In our Learning Analytics framework we include 'Competences' as a vital dimension for reaping benefits from data analysis. This includes interpretative competences such as understanding visual semantics encapsulated in infographs or other visualisation interfaces. In a world of data, visualisations are a commonly used tool to ease access to complex information, but  they also require specific competences to exploit this information. As humans we acquire visual semantics in a similar way to first language acquisition. In contrast to aural or acoustic access, the visual access to information dominates our perception of the world as well as our memories. It is quite important to note that reading too is a visual medium - an abstraction and representation of acoustic signals. This can be easily understood when looking at hieroglyphs and other pictographic forms of writing. The field of visual linguistics is, somewhat surprisingly,  rather underdeveloped if it exists at all. Image semantics like aesthetics, functional and emotional content of visualisations are mainly covered in Design studies and the Fine Arts. However, cognitive and emotional studies of visual languages are quite sparse, especially with relation to technology and the cognitive load it carries. One thing, though, is clear: pictures speak a language, and our capability to understand and interpret information in images follows rules that we acquire alongside our linguistic abilities. These rules are also culturally bound; e.g. the meaning of colours, etc.

This excellent 8-part series of posts by Carla Casilli explains how the visualisation of information has developed into a new visual language. A very stimulating read, to which I'd add that the 'newness' in visualisation languages occurs in the design. This is because of the semantics that gets packaged into a single visualisation. It is the amount of information encapsulated that makes visualisations a powerful and compact tool for transporting meaning. At the same time, however, they carry bias towards priority information, and the interface by which it is presented can pose serious barriers to making the meaning intelligible. This poses an ongoing challenge to visualisation designers as well as to users who still have to learn the language first.