When the apprenticeship model breaks down - or may break down. New-comer's periphery never evolves because participation doesn't develop/increase/become more complex.

Eg, 1st and 2nd yr grad students working on studies involving coding analysis often get assigned coding tasks because it can be done with relatively little experience. However this can result in them not becoming a full member of the research team until a new student comes in, is trained for the coding job and the older student moves on to a new task.

Oscar shared his group's observation of people buying a metrocard last week. One of the machines was not accepting bills and that changed the expected experience of many subway riders who used that machine. Using the LW language, riders interacted with the machine in a renegotiation of what it meant to go through the process of buying a metrocard when added decisions (continue or end) was involved. We can remember that different riders ended up with different results (meanings) and actions. Some left the machine without doing the extra steps while some continued on in a new interaction with the machine.