In their study on seeing as a situated activity in airport control rooms, Goodwin and Goodwin[14] propose to do some imaginative work on how baggage loaders identify and select aeroplanes. We propose to start looking at key work in a similar way. Consider someone, who arrives for the first time in the concierge station and has to find a series of keys in the smaller cupboards: for example, the keys for flat 4 on landing 16, the key for flat 1 on landing 24, and the keys for flat 2 on landing 25 in block 183; say to close down the water supply in these flats. The only thing s/he has at hand is a list with three numbers; 183 16/4, 183 24/1 and 183 25/2. In the key cupboard s/he can see around 20 keys of similar shapes, materials, colours and sizes. S/he does not distinguish different keys from each other, since for her/him they all look the same, although some of them might be differently shaped. There are several ways of getting access to a flat. S/he could, for example, take all 20 keys from the cupboard, take the keys into the building, identify the right block, landing and finally the flats on the list, and try them out one after the other. But there is of course a simpler way of finding the right keys. Keys are hanging on hooks; each key is related to its lock by an inscription on a tag. To find the right keys our stranger will have to situate the keys within a 'relevant web of meaning'[15]. S/he he does so by looking at the list, reading the numbers of the keys and by identifying the right key in the cupboard. When we consider that access is more than just selecting a key from the station cupboard, we get a first insight into what key work in Red Road involves. Our stranger will have to bring the selected key to the corresponding lock in the building and open the door of the flat. From this idealised observation we learn at least three things:

Firstly, access consists not simply of supporting objects (cupboards, hooks, keys, doors, locks, valves), inscriptions (tags, numbers, key shapes), building parts (block, landings, flats) and users (residents, concierges, builders, housing managers, housing officers), but involves the practicalities that associate one with the other. Secondly, access is not achieved by a single selection of a key, but is shaped in and through larger courses of action. We must expect that keys are put into question (involving, for example, changing, re-selecting, trying and loosing keys). Finally, webs of meaning in a transitional place such as the partially occupied high-rise under clearance offer various instances of non-coherence and trouble; from wrongly tagged keys to missing locks and doors. We can expect that concierges in Red Road have always a little more to do, for access to become possible.