What do rules of scientific visualization tell us?

At our Paris station, we moved from scientific objects to scientific collections and from museums to laboratories. The themes of our visits to the National Institute of Agronomic Research (INRA) and to the Herbarium started from two specific questions: What are the ways collections work as “scientific objects of enquiry” in current laboratories? What are the epistemic differences between building and investigating a model, and building and investigating a collection of data? In the introductory lecture, B. Strasser pointed out that during the last two centuries physics and natural sciences frequently exchanged methodologies in both directions. In a way, the visualization of data seems to be a mediating point. Both during lectures dedicated to genomic researches at INRA and during talks about the way botanists work at the Herbarium, visualizations were present in great number. In one case, they were mainly visual elaborations of great amounts of statistical data; in the other, maps of distributions of different species and/or visual tools to aid in morphological identification. In both cases, the visualizations served to reduce great amounts of data to synthetic maps that are univocal presentations, appropriate to sharing knowledge. Roughly, we can say that visualization itself is becoming the (one) model to present results on collective enquiry.
Visualizations and knowledge-sharing work thanks to substantial efforts in standardization and in “creating grammars”. That is why, for instance, visualizations of processes in genome mapping could directly reveal machinery biases by the construction of statistical curves of distribution (and consequently visually correct them). Standardization and common grammars are also the premises to quickly create world communities talking the same language, and to present new suitable rules in interpreting data as discoveries themselves, that are rapidly adopted by the researchers.
In this way, visual grammars are today seen by scientific communities not as “objects”, but as successful tools to investigate objects. They are an ideal medium to create and disseminate objectified knowledge. Their organization is about the way things are discovered. On the other hand, a historian is more used to see visual grammars not as tools, but as very powerful objects to investigate, able to reveal how scientific knowledge organizes itself in different periods. Rules of visualization have long fragmentary histories, which date back at least to the seventeenth century, and touch, with uneven successes, different fields of knowledge. As every polarizing split, the one presented above could be considered coarse. But, in the end, what do rules of scientific visualization tell us?