Translation between software designers and users

W qlen asked how ~:hey create fo r an unfamiliar domain, software designers typically answer: "Either I have to learn enough about what the users do to be able to tell them what they want, or' they have to learn enough about computers to tell me." Several assumptions underl ie this response. One is that software design is concerned primarily with the user's task, in isolation f rom the workplace context in which it is; performed. Another is that e i ther the engineer or the user will have the opportuni ty, resources, ;(nowledge, and skills to learn enough about the other's domain to engage in useful dialogue. We have developed the not ion Of translation between designers and users to address this second assumption. PD requires effective communication between individuals with d i f ferent kinds o f training, d i f ferent goals, d i f ferent languages, and d i f ferent workplace cultures. Users cannot be expected to describe their work and needs in the language and f rom the point of view of an engineer. Conversely, engineers seldom have an intuit ive grasp of their users' working life, or of the environment in which a product will be used. Some kind of two-way translation is needed bet~Neen the user's domain and the software designer's. Here we explore the role that can be played by a third party who serves as translal:or between the users and the engineers. For dialogue between a user and an engineer to be meaningful, the discourse must include model5 that both parties can understand. These models are metaphors which act as a bridge between one domain and another [2]. A good translator can create metaphors that are meaningful both to users and to engineers. 13y enabling the user to part icipate in design, the tran¢.;lator may also be able to overcome t i le user's reluctance to confide in a developer and t rust him or her to turn privileged information into a useful tool. we have Inad the oppor tun i ty to work both with and as translators. Our study o f translation in PD focuses on our own experiences, as well as on the experiences of participants in a workshop we recently gave [5,6]. We are performing retrospective analyses o f several design ef for ts that were somewhat extreme in their need for translation. It seems unlikely that these ef for ts would have succeeded wi thout the translating services o f a person with work experience both in the user's domain and in software design. The users in our case studies are as diverse as physicists, English teachers [3,4[, psychiatric nurses, and CAD engineers [1].