At the heart of learning theory lies the concept of breakdowns. When trying something for the first time (e.g. riding a bike) you experience breakdowns in your use of the bike (leaning too much to one side will cause you to loose your balance), and we use these breakdowns to experiment with reality. When something goes wrong (a breakdown) we are given an opportunity to learn as the breakdown reminds us of the discrepancy between our actions or expectations and the world.
A breakdown causes a shift of focus. For example, when searching an online database for scientic articles your internet connnection dies (you experience a breakdown in the use of your tool, which is the internet/the database/or likewise). Before your internet connection stopped working, you were not particularly concious about the connection but focused on finding articles in the database. The breakdown of the tool brought about a shift of focus from the finding of articles to the workings of the internet connection (or rather, how to get it working again).
This is related to Heidegger's notion that objects and properties are not inherent in the world, but arise only in an event of breaking down, in which they change from "ready-to-hand" to "present-at-hand" (Winograd and Flores, 1986). When an object such as your internet connection is ready-to-hand, your are using it to fullfil your objective (finding an article) without conscious reflection. Only in the event of a breakdown, your internet connection becomes "present-at-hand", i.e. you conciously reflect on it instead of focussing on finding articles. Before the breakdown you had given no thought to the presence of a TCP/IP protocol stack on your computer, but now it emerges as a very relevant and important part of your activity.
Breakdowns can be used constructively in the design process and are not a negative situations to be avoided. Instead, a breakdown is "a situation of non-obviousness" (Winograd, and Flores, 1986: p.165). The breakdown uncovers an aspect of the design task and is a source of learning.
Fischer, G., Nakakoji, K. and Ostwald, J. (year unknown): A New Understanding of Design and Its Computational Support. Unpublished work, but may be published in the future through University of Colorado