Does a Glossary Really Ensure Translation Consistency? A Few Myths from the Translation Industry

Anyone who has ever needed a large amount of technical documentation translated within a very short timeframe has probably come across translation agencies saying things like:

“Thanks to our glossaries, we ensure terminological consistency even in projects handled by multiple translators.”

It sounds reasonable. But is a glossary really such an effective tool? Or is its role sometimes overestimated?

In this article, I take a closer look at several assumptions that frequently arise in projects carried out by several, and sometimes even a dozen or more, translators. From the client’s perspective, these assumptions sound convincing. In practice, however, reality proves to be more complex, and the mere fact that every translator receives a glossary says very little about the quality of the final translation

What exactly is a glossary?

Before we go any further, let’s define the main subject of this article. What exactly is a glossary in a translation project?

Simply put, it is a list of agreed terminology that should be used consistently throughout the entire project. It may include the names of machine components, functions, on-screen messages, safety features, or the terminology adopted by the manufacturer.

Such a document helps everyone involved in the project make consistent terminological decisions. It is particularly useful when more than one person is working on the translation or when the project extends over many months.

So the idea itself is an excellent one. The real question is how much it can actually achieve.

A Real-Life Business Scenario: Several Hundred Pages of Documentation for a Modern Industrial Production Line. The Translation Is Needed by Next Week.

Sooner or later, almost every internationally operating company faces a translation project that simply cannot be completed within the required deadline unless the work is divided among several translators. These projects typically involve such large volumes of text and such tight deadlines that they simply cannot be handled by a single translator. In such situations, the natural solution is to choose one of the available approaches to multilingual project management and outsource the work to a translation agency. The agency takes care of the organisation: dividing the text into sections and assigning each part to the required number of translators.

If a glossary has been prepared for the project, the assumption is that every translator receives it. In theory, everything looks perfect. Everyone works with the same terminology, so the final translation should be consistent, the terminology should be used uniformly throughout, and the final document should read as though it had been translated by a single person. In practice, however, there are several problems that are rarely discussed.

If All the Translators Had the Same Glossary, Why Did the Project Still End Up Inconsistent?

Does a Glossary Really Guarantee Translation Quality?

A glossary certainly helps. It reduces the risk that different translators will make different decisions when referring to the same element. For example, the name of a machine, one of its components or an operation within its cycle.

However, useful as it is, a glossary is not a magic tool. It does not make decisions on the translator’s behalf, it does not solve every linguistic problem nor does it eliminate differences in professional expertise. Put simply, it will not enable someone with no experience in translating technical documentation to produce a translation of the same quality as someone who has been translating such documentation for many years.

That alone shows that a glossary is not nearly as powerful as it is often claimed to be. And the more time passes, the more apparent this becomes.

This does not mean, of course, that glossaries should be abandoned. Quite the contrary. In projects divided among multiple translators, they are indispensable. The problem only arises when people begin to expect them to do something they simply cannot do.

What Happens to a Glossary When a Project Lasts for Months (or Even Years)?

First of all, when a project runs for many months, or even years, and is carried out by external specialists, the composition of the project team inevitably changes over time. The longer the project lasts, the more members of the original team leave and the more new people join it.

The agreed terminology often changes as well, and sometimes the glossary simply becomes outdated. In my experience, however, very few people actually think about keeping it up to date.

Many Participants, One Set of Rules. Is That Really the Case?

There is another aspect to this, one that is something of an open secret within the industry. Clients often assume that everyone involved in the project works according to the same rules. The truth, however, is that this is not always the case. Sometimes, some translators never receive the glossary at all. Others forget about it or fail to use it consistently. The longer the project continues, the more common such situations become.

Over the years, I have joined multi-translator projects at various stages for manufacturers of construction equipment, agricultural machinery and plastics processing equipment. These are well-established brands on the Polish market that produce documentation for their machines in multiple languages. Yet I could count on one hand the number of times I received the glossary before carrying out my very first translation. In all the remaining cases, I had to ask for it myself, and even then I did not always receive it. This shows that, over time, even the best organisational assumptions simply lose out to the constantly changing composition of the team carrying out the project.

Does a Larger Glossary Mean a Better Translation?

In my opinion, no. I would even go so far as to say that an overly extensive glossary can have the opposite effect to the one intended. It is difficult to say exactly where the line is beyond which a glossary becomes too large. Every project has its own specific requirements. However, if a glossary is compiled without much thought, following the principle of “the more, the better”, there is a risk that it will eventually include too much, too many mandatory equivalents.

At a certain point, instead of keeping the project under control, it begins to make it unnecessarily rigid.

The problem is not the number of entries itself, but what happens when the glossary starts to include terms that cannot be unequivocally assigned to a single equivalent.

Not Every Term Belongs in a Glossary

A glossary works extremely well for names that should remain consistent throughout the project: machine names, component names, workstation names, stages of the operating cycle, or messages displayed on screens. These are elements specific to a particular machine, and they are precisely the ones that require consistency.

A simple example is the English term welder / welding machine. Although it is a basic term that can be found in any dictionary, it is worth specifying in the glossary that, in this project, the machine should be translated in Polish as spawarka rather than zgrzewarka (or vice versa). In projects involving multiple translators, not everyone receives the entire document or has access to the full context, and the English term welder alone does not always make it clear which type of machine is being referred to. Unless this is established in the glossary, one part of the documentation may refer to it as a spawarka, while another calls it a zgrzewarka. Yet throughout the documentation, it is the very same machine. Once a terminological decision has been made, it should therefore remain in force until the end of the project.

The situation is different when it comes to terms that do not refer to one specific machine component but instead belong to general technical language. In such cases, it is impossible to assign one correct translation once and for all.

In many cases, a term in the source language does not have a single Polish equivalent. Its meaning depends on the context, and only after understanding the element being described can the translator determine which translation is the correct one.

A good example is the Italian word giunto. Depending on its function, it may mean a joint, a coupling, or a seam. If the glossary prescribes only one of these equivalents, the translator is faced with a choice: follow the glossary or follow the actual meaning of the text. And that choice is not always straightforward.

Glossaries are highly effective at standardising decisions that only need to be made once. They cannot replace the decisions that have to be made anew each time on the basis of context.

This is precisely where the limits of a glossary lie. A good glossary does not answer every question. It only answers those for which there is genuinely a single correct answer. Wherever the meaning of a term depends on context, a glossary entry cannot replace the translator’s expertise or careful analysis of the text.

For the client, this means that even a very extensive glossary does not guarantee that a translation produced by multiple translators will be consistent throughout or that the finished text will be accurate. At a certain point, what matters more than the number of glossary entries is the person who is able to determine the meaning of a term within the context of a particular text.

What Should a Glossary Really Contain?

From the perspective of more than 20 years of translating technical documentation, I believe I can confidently put forward a rather bold proposition, one that goes against long-established industry standards:

A glossary should contain project-specific decisions, not replace translation decisions.

Because:

• the name of a workstation is a project-specific decision;
• the name of an operation is a project-specific decision;
• the name of a particular machine is a project-specific decision.

Whether giunto means a joint, a coupling or a seam in a given sentence, however, is not a project-specific decision. It is a translation decision that arises from the context.

In a well-prepared glossary, the number of entries is not an objective in itself. The objective is to eliminate those uncertainties that can genuinely be resolved once and then applied consistently throughout the project. All remaining decisions should be left to the translator, because only the translator has access to the full context of a particular sentence. Ultimately, it is the translator who is responsible for ensuring that the client receives a document that accurately conveys the meaning of the original and is fit for practical use, not a text produced simply by applying the same equivalents consistently every time.

This becomes particularly important in the case of operating manuals, where translation quality has a direct impact on usability, safety and legal compliance. You can read more about this in my article on operating manual translation for export.