in Excel table format.

However, in the case of Computer-Assisted Translation (CAT), turning the Glossary into a CAT System readable format is required. There may be a variety of problems (the world of information technology is far from perfect — the way the programmers are sometimes being called - "tech ninjas", or their jargon phrase -"to monkey around" are no accident!). Encoding can fail, and in this case, you'll see some "mojibake" instead of letters. Not all letters in a row can end up incorrectly converted. Some reserved characters may also get misplaced - a lot of things can happen. Therefore, even if you have an excellent, professionally prepared Glossary, it is better to entrust its conversion into a terminological database to the experts. After all, they're well aware of common problems of converting data into one or another format for different languages, and are well aware of how to professionally handle them.

Glossary: That's not the half of the story - ask for a Translation style guide

We develop translation style and execute guides for demanding customers who take care of their brands expeditiously. The words in the Glossary can be correct, but it is still possible that they may end up connected to common vocabulary in the sentences in such a way that the text will still not be understandable due to use. Or the text will be "like the other one" - it will be grammatically correct, but not at all interesting to read. When this occurs, a specific document as style guide comes to the rescue. In short, it describes everything you cannot fit into a rigid Glossary structure - something without which the customer will never accept as a ready translation. In particular, the style guide defines the requirements for translation of standard structures (headings, interface elements, names, titles, messages), a list or description of untranslatable elements (names of companies, products, abbreviations, etc.), the types of fonts used, and so forth. It takes into account linguistic and cultural characteristics of the target audience, specifics of a project and any wishes and requirements of a client.

Creating comprehensive style guides is a separate job entirely. The main difficulty here is that the customer is often unable to articulate what they want or what exactly they do not want in the translation. We are often told by customers to "Translate it like in that document," or, "I want the tone of voice like on that website..." For such cases, there is a style guide which carefully gathers (on the basis of hundreds and thousands of translated projects) demonstrative examples and a representative idea of what the style actually is, what it is composed of, and how to transform the associated vibe (feeling) into clear instructions.

Many years ago, when Logrus Global LLC was just starting its work in the translation industry, we produced a landmark book - "The Translator's Bible"- which laid the basic groundwork for future translation style guides.

The Glossary stands alone, the translation stands alone, and other aspects of linguistic quality control

If you already have the Glossary, we can help you to improve it: to reveal the needless synonyms, contradictory translation terms, and even orthographic mistakes.

A ready-made Glossary can be supplied by the "correct" introduction of new terms from newly written documents.

Or conversely, any available translations can be validated against the provided Glossary. And not just against the Glossary, but also by taking into account all the additional rules and exceptions described in the Translation style guide. As a result, a customer receives the full report detailing detected errors and corrected translations as well, if necessary.

A separate task is to reconcile translation memory with the Glossary. If the customer has all the translations stored in one database, some rarely reused translation fragments (segments, units) may become obsolete in relation to the terms they contain due to the time of use. By random (manual or automatic) substitution in a new document, such outdated terms can be skipped by the translator and, as a result, the translation will likely be wrong from a terminological view point. And since automatic translation of 100% substitutions without their subsequent necessary proofreading has become rather the industry's rule, the terminological clearness of translation memory is now becoming a much more pressing issue.

By reconciling translations with the Glossary, we use our own and third-party automation tools, which recognize different forms of one word (typical example here is the case-endings in a number of languages), because the terms in the Glossary have their root form (nominative case), but in the text their word formation duly begins.

In fact, multiword terms can be "split" by reserved words that make them more unrecognizable. Our tools also work in these highly complicated cases.

The administered steps mentioned above provide a general idea of the common issues associated with creating a Translation Glossary and do not necessarily describe the whole variety of terminology services provided by Logrus Global LLC. We have a wealth of experience in dealing with often nontrivial terminology tasks. We also maintain a large number of vocabulary-based automation tools developed by our own team of programmers and engineers. Regardless of the subject area complexity, the rarity of the language pair for translation, or the data format ingenuity, our terminologists will provide the right product language to meet your corporate needs!