If you’re here and reading this you’re probably already aware of Guerrilla Translation and our focus on careful, deliberate, artisan translation. We don’t use machine translation and all our translations (and original content pieces) are edited by another member. We take whatever time and energy is needed to sensitively and accurately translate the meaning, rather than just the words, of whatever we are working on. This level of engagement has, on occasion, presented something of an ethical dilemma, because when a text uses language, terminology or concepts that are problematic, offensive or outdated we have to ask ourselves, what is the responsible way to translate it? We would obviously flat out refuse to translate anything that contained outright slurs or promoted any form of intolerance, but this is an issue that often runs deeper than we think, and occasionally requires a sensitive touch. All of us have, at some point, unwittingly used language or ideas that are derived from ignorance or intolerance, and we all have to work hard to eliminate and modify behaviour and language that doesn’t align with the values we aim to uphold. This is also something that evolves and shifts over time — what was OK twenty years ago might not be now, what is accepted today might not be in the future — and it takes humility to recognise and accept the role we all potentially play, as well as the need to adapt and change our behaviours.
This question has generated a lot of discussion among GMC members, and we’ve decided to write this article for two reasons. Firstly, we want to clearly outline our position on this matter and how any potential client can expect such issues to be handled in the event that they come up. Secondly, we want to give a few examples of overarching concepts which crop up in English and Spanish texts (and probably other languages as well), with a view to creating separate articles to address more language-specific issues that can serve as a loose toolkit or style guide for anyone who might be interested.
It must be stressed that we will never modify the structure or meaning of a text without the author’s permission. A note will be made on the final copy of a translation highlighting any light modifications, and any large-scale issues will always be addressed with direct input from clients, publishers or authors.
The Ethics of Translation and Our Role
This discussion began while translating an article from Spanish to English that hinged one of its arguments on a distinction between the “civilised world” and the “developing world”, as well as repeated references to the “First/Second/Third World”. To deny that such a distinction exists would be wrong and extremely harmful, but the outdated language and idea raised eyebrows for both the translator and the editor. Leaving such terms in place, directly translated, was not an option, and so more appropriate terms — “global north/south”, or “majority/minority world” — were judiciously employed with minimal impact on the fabric of the text. We were certain that the author would not wish to knowingly promote such ideas, and we would make similar assumptions of anyone who hired Guerrilla Translators given that we wear our political and ethical convictions on our sleeve. We are also aware that such ideas vary in their evolution between cultures and languages, and had no interest in aggressively confronting the author on the subject, but instead chose to gently steer the language in a more inclusive direction, with a note in the final correspondence highlighting the changes made.
The questions this incident raised were: how much ethical responsibility and freedom do we have here? Where is the line between tweaking/updating language and changing meaning? We landed on the following:
We refuse to perpetuate or recreate language and ideas which reinforce discrimination, prejudice or hierarchy in any form.
We understand that this may aspire to an unattainable standard of “wokeness”, given how ingrained a lot of this language and behaviour is. It is also true that superficial language, symbols and gestures can easily be exploited for virtue signalling: the visible but hollow lip service to progressive ideas which disguises true beliefs and behaviours (think corporations dressing up in rainbow flags during Pride month, or fossil fuel companies shamelessly greenwashing their public images). This may all be true, but we prefer to aim high, and we believe that, given its ubiquity, language plays a big part in regenerating harmful ideas. Translators sit at a cultural bottleneck through which all content has to pass on the way from one language to another, and we therefore have great power and responsibility. If we shrugged with indifference and translated these ideas directly, we would be helping to spread them across borders, potentially to hundreds or even thousands of readers. We don’t want that. As Taiki Waititi puts it, “What if everyone stopped giving to racism? What kind of future would that be for our children?”
Something that cannot be ignored in this discussion is the impact of Artificial Intelligence and machine learning. CAT (Computer Assisted Translation) tools are used extensively in the translation industry, and unquestioningly recreate the exact prejudices held by humans. When a text is machine translated, objectionable ideas are simply passed from one language to another, thus perpetuating their usage and the segregated, ignorant worldview they uphold. This is, in fact, one of the central ethical problems AI presents — machine learning is trained to recreate the logic of humans. It is incapable of critical thought, and will always recreate any bias (there is always bias) in the logic it was trained on, with no second thought. Any underpaid, overworked human post-editor of the text would not have the time or resources to interrogate such ideas, let alone construct new ones to take their place. As humans, we are able to break the cycle and stop these ideas from spreading through creative, culturally sensitive modifications. A machine, no matter how intelligent, will never be able to do that.
A Far from Exhaustive List of Examples
The following are some examples of commonly used or accepted ideas which we would change, with an explanation of how they would be handled. This list is far from complete, but it gives an idea of our level of involvement with every text that crosses our inbox.
First/Second/Third World, Developed/Developing/Undeveloped World, Civilised/Uncivilised World
The one that started it all. These distinctions are problematic because they exist to establish a hierarchy, and then place former and current colonial powers at the top. They are rooted in the assumption that the Western model of development is superior and that any alternatives are automatically inferior. We cannot deny that a global divide exists, but to define it in these terms reinforces the binary, hierarchical logic of European colonialism and white supremacy. As a rule, we would refer instead to global north/south, or majority/minority world, but would adapt depending on the specifics of the topic at hand (i.e. reference to HDI figures, quantifiable aspects of infrastructure development and so on).
We should point out that a lot of this language was originally coined by Marxist theorists, many of them in the Global South, and it was closely connected to the “aligned/non-aligned” divide during the Cold War. While we can respect that some writers continue to use these terms in an academic context, the average reader will probably not be aware of this usage, and will likely interpret them as reinforcing, rather than questioning, the established order.
The use of the LGBTQIA+ acronym is not in itself problematic, far from it. It can, however, be manipulated to exclude certain groups, a common example being the TERF preference for “LGB”. This example weaponises the acronym in a malicious and extreme way, but authors who are less aware of this topic may, completely inadvertently, stray into similar territory. By default we use LGBTQIA+, incorporating any other elements of the acronym which are included in the source text or which are relevant to the topic, region and so on.
Dog Whistle Language
This refers to language that, on the surface, seems inoffensive but in fact signals regressive ideologies, and is used heavily by populist politicians. It varies widely among times, cultures, nations and languages, and has too many examples to list here, meaning that it cannot usually be substituted very easily. It can also be difficult to identify and adjust, because it can encompass so many different ideas. Any text that relied heavily on dog whistle language would likely not make it into our inbox, given that it normally masks a deeply conservative worldview, but these terms can easily creep into everyday usage or become accepted.
A perennial, commonplace example is the phrase “family values”: a seemingly innocuous term (most people experience family in one form or another) but one which almost always adheres to a very narrow definition of “family” (usually Christian, nuclear, heterosexual, monogamous, middle-class, conservative, patriotic, childbearing etc) and is used to exclude any non-normative family structures or to dismiss forms of culture that question or deviate from what is considered “normal”. Such a term reinforces an institutionally embedded hierarchy, and politicians (for example) who claim to speak in the interests of “working families” are usually tugging on this conservative, regressive strand of thought.
Often in texts we work with these ideas are being questioned, or at least not championed, and a simple solution may be to place the language in quotations, or add “so called”, to indicate awareness and cut through any doublespeak. Evidently, extensive use of dog whistle language, however accidental, would warrant consultation with original authors. This Wikipedia article contains many more examples and explanations.
The term “aporophobia” was coined by Adela Cortina in the 1990s to describe the fear and rejection of poor people and poverty. This fear drove various cultural migrations such as white flight in the USA, and animates a lot of the widespread gentrification and displacement in urban areas today. The dichotomy of “rich vs poor” is problematic because it establishes a person, territory or community’s worth on a solely economic basis, but the binary can also mask a lot of subconscious racial and class-based bias (search “poverty” on Google images to see this for yourself). As an alternative to this segregated, messy binary we would aim for precision in terms of the aspect of wealth or poverty being discussed: low/high income households, low/high paid jobs, advantaged/disadvantaged areas/regions and so on.
In handling any issue such as these, we would always take the stance of working with and never against the client or original author. We would take it as an opportunity to be humble, to educate and be educated, and the end result will always be far superior to any rushed, machine-translated text.