Languages As Data Jewellery

Languages As Data Jewellery

Data jewellery is often used to physically represent personal data, in our case: the languages a person speaks. The encoded data consists of the languages, their corresponding language families and the person’s proficiency for each language.

How It Started

This work started when I was a research intern at the Innovis Lab at the University of Calgary, Canada. Inspired by the internationality of research labs and conferences and the creative uses of physical visualizations, also called data physicalizations, my two collaborators and I decided to visualize the different languages we speak as decorative jewellery.

At first, we started by sketching different versions of how to represent language data. The first sketches were more traditional visualizations and evolved into designs for physical visualizations, more specifically jewellery.

 

The Necklace

Our goal was to design a gender neutral piece of jewellery. The pendant encodes several aspects of personal data, such as the number of spoken languages, the family of each language, and the level of proficiency in each language.

For the overall design of the pendant, our intention was to use a gender neutral appearance. The design and layout of the single elements allow multiple ways of wearing the jewellery. It can be worn as a bracelet, a short or long necklace, or as a keychain. Furthermore, our goal was to create an abstract pendant that does not immediately reveal the encoded data, which makes it incomprehensible for people who do not know its real purpose.

Language Encoding

Every spoken language has a classification which consists of multiple levels of language families. Each family includes a number of languages which have the same origin. We decided to focus on the most commonly spoken languages in Canada.

We chose to use color as a visual variable for representing the spoken languages. As humans can only distinguish and name a very small number of colors and there are far more languages than could be represented by color, we chose to take advantage of two visual variables for encoding the spoken languages: We used shape as a visual variable for representing the language family and color for demonstrating languages inside a family. Therefore, it was possible to use of the same colors in combination with each shape which leads to a higher number of representable languages.

We decided on different shapes to represent four language families including Indo-European (circle), Sino-Tibetan (triangle), Afro-Asiatic (square), and Japonic (pentagon) which are common languages among the most spoken languages in Canada.

Shapes to represent four language families: Indo-European (circle), Sino-Tibetan (triangle), Afro-Asiatic (square), and Japonic (pentagon)

We chose to use color as a visual variable for representing the spoken languages. As humans can only distinguish and name a very small number of colors and there are far more languages than could be represented by color, we chose to take advantage of two visual variables for encoding the spoken languages: We used shape as a visual variable for representing the language family and color for demonstrating languages inside a family. Therefore, it was possible to use of the same colors in combination with each shape which leads to a higher number of representable languages.

We decided on different shapes to represent four language families including Indo-European (circle), Sino-Tibetan (triangle), Afro-Asiatic (square), and Japonic (pentagon) which are common languages among the most spoken languages in Canada. 

In order to easily differentiate languages in each family, we utilized the following nine different colors: Yellow, orange, red, blue, green, pink, purple, aqua, white/black.

Proficiency Encoding

People have different levels of proficiency in the languages they can speak. We considered three levels of language proficiency: basic knowledge, intermediate knowledge, and fluent. The level is encoded by the number of shapes. The more shapes the more advanced a person is, e.g., three triangles inside each other show that either the person is fluent in the language, or that it is their native language. Two nested triangles demonstrate intermediate level language skills, and a single shape shows basic level of language proficiency.

The number of shapes represent the language proficiency - one shape: basic knowledge, two shapes: intermediate knowledge, three shapes: fluent or native language.

Usage

Below are three examples of the necklace. The first one is my own: I am a native German speaker, I am fluent in English, and I have basic knowledge (I don’t dare to say intermediate) in French and Spanish. Since all of these languages are European, the pendant only consists of circles. The middle example belongs to a person whose native language is Mandarin and is fluent in English. Let’s name them Reese. They also have intermediate knowledge of Spanish and know basic Japanese. The last example represents the data of Jordan who knows three languages. Their native language is Arabic and they’re also fluent in English. Their third language on an intermediate level is Persian.

We thought that the necklace could be used as an icebreaker at events and could engage people in a conversation. The conversation could either be about the languages or about the pendant itself if not everybody knows about the data it encodes. Another use is to communicate shared spoken languages. If I were to meet Jordan and Elliot at an event, I could immediately see what languages they speak and which one to chose to communicate. Most likely this would be English. However, if I would like to improve my Spanish, I could try to learn from Elliott.

Lessons Learned

Overall, designing a piece of data jewellery was a fun little project and I personally like it a lot. However, not all our decisions were ideal. For instance, using color and shape as a double encoding for the language and the language family results in some issues. Since the colors seem to be more dominant than the shapes for the pendants, they suggest a connection between two languages of the same color, such as Mandarin and Spanish, which does not really exist. Instead, the shapes represent the connection between languages of the same family which is unfortunately not as prominent as the color. A possible solution would be to use a certain color scheme for each language family, such as blues for the Sino-Tibetan family and reds for the Afro-Asiatic family. Thus, colors would not be used twice.

Furthermore, the colors for the languages are hard to remember and a legend is always needed. This was done partially on purpose to first, avoid associating a specific country with a specific language (English is only associated with the UK and not the US or Canada) and second, to not perpetuate stereotypes about countries.

The last point is not a lesson but more a missed opportunity. If 3D printing had been as common and as feasible as it is today, I would’ve printed some of the examples and maybe even played around with different sizes, encondings, and colors to optimize the pendant as a piece of jewellery.

Further Information

The project was part of a small research publication at a Personal Visualization workshop at the IEEE VIS conference in 2015.

Tamara Flemisch, Fateme Rajabiyazdi, Mona Hosseinkhani Loorak, and Sheelagh Carpendale. 2015. NeckLan: Language as Jewellery. In Electronic Proceedings of the IEEE VIS 2015 Workshop on Personal Visualization: Exploring Data in Everyday Life. http://www.vis4me.com/personalvis15/