Equivalence languages

Project by Ema

During the prototyping sessions of SI29, in which the theme was File Sharing, i started to think about file contents and their interpretations.

Going more towards the philosiphical and less technical side of file sharing, my focus of research and experimentations was set onto codign languages.

When speaking about Files, you are facing a huge pletora of formats (.extensions). Whens coming to share an image with someone, most of the time the extension of such is not important for the receiver. Our devices can already interpret and convert most of the image file formats, so when sending-sharing an image to someone else, what is important is what gets displayed, not its file extension.

On that, the content of a file cannot always be easily understood, like something that is visual. There are many file types that doesn't deal with pixels but with written texts.

Coding languages for example rely on a very specific vocabulary and thus extensions (.html, .css, .py...). The content of such files is basically text.

So I was interested in the idea of the interpretation and translation of those. I wanted to explore the friction between language, vocabulary, translation, text and instructions, to use them less as what they're just intended to be: a set of instructions that have a functioning output.

So... when i send an image of something to whoever, they will interpret this image based on their cultural background and set of knowledge. When I send a text, they do have to know the language and the vocaboulary of that text in order to understand it, otherwise you're stuck... lost.

Coding languages are languages. Are they really? to what extent? Are they languages or a set of instructions?

-- for the sake of the topic presented here and in which i'm delving into, i would redirect you to their definitions: Language - Instructions

---

Why i titled this project Equivalence Languages? From Wikipedia:

Translation in the field of language is the communication of the meaning of a source-language text by means of an equivalent target-language text (also called 'receptor language'). The English language draws a terminological distinction (which does not exist in every language) between translating, which refers to written texts, and interpreting, which denotes the oral or signed rendering of speech between languages.

Producing an Equivalent I think is a key component when sharing something with someone. In semiotic is a process that always happens in communication.

The idea of Equivalence relates on how computers work. A coding language is always a kind of language, a set of instructions, built upon the common ground of the Binary Language-Code. A coding language is a set that allows us to faster produce the what we want, in a way that is more suitable for us to work with. That should be "easier" to both understand and produce --meaning to speak --> and write in this case-- but as you can imagine this also carries frictions.

The protoype

I selected CSS as the coding language to test this idea. Why? Cause CSS is very based on words rather then numbers. CSS also works in english and in english only.

For the first test i used the CSS code of the Special Issue 28 landing web page, that i took from Git. Because: it was a CSS code that was already done and contained many functions-->vocaboulary-->different words. Being a CSS newby myself, it was easier for me to start with something already done then to start from scratch writing something myself.

• I took the code
• saved it in .txt --> why? cause i didn't want to influence the receiver from the file extension. I wnated the receiver to interpret it mainly as text.
• asked them to translate everything that they could in their native language
• once finished, i sent it to someone else speaking their same language, and asked to translate it back to english

Left: original - Center: italian - Right: back to english from italian

What happened? You can check for yourself!

Many things have been translated correctly back and forth, but since CSS relies on a specific vocaboulary, many things do not work when translated back to english, cause the word used is not the same, eventhough their semantic is the same. Like for example in line 2: overflow becomes eccesso which comes back as excess. Overflow and excess means the same thing, but of course for a computer they can't be interpreted as the same. ahhh those computers...

Up next

I am now continuing the prototype by asking to other people speaking different languages such as Portugese, German, Dutch and Chinese.

To do next: render back those pages to see what is working and what not.

Feedback from Manetta & the class

- Manetta referenced ways to use Python for example, to "interpret" the different words with the same semantic to make them working. This, however, means that the code has to be compiled back into the proper CSS language. What CSS really "wants"/"needs" in order to work.

- Sassy CSS (SCSS) was mentioned as maybe a way on how to code in CSS without using english, but by reading what SCSS actually is, that might seems to not be the case - look here

- The poetic aspect of this kind of project emerged, and Manetta suggested that of course it would be cool to come up with the starting text myself. To delve into the poetic aspect of coding caught more my interest, since I feel it gets closer to the interpretation of those coding languages, which as I previously said, they have this friction between language and instructions, thus to to better frame them by taking a step back towards the language itself.

Going down the rabbit hole of its vocaboulary and poetic is for me a path that i'd like to take!

By doing some research i cam across this publication, called Code Poems. An insightful project carried out by Ethos Lab IT Univerisity of Copenhagen.

Preface of the text:

Code poetry is a form of experimental writing and coding. Bothwriting and coding pay attention to the poetics of syntax, grammar, and punctuations, considering natural and computer languages as playful and aesthetic materials. This mix of languages is called “codework”, a genre coined by poet-theorist Alan Sondeim which he describes as "the computer stirring into the text, and the text stirring the computer” (2001). Though textuality is an important element in language that contains meaning yet it requires interpretation, I want to point to the aspect of performativity in computer code and natural languages. One might draw upon John Langshaw Austin’s Speech-Act theory in thinking about the things that you can do with performative speech, in which the utterance of words carries inherent actions beyond simply making statements. However, code execution is different from the speech-act, because computer code, especially high-level programming language, is written for both humans and machines. When a piece of source code is executed, the computer is doing something immediately, for example, the translation of high-level code to binary code, or to display certain visual and textual materials on a web page. In this way, there are different kinds of readers beyond the human.

Other referencies: CSS Reference by W3 school - CSS CheatSheet vocaboulary