Translating a terminology system, not a list of words
You have provisionally translated Aufnahme as holder. A few paragraphs later, the patent introduces a Halter that fits into that Aufnahme. The second component holds a sensor; the first receives the second. Suddenly, the English noun that looked perfectly reasonable on its own has become a problem in the company of its neighbours.
The difficulty is not simply finding another word for holder. It is deciding whether the patent distinguishes two components, two functions, or two ways of describing the same thing. That decision cannot be made by considering either German term in isolation.
This is a different test from asking whether an individual translation is correct. Several plausible English terms can, when adopted, make the relationships in a patent harder to follow. German patent terminology therefore has to be translated as a system: the choices must work together, preserving not only how each feature is named but also how it relates to everything around it.
When one translation constrains another
Consider a simplified example constructed for this article:
Der Träger 10 weist eine Aufnahme 12 für einen Halter 14 auf. Der Halter 14 hält einen Sensor 16.
Suppose the surrounding description establishes that the Aufnahme is a receiving portion of the Träger, while the Halter is a separate component inserted into it. A workable rendering would be:
The support 10 has a receiving portion 12 for a holder 14. The holder 14 holds a sensor 16.
The English now preserves a chain of relationships. The receiving portion belongs to the support; the holder fits into that portion; the sensor is held by the holder. The nouns have been chosen with those relationships in view.
The question is not whether holder is always wrong for Aufnahme. It is whether that choice fits the function established here and the distinctions the patent maintains. If Aufnahme and Halter both become holder, a sentence about a holder receiving another holder may remain intelligible, especially with reference signs. Nevertheless, it may obscure a distinction that the German maintains quite economically.
Nor should a translator invent a more specific noun merely to make the English labels different. Socket could imply a particular receiving geometry; bracket could suggest a component form the source has not established. Receiving portion is defensible in this constructed example because the surrounding description supports it, not because the translator needs an unused word.
Each choice therefore constrains the others. A good equivalent must fit its own concept and remain compatible with the terminology selected for neighbouring concepts. The requirement is not simply to use a different English headword for every German label, but also give consideration to the distinctions between them and how the terms relate to each other.
A part of something is not a kind of it
Terminological relationships are not all of the same kind. A screw, for example, can be a type of fastening element, but it can also form part of a fastening arrangement. These are different relationships: the first is classification, the second is part-whole.
In terminology work, these are called generic and partitive relations. For the patent translator, the key is spotting whether a term names a category, a type within it or part of a larger whole.
Imagine that the sensor holder in our example is secured by a Befestigungseinrichtung. A further invented passage reads:
Die Befestigungseinrichtung 20 umfasst zwei Befestigungselemente 22. In der dargestellten Ausführungsform sind die Befestigungselemente 22 Schrauben.
A translation that preserves the structure would be:
The fastening arrangement 20 comprises two fastening elements 22. In the illustrated embodiment, the fastening elements 22 are screws.
Here, Befestigungseinrichtung names the fastening arrangement, while Befestigungselement a component within it. A translation that uses ‘fastener’ for both may read fine in isolation, yet still flatten a necessary distinction. Schraube is narrower still, naming the specific type of element used in the embodiment. The translator needs to preserve not just word meanings, but also the relationships between terms and the level of generality at which the source is speaking. That shows how plausible word choices can still disrupt the terminology structure.
The document can change the relationship between terms
The terminology system is not always the one a translator expects from general technical knowledge. A source may expressly explain how particular terms relate within that document.
Regelung and Steuerung illustrate the issue. A translator may expect a contrast between closed-loop control and open-loop control. Yet recognising that possible contrast does not establish how a particular patent uses the terms: its own definitions still need to be checked.
A concrete public example is WO 2007/085293 A1. Its German description expressly uses Steuerung to cover an open-loop system, a closed-loop system, and a system performing both functions. For translation purposes, that statement makes the word an umbrella term in that document rather than the exclusive opposite of Regelung. Mechanically translating every Steuerung as open-loop control would restrict the breadth the author has explicitly assigned to it.
The relationship between the two concepts may differ from the conventional distinction in a particular patent.
Recognising an author’s usage is not the same as endorsing it as good drafting. The EPO’s Guidelines caution against using established terms with different meanings where that is likely to cause confusion. However, the translator still needs to understand the source’s intended system, while keeping any drafting concern separate from the translation decision.
Consistency is not a one-word rule
This explains why one German word, one English word is an unreliable definition of consistency. It can fail in both directions.
An optical inspection patent could use Aufnahme for a mechanical receiving feature and Bildaufnahme for image capture. The shared German form does not make these the same concept. A comparable example of this distinction in the meaning of Aufnahme can be seen, for instance, in WO 2018/206484 A1: claim 1 distinguishes an Aufnahme in the vehicle body from a camera-related Bildaufnahmeabschnitt. Applying one English equivalent to both uses would create uniform wording at the expense of meaning.
Conversely, two source expressions might genuinely designate the same component. Perhaps the description introduces a long functional name and subsequently uses an explicitly defined shorter form. Treating those as separate concepts simply because the strings differ would be another kind of mistranslation.
Neither lexical similarity nor lexical difference is the defining factor here. The translator has to establish what each occurrence refers to, whether the document defines an equivalence, and whether any apparent distinction is sustained by the technical context. Where the evidence does not settle the matter, the translation should not quietly resolve it by inventing a tidy naming scheme.
This is also why a preferred English term from one patent is not automatically the preferred term for the next. The underlying object may be similar, but its position in the document’s terminology system may differ. A label that distinguishes two neighbouring features neatly in one application may merge them in another. Consistency is something to establish within the invention, not merely to import from an earlier bilingual list.
One revision can affect the whole set
Once terminology is understood as a system, revising one English term can require more than simply replacing a repeated word.
Suppose receptacle was initially used for Aufnahme, but later passages show that the term actually refers to an integral receiving region rather than a separate, container-like component. Changing the translation to receiving portion may solve the immediate problem, but the consequences do not end there. Related terms and phrases now need to be checked against that revised understanding.
If the source also refers to an Aufnahmeöffnung, for example, the translation should still make clear that this is an opening in or associated with the same receiving portion. Descriptions of components entering, engaging with or being supported by that region must remain consistent with the new interpretation. A change in one core term can therefore require a targeted review of the terminology around it.
This does not mean mechanically reproducing the same German stem wherever it appears. Natural English may express related concepts using different words. What matters is that the relationships between those concepts remain clear and that the translation does not accidentally turn a feature into a separate component, merge two distinct features, or change the level of generality used in the source.
When revising a term, I therefore ask what the feature belongs to, what it receives or acts on, which other features are defined in relation to it, and whether the passage is speaking generally or about a particular embodiment. Those relationships often reveal which other parts of the translation need to be checked once an interpretation changes.
What a useful terminology record preserves
A two-column glossary can record a chosen equivalent. A more informative entry will also record what the term denotes in this patent and how it relates to neighbouring concepts. Aufnahme as receiving portion becomes more useful when accompanied by a note that it is part of the support 10, receives the holder 14 and is not the holder itself.
This does not require an elaborate database for every job. It requires enough context to prevent a later decision from undoing an earlier distinction. A brief definition, the relevant reference sign and a note about a broader, narrower or otherwise related term may be sufficient.
There is also a formal reason to take consistency seriously. PCT Rule 10.2 requires terminology and signs to be consistent throughout the international application. The EPO’s guidance on the description likewise emphasises consistency across the application. Such requirements establish the importance of coherent terminology, but they do not tell the translator how to achieve it across two languages.
The translator still has to do that work: distinguishing a receiving feature from the component received, an arrangement from its constituent parts, a category from a particular implementation, and genuine synonymy from an unresolved inconsistency in the source.
The result is not simply a list in which every German term has an English partner. It is a translation in which the reader can follow the same conceptual structure as in the German. The terminology is successful when those relationships survive the translation, not merely when the chosen words recur.
Leave a Reply
Want to join the discussion?Feel free to contribute!