Brackwell College, a fictional further-education college. For three years there have been reports about fumes, headaches and a persistent smell in Workshop 3. Nothing has been done. These five items are everything a review found, and every lens gets all five.
The evidence never changes. Only the question you are asking of it does.
| Lens | Brings into focus | Leaves in shadow | Remedy it implies |
|---|
A framework's blind spot is not a fault that could be patched. It is the direct consequence of the thing that makes it useful, and each of these three buys its clarity with a specific commitment.
Commits you to the idea that there are real interests being thwarted - that the technicians would want something different if the situation were different. That is a strong claim about people's real interests, made by an analyst who is not them, and it has been the standing objection to this account since it was first put.
Strongest where somebody could have acted and did not. Weakest where nobody is in charge of the thing doing the damage.
Commits you to treating power as productive and dispersed rather than held. That is what lets it see a definition or a classification as an exercise of power. It is also why it struggles to say who should be held responsible, or whose interests were served, without importing a vocabulary it has set aside.
Strongest on the machinery that keeps running by itself. Weakest where somebody needs to be answerable.
Commits you to locating the wrong in an epistemic relation: who is believed, and what vocabulary is available to say things with. That makes a specific and often invisible harm nameable. It can also make an institutional problem look like a problem of individual virtue, and a budget line or a performance measure will keep working whatever anybody believes.
Strongest on what happens to a speaker. Weakest on machinery that does not need a speaker at all.
Ferrand Borough Council, also fictional. A three-year backlog of repairs on one estate has been reported to councillors as a resourcing problem. A review establishes four things. The repairs contractor is paid against a single measure, jobs closed per week. A job that is reopened within twenty-eight days is counted as a new job. Residents' calls are logged by the first keyword an operator selects from a fixed list of twelve. The tenants' association has asked in each of the last four years for "damp" to be added to that list, and no decision on the request has ever been recorded.
Three accounts of power pick out three different things in the same case: a decision nobody had to make, a definition that produces the outcome a decision would have had to produce, and a speaker who is not believed. Each brings some of the evidence into focus and leaves the rest in shadow, and the remedy each implies leaves the others running. Which lens you reach for decides what counts as the problem.