This question can be a bit difficult to answer because some problems will depend on which languages you are translating between. For that reason, I will try to keep the problems listed more generalised but use examples for clarity when I can think of them/find them. Let’s jump in!
1) Context – Specialised vocabulary is often used in a specific context. While it may be obvious that specialised vocabulary and specific contexts go together, this is not always the case. An example where they do go hand-in-hand would be in a mechanic shop where two mechanics are discussing what is wrong with a car. Their conversation is almost guaranteed to contained specialised automotive vocabulary that would not be used in other contexts. An example that goes against this would be someone who comes up to tell you that they have broken their phalanges. In this context, there is no apparent reason for them to use a word like this because it is specialised vocabulary (i.e. medical) and the more common term would be fingers or toes (or bones within the fingers or toes). They might have used it because they didn’t know/forgot the word for fingers or toes and are trying to use a word that may be recognised on a multilingual scale as medical – especially Latin anatomical names – vocabulary often is.
So what does this mean for translation? When examples of specialised vocabulary appear outside of their obvious context (e.g. automotive vocabulary outside of an automotive setting), it can sometimes be difficult to identify, especially for machine translation (think Google Translate, DeepL, etc). While the word phalanges is not used in a general context, there are specialised terms that are used in both general English and one or more specialised fields. In these cases, the same word represents different concepts. An example I like to use in my classes the word “power”. “A government in power” and “electrical power” are not the same thing; in these two instances, the word “power” means two different things but someone who does not realise that there is a difference would not understand. I recently did a translation of a rental agreement from Austrian German to UK English where “Beilage” was used to mean attachment. It was an MPTE translation, so a machine translator made an attempt first and I was supposed to come back and improve that translation. It thought that “Beilage” was “garnish”, which was obviously wrong.
2) Ability – One of the other answers simply said “ability to do it”. This answer is woefully incomplete but it does point to our next point (or several points, depending on how you look at it). The first “ability” problem is the ability to recognise that a word is a specialised vocabulary. Have you come across this specialised vocabulary before? Do you (or the translator) recognise that it is specialised vocabulary? Does it also have a use in general language? Is there a difference between how it is used in general and specialised contexts?
Let’s say you have been able to successfully identify the specialised vocabulary that you are going to translate, how it is used and also in which context(s) it is used. Our next “ability” is thus: do you have the knowledge to translate it into your target language? Do you know the correct word/phrase/translation to represent this concept accurately?
What happens if you don’t have this “ability” (i.e. knowledge)? Well, fortunately, it isn’t the end of the world. For many languages, there are multi-lingual dictionaries, termbases and other resources that can help you gain this knowledge. Not having our second ability can be overcome by having a third ability: the ability to do research. Do you know the resources that are available for your language pair? If your language pair is more obscure, do these resources even exist? For example, how many Hindi-Klingon dictionaries are in existence?
When I searched for “Hindi Klingon dictionary” on Google I have over 1 million search results. However, only the first two had connections to both Hindi and Klingon. Incidentally, both were labelled as “AI translation” services. To be honest, I was even surprised to find this many results; I was expecting 0 relevant ones. Out of curiosity, I decided to test this AI to see how it would do. I choose the Hindi word for “port”, meaning the direction left on a ship, which I believe is बायां (baayaan – Hindi speakers please correct me if I am wrong!) and put it into the Hindi-Klingon translator in the first result, which is powered by Microsoft Translator™ NMT. The result was “rut mutlha’taH”, which I can tell you is incorrect. I don’t speak or know Klingon but I do speak Spanish and Portuguese, which was also provided by this AI as other options I might be interested in. The option given here for Spanish was “puerto”, which is a valid translation for the word “port” in English. However, “puerto” is not the nautical direction but “harbour”. While English uses “port” with several different meanings, even within the same specialised field (i.e. nautical), Spanish uses “babor” for the nautical direction. A different AI from English to Klingon gave me the Klingon word “poS”, meaning left (or the nautical direction “port”). Several other sources confirmed this, so I believe it is correct (but please don’t crucify me if it is wrong! If it is wrong, it actually makes the point for me even better than if it were correct).
In a case like this where resources between two languages are shown to be unreliable, our job as translators becomes harder because we have to do a sort of relay if possible. In the past example as I was doing research to figure out what to write, I was primarily going from Hindi to English and then English to Klingon OR Klingon to English and then English to Hindi. The resources for English-Klingon and Hindi-English translation are well established but going directly from Klingon to Hindi was problematic.
A quick warning for this situation: reliable direct resources are always superior to this relay process because a translator is a lot less likely to make mistakes if they have to take fewer steps during a translation. In other words, there is a lower chance to introduce error, such as the one we saw above with “port/babor/rut mutlha’taH”. If there is a reliable source for your language pair, use it. This relay process becomes unavoidable when the resources don’t exist or are unreliable. In this case, it can actually be an opportunity for you: you have to do the research so you can create (or help) create a resource to fill this hole in humanity’s knowledge.
Sorry to go off on a bit of a rant! Let’s get back to our list:
3) Language – While languages are generally very robust, it is not always as simple as replacing the specialised vocabulary in one language with an equivalent in another. Sometimes this is because the word doesn’t exist. An example is “Schadenfreude”, which is German for “the joy felt at the misfortune of others”. In English, we don’t have a word that represents this concept. For a translation, you would either have to use the phrase I just mentioned or, because English is fine with using loanwords, simply use the word Schadenfreude (which incidentally is a well established German loanword, just like “kindergarten” and “Zeitgeist”). For the English language, this is acceptable but this doesn’t work for every language. Some languages, like French, are “supposedly” controlled by a language institution (for French, it’s the Académie Française) and are notoriously against borrowing words for other languages. I say “supposedly” because native speakers do what they want – see “la fin de semaine” vs “le weekend”. For these languages, there is often a process of adapting/assimilating the word from another language to follow the rules of French, coming up with their own word or repurposing another word. An example is “ordinateur” in French meaning “computer”. It was coined in 1955 to represent this machine, which was more than just a “calculateur” (calculator) and could be used to “organise” things (Latin: ordinator – one who organises).
3.1) Language equivalency – Another issue that is faced with translating from one language to another is the fact that words or phrases are not always exactly equivalent. While the following isn’t an example of specialised vocabulary, I believe it illustrates points well in general terms: colour.
Different languages classify colours differently. An example is comparing Russian with English: English has one word for blue while Russian has two. While English can represent these two Russian colours with “light blue” and “dark blue” (and therefore achieve equivalency), it is not always as simple as that. Some languages (some examples that come to mind lie within the Arctic circle) only have words for black, white and red. Does it mean they can’t perceive other colours? Absolutely not! Their biology is the same as ours but they have no need for specific words for these colours. In fact the words they have already are used to cover the colours that we may call “blue”, “green”, “yellow” or some other colour. Here is an interesting video that explains this phenomenon:
Another example was in an interview I saw approximately 10 years ago with a native speaker of a very small language in west Africa (unfortunately I was unable to remember the name and find the interview/documentary). When I say “small”, I don’t only mean that there are not many native speakers (I believe it was spoken in a few villages in a very specific region) but also that the language itself is very small in terms of vocabulary. The number that sticks in my head (because it was so unbelievable) was that the language had 200 words in total! The words represented things that were common to their lives, like home/house/hut, food, etc. If they needed to talk about anything that their language could not handle, they would either use French words or codeswitch completely into French. It was an amazing programme and I really wish I could remember the name of the interview/documentary.
Whether this language exists exactly as I have described above, let’s say that we now want to translate a technical manual from English into that language (and only using that language. No French of any kind is allowed in this situation). How can we represent the specialised vocabulary in a language with only 200 words? I doubt they have words for some of the tools that would be needed, let alone some of the other concepts that we will need. This is an example of where concepts or vocabulary are missing from a language and can cause headaches as you try to describe concepts within the constraints of the target language.
That example might have been a bit drastic but I come across this daily translating from German to UK English, two languages that are in the same language family. Governmental structures especially can be difficult because Germany has a federal system but the UK has a unitary system of government. While referring to federal systems in countries that speak English can help to find similar vocabulary, there are some things that still don’t work. A “Bundesstaat” is literally a Federal State or, more normal in English, just a “state”. How about “Freistaat” (literally Free State)? There really isn’t an equivalent in English. The official English names for a lot of these places are usually “Free State of X” but, to me at least, this seems to be a bit opaque. Does that mean there are non-free states? What does that entail exactly? This type of distinction between “free states” and “federal states” doesn’t exist. Even within the “Freistaate” in Germany, there are differences. Bavaria functions differently from Thuringia and Saxony. To make matters more complicated, the meaning has also changed over time. How can we represent the changes from meaning “republic” to “synonymous with Bavaria” to whatever it is supposed to represent today (since the 3 free states in Germany don’t have a common use for the term “Freistaat”)?
As you can see there are many problems and challenges faced when translating specialised vocabulary from one language to another. While it may sound complicated, don’t let it put you off. Translation can be a very rewarding and well-paying job (depending on your language combinations, experience and specialisations). Good luck!