Technical subjects can be approached at different levels of sophistication. Banana coding is a visual cue to difficulty of a text: both the level of knowledge that is imparted, and the level of knowledge needed to understand it. Three levels are defined: Green banana for the new reader; yellow banana for the riper reader needing technical insight; and brown banana for the experienced user who need deep technical detail.
The job of a green banana text is to tell you what the subject is and what are the important differentia in the subject. For example, it might show a set of road signs for different classes of bend – and by grouping together the signs for single bend, double bend and series of bends, it emphasises the features that differentiate the signs.
A yellow banana text is there to provided the logic that explains why thing is the way it is and provide additional detail. That for road signs might explain that the bend signs warn people to slow down before they get to the bend, especially where the sharpness of the bend is not evident. The text will point to additional sign such as chevrons on the bend itself, advisory notices about maximum speed, and additional warnings such as “adverse camber” or “mud on the road” which warn drivers to slow down more.
Brown banana gives the deep technical detail needed by specialist to use the information effectively. For example, a road sign text might list the criteria used to decide whether to erect a sign, such as the sharpness of the bend, the angle turned through, how wide and how busy the road is, and so on.
The original banana coded site dated from the early 2000’s, and was an Intranet site for ISO 10303 (better known as STEP – the STandard for Exchange of Product data). Green banana text got users to an introduction to STEP, the names of the main product areas covered, and some case studies of the business benefits. Its main aim was to provide an easy route for managers and engineers to find out what STEP was about and to see whether it was relevant to their workplace. Yellow banana was aimed at a more technical group, who already had some exposure to STEP, and needed to know whether it would do the job they needed, whereas brown banana was for those who developed the standard or implemented it in design software.
The site solved two sorts of problem. Firstly, how to distribute technical information across a large company without creating a barrier to new users. Secondly, there were a large number of resources published through the site (of the order of thousands of documents) and banana coding hid that level of detail to avoid overwhelming the new user while giving the experienced user easy routes to new or significant documents – at the time, search capabilities for such sites were limited, and navigation was mostly through index pages of hyperlinks.
I do not know of any objective criteria used to grade text difficulty in historical Knowledge Management approaches. These noted the existence of experts, and tended to focus on the elicitation and formulation of expert knowledge, without mentioning any direct metrics on the complexity of the knowledge so captured. As an ad hoc suggestion I propose the following criteria:
Green Banana – facts acquired in schooling to the age of 16 (GCSE level) such as could be tested by multiple choice questions, and of no more quantity than a reader could cover in half a day’s study – say 50 facts;
Yellow Banana – facts and techniques as acquired in preparation for university (A-level, Baccalaureate), with an individual topic taking a few minutes to study, and possibly relying on an average of half a dozen more yellow banana texts – perhaps like the opening sections of the more technical Wikipedia articles;
Brown Banana – facts and techniques requiring university level study, such as the technical articles in Wikipedia on engineering, computing or maths, which require some deeper understanding of a topic and its relations to other topics in the same domain – easy reading only after five or so years studying the domain.