One of the core concepts when discussing a competency model is the ability to compare the current state, or what we have, with the predicted or future version of the same role. In other words, to see the gaps, the missing components. If the model is presented in a descriptive format outlining the required abilities and competencies for the role, finding differences between present and future is not that complicated. But if the model changes its format from a text-based to a skill-based version, it becomes challenging. How to comfortably discuss something that is invisible?

In Pape Satàn aleppe Umberto Eco—yes, I am still getting a lot of inspiration from this book—has a piece titled "About an unread book" in which he brings the attention to a French author, Pierre Bayard and his book Comment parler des livres que l'on n'a pas lus?  If we agree that libraries hold millions of books and more books are published every year, even if an avid reader consumes on book per day, that will be only 365 books per year. A fraction of the available volumes. Of course, the idea to read everything seems odd. Even if you specialize in a narrow discipline, it is still hard to keep up. Bayard admits to not reading Ulysses (Joyce), but it doesn't stop him from discussing the masterpiece and to make references to Homer's Odyssey (by the way, he hasn't read it either). According to Bayard, "it is the relationship with other books that tell us more than we will get from reading it." Eco concludes that the most interesting theory Bayard brings in is the notion that we forget anyway most of the books that we read. What is left in our minds are some sort of virtual model, not exactly describing what the book was about, but the impressions that it left on us. Bayard, who apart from being a literature scholar is also a psychoanalyst, plays with an idea of a school in which students would discuss only books they haven't read. In this self-discovery process, students would "invent" books not necessary to read. He concludes that perhaps there is a "better way" of reading (if you are reading) to encode the contents, but to Eco's surprise, the theory of not reading and discussing is in such a strong opposition to the fact that Bayard's analysis is quite thorough and as such, it suggests that Bayard hasn't read the book he wrote.

Now, I haven't read Bayard's book, but I am writing about it based on what I found in Eco's book. I am not sure how comfortable I am talking about a book that I haven't read. Still, every day we discuss topics that are not always our domain. The idea of looking for clues is based on limited variables. Design often challenges us to come up with a full story based only on available fragments. Archeology or anthropology does the same. How often you remember something, a passage or a concept, but can't place it. For this post, I wanted to refer to a post I wrote several months ago, and I couldn't find it. What I do remember though is what it was about: the idea of seeing something based on references existing in other sources. Astronomers often use this method.

Last week, I came across Picular, a search engine allowing you to look for colours based on pretty much any query that is entered into the search box. Look for sad or winter or relaxed. It is quite impressive how the system suggests colours based on the association with the keyword. However, it doesn't go deep into what is the relationship between the keyword and the colour. It looks only for a dominant colour in a photograph. I wonder if it checks as well the tags that this photo may carry. Also, in the right side panel, what is the relationship with the other colours presented? What if we swap the colours for skills? Again, a Competency Model can be based on a description. From this description, we can extract skills. We already have this functionality in TeamFit's advanced search. A resume or a role outline can be uploaded, and a set of skills and associated roles will be extracted. We are getting quite comfortable while discussing hidden relationships. The question is, how accurate we can be in predicting the missing elements or suggesting for that matter. Part of a good Competency Model is the action plan showing the path to improvement. I think this is the critical component - knowing existing competencies is great, understanding the gaps is even better, but the ultimate piece of information is the learning path.

We are in the process of collecting thought related to the design of Competency Models. If you have 30 minutes to spare, here is the link to our survey. We will share the results in a future blog.