Algospeak is rampant, but how respectful is it?

Anyone who hangs out on social media knows this: they know that there is a parallel, encrypted language, the so-called «algospeak», which, using emoticons, asterisks and puns, tries to evade censorship, both the more harmless one of Meta and the heavier one of totalitarian regimes.

And so, to talk about Palestine we use the emoticon of a slice of watermelon, while for Ukraine there is the sunflower. Beyond the little drawings, it is a serious matter, which bypasses the «shadow ban», that shadow ban that tracks the contents spread on the web and makes them invisible to others, without leaving traces, without raising visible walls.

And so one appears on Tiktok cob which alludes to the male sex, while the vagina becomes a lipgloss. The “forbidden” words are avoided, circumventing the problem in a joking tone.

The S’s are replaced by the dollar sign ($): you will then find «le$bica» and «se$$o», a simple and intuitive way to change spelling but not meaning.

But the question is: isn’t the risk of not committing to more respectful language, but rather, if anything, perpetuating offensive language anyway?

AND Marco Di MarcoCEO of Takethe communications agency he developed Inclusive Talksa refined tool based on artificial intelligence, which aims to stimulate collective reflection on the impact of inclusive language, because words are really important and impactful, they know how to kill and cure, it depends on how you use them.

Let’s start immediately with the value and importance attributed to the practice of algospeak: does it make sense? And, what’s the point?
«The tendency to use encrypted codes has always been inherent in the history of language. From generational jargons to cryptography, this practice is the result of the human need to feel part of a group that not everyone can join. Algospeak is the current transposition, with some subcategories that make you think. Its use serves to circumvent two types of censorship: the first, traditional, practiced by totalitarian and illiberal regimes. In this case only the technical way of developing the code changes compared to the past, but the objectives are the same. Affirm your freedom of thought and rebellion. The second, however, completely new, concerns non-ideological, but pseudo politically correct censorship, which is applied by private entities within the platforms they own. Here the matter becomes more complex, because there is no enemy to hide one’s ideas from, but subjective interpretations, transformed into rules by algorithms, still too “incompetent” to understand the context of the sentences on which they intervene. The result is that there are words prohibited by digital censorship or even by preventive self-censorship on the part of the user, terrified of the possibility of being banned from their social life. The term ‘sex’ is a clear example of this.”

But, in general, is it a practice that leads to the adoption of a more or less inclusive language?
«It is certainly a non-inclusive practice, but a necessary one. Freedom of expression is one of the cornerstones of our civilization and censorship is created to be circumvented; there are themes and events of such importance that the free circulation of ideas is well worth the use of encrypted terms.
The risk, however, is that of an excessive use of algospeak, which can lead to the proliferation of an uncontrollable number of coded sublanguages, increasingly limited and difficult to manage even for those who use them. L’over algospeakcould then lead to paradoxical situations, in which the algorithms, which are by definition always self-learning, will begin to censor completely harmless words, the only fault of which is that of being used in place of the prohibited ones”.

What role does artificial intelligence have in today’s language?
«There is a lot of talk about the infinite possibilities of Artificial Intelligence and how it can be useful to us. There is also a lot of discussion about how inclusive AI actually is and on the basis of which parameters it decides (because it is programmed) whether to obscure certain profiles and contents. From our point of view, we tried to understand how it could be useful to the cause of inclusiveness, even if we believe it is essential to underline the great contribution of the “human” work of supervision, analysis and comparison that was necessary to implement the tool. Inclusive Talk works on language, fishing through the language and tracing the origins of words and expressions to identify which ones might be problematic or inappropriate. Because the story we tell about something and the words we use determine our awareness of it and the message we want to get across. The distorted way in which feminicides are reported in the news is a clear example of this. For us it is above all a help to reflect on what prejudices we carry with us (sometimes even unconsciously) and how we can overcome them, expressing ourselves in a more respectful way.”