Being pigeonholed is sometimes a good thing – especially when it comes to categorizing texts that need to be translated. In the world of translation we talk about text classification, where every type of text that a business produces and needs to get translated is classified using certain criteria. That way, the ideal service level can be found for each category in order to ensure top-quality results and save money.
There’s no place for one-size-fits-all solutions in the translation industry. Every text comes with its own challenges and needs to be analysed in detail before the translation begins to establish which service level and translation workflow are best suited for translating it.
What do we mean by classifying texts?
Classifying texts means dividing them into various categories on the basis of specific linguistic, communicative and content-related criteria, such as field, purpose, style and target audience. Some examples of text types are legal documents, technical instructions and manuals, advertising texts, literary texts and scientific articles.
How does classifying texts relate to translations?
Businesses usually want to get the same types of text translated, whether it’s technical documentation, software texts, press releases, internal communications, marketing brochures or contracts. If they work alongside their translation agency to classify their texts, the right service level for each type of text can be identified – and they can also factor in the risk that would arise in the event of an incorrect or poor translation. For example, they might decide that internal communications, and technical documentation as far as possible, will always be translated using machine translation with post-editing. By contrast, marketing texts might be translated using human translation without review, and for contracts they might opt for human translation with a review by a second translator. This is one potential scenario where classifying texts can be very helpful: the business knows which services they’re getting, and the translation agency knows the requirements for the specific types of text. So the texts are being pigeonholed – but for good reason.
How are texts classified?
Ideally, the business will send the translation agency examples of the texts they usually need to get translated. The agency can then take a close look at the texts to analyse them and classify them on the basis of the criteria mentioned above. Most translation agencies are very experienced in terms of knowing which service levels are best suited for which types of text.And their team can also tell pretty quickly which of their client’s texts are suitable for machine translation with post-editing.On the basis of their experience and the analysis they’ve conducted, the agency can then give the client a detailed breakdown of the ideal service levels for each text type.
What are the benefits of classifying texts?
This analysis of text types can benefit clients in lots of ways, especially when these texts need to be translated. These include:
- Better consistency and quality: Sorting texts into categories allows clients to specify the style and terminology for particular types of text, so translators know exactly what they need to do. For example, technical manuals and instructions require clear and precise language, while there’s more freedom for creativity when translating advertising texts. Clients can also create style guides for each text type for the translators to adhere to, in order to increase consistency within the text types.
- Faster project delivery: Sorting texts into categories allows translation projects to be assigned and delivered more quickly. Translators with specialist expertise in particular fields (e.g. those with a technical background or experience in marketing) can be assigned to the relevant projects, meaning less research is needed. This accelerates the translation process itself, as well as improving the precision and accuracy of the translations.
- Better use of resources: Sorting texts into categories allows specific translation memories and term bases to be created for each text type. This ensures that each database only has relevant entries for each text type, which means terminology doesn’t get mixed up, and overall the translations become tidier.
- Optimized project management: Sorting texts into categories allows clients and agencies to decide in advance whether complex review or proofreading workflows are needed for the translation. Many clients prefer to have a second translator carry out a review on legal texts, or to get one of their own reviewers to take a final look at marketing texts – the ability to plan ahead means the translation process can be optimized, projects can be scheduled realistically and there’s a better chance of meeting deadlines.
How does classifying texts help the translation workflow?
The above benefits of classifying texts go hand in hand with their practical benefits for translation. Here are some examples to illustrate how it can optimize the translation workflow:
- Better preparation for translators: If the translator knows what type of text they’re dealing with, they can make specific preparations, read up on the subject and research specialist terminology more effectively, and look at the relevant style guide. This in-depth preparation inevitably results in a faster, more efficient translation process and a higher-quality translation. Plus, if text classification means the translator can familiarize themselves with the requirements and the subject matter in advance, it’s less likely that they’ll have any queries when working on the text.
If it’s a technical, medical or software text – areas in which MEINRAD specializes – it’s important to identify that right at the start and to put the text in the appropriate category. That way, the right workflows can be determined and the right translator can be chosen to ensure the best possible results. For medical translations, for example, MEINRAD only uses translators who specialize in the medical sector and are familiar with the relevant technology. Likewise, for technical translations we only use translators with the required experience in the respective specialist field, and for software translations our translators must know all about the requirements of software texts and be able to use the appropriate language. - More automation for processes: Classifying texts means templates can be created in translation memory systems and CAT tools for specific types of text, which saves time by ensuring the right resources are automatically attached to projects. And interfaces can also be set up for some text types to make life easier for everyone involved.
- Use of machine translation: If you classify your texts, you can decide in advance whether they’re suitable for machine translation – because it isn’t always the right choice. Texts with simple language and short sentences, such as technical documentation, legal texts and informal messages, are usually ideal for MT, while creative texts with wordplay and playful language (like marketing texts) tend to be less well suited.
- Quality assurance: You can tailor the quality assurance steps for the respective text type: for example, legal texts and technical manuals might need to meet different, more stringent criteria than marketing texts, which are usually translated more freely.
Summary
Classifying texts is a key tool that can considerably accelerate the translation processes and give you much more consistent, higher-quality results. Systematic division of texts into categories allows the ideal service level to be identified and means workflows can be planned more effectively. And it improves resource management, giving you tidier translation memories and term bases in the long run.
Main image © Adobe Stock
