In the 1980’s a wave of research in cognitive science examined the role of preconceptions, in the form of naive theories and mental models, in shaping our understanding of the world (Gentner & Stevens, 1983, McClosky 1983, ). Research on naïve physics, on folk theories of psychology, on mental models , convincingly showed that people do not face a new reality from scratch but use their pre-existing knowledge to approach it intelligently. People tend to build naive theories of new situations and objects spontaneously and generally their theories emerge from similarity between the new situation and some situation experienced in the past. Naïve models can be more or less correct or close to an expert theory, but in all cases they offer a frame of interpretation of events and behaviours that is used to apprehend new phenomena. Research on naïve theories, which coincided with the advent of user centered design, was exceptionally influential in defining one of the main tenets of the discipline of HCI : capturing and understanding the naïve models of users is an absolute prerequisite to design a new artefact that meets users’ needs and expectations.
Understanding naïve functioning models of users is essential in order to :
respond to the expectations of the user ;
design an interface that triggers analogue recognition and action ;
ensure a comfortable transition from an old experience to a new one.
The shopping carts, chat rooms, wastebaskets, scissors, that populate our screen environment are the direct outcome of that wave of esearch. From individual models to cultural representations Most of the research on mental models at the time was couched in a cognitive science paradigm and was essentially interested in the individual process of model construction and revision (Johnson Liard 1981).
This research was influenced on the one hand by Piaget’s constructivist theory of how children develop new concepts through their individual direct experience with physical phenomena, and on the other by models of memory processing and information processing. As a consequence of this individualistic paradigm, HCI inherited an individualistic process of enquiry : individuals are interrogated and observed in order to capture the models and theories they entertain regarding the objects and concepts they manipulate during the execution of some task or activity. The most general models are identified and used as a basis for designing or redesigning new artefacts or features. Another current much less known by the HCI community has been investigating peoples’ beliefs and representations but from a social and cultural perspective. Anthropologists have been studying how communities build shared representations of events, situations and artefacts. I have come to the conclusion that this paradigm has much to bring to our discipline and in particular to understand the phenomena of diffusion of innovations.
The construction of a cultural model of AIDS in Haiti Paul Farmer an outstanding physician and researcher in medical anthropology, (Professor of Medical Anthropology in the Department of Social Medicine at the Harvard Medical School, and a co-director of a charity hospital in rural Haiti, the Clinique Bon Sauveur), was present during the outbreak of the AIDS epidemic in Haiti and followed the diffusion of AIDS from the cities to the rural areas from 1983 to 1990. He carried out systematic interviews of the village people of Do Kay to capture their understanding of this new illness. Framer (1999) describes the discovery of a disorder previously unknown and how the growing awareness and progressive exposure to the illness or to the rumours surrounding it forges the cultural representations shared by the majority of the community.
Through a series of interviews with the same 20 informants over a period of a few years, Farmer documented the rate at which a consensus was formed in the community and the phases of understanding that the community went through. In 1983 Sida (the French and Haitian acronym for AIDS) was first diagnosed in Haiti and the popular press in the US had started to blame Haitians as the originators of the American epidemic. This rumour dramatically reduced tourism on the island and thousands of people lost their jobs. Sida was therefore a public issue and widely discussed on the media but only city dwellers had direct experience of the syndrome. Villagers on the contrary didn’t feel concerned by this illness which they saw as a city illness somehow associated with diarrhoea and with homosexuality. By 1987 after a few cases appeared in Do Kay, a protomodel of the illness was formed in the community. In fact, the first three cases of AIDS which hit the Do Kay community proved influential in the elaboration of a common shared representation of the illness.
The model was strongly influenced by the community’s understanding of tuberculosis (a largely widespread and deathly disease in central Haiti ). Similarities in the symptoms and the fact that subjects with AIDS often develop tuberculosis, made the parallel very strong. The first person to fall ill was a teacher of the local school, the second a young girl who had gone to work in the city, the third a young man working in a well to do family. Both men had good jobs that may have created jealousy in someone: their illness was therefore sent by someone. The girl on the contrary was so poor that no one could be jealous of her therefore her illness was caught by infection of the blood. For most villagers therefore two related deseases existed, the illness caused by sorcery and the illness caused by an infection. Just like tuberculosis a “sent” sickness to be cured requires the identification of the sorcers, while a “natural” sickness can be treated by medecine.
The construction of a cultural model of the Internet in Europe.
In 1996 we had the unique opportunity to observe the diffusion of a radically new media , the internet, in Europe.As an example of a new technology entering a mass diffusion stage, Internet was a perfect case-study to observe how naïve theories of a new medium were constructed and how they evolved with time and direct experience.
In a series of studies in France and Italy we analysed what the public really understood and knew about the nature and the functioning of Internet : what the public thought the net is used for, how the net is organised, who publishes and how information circulates on the net… We analysed the knowledge of the participants and developed an in-depth understanding of their mental models of the Internet. We also confronted the participants to other functioning models of the net in order to evaluate the stability and coherence of their own representations.
We interviewed 60 participants in France and Italy, from various social and professional backgrounds who were all regular computer users. The criterion was to have been using computers for at least two years but to have never used the Internet. Computer users were expected at the time to be the more likely section of the public to move onto the Internet. We were astonished to find essentially two models or theories of the Internet, incidentally both very distant from the reality : a model according to which all websites and information resides on a central computer, and a second model according to which information floats about on the phone lines.
Two main models of Internet.
The results of the research clearly showed that there were two main models of Internet: 45% of the participants imagined the net as a system of individual computers freely interconnected like telephones ; on the contrary 35% of the participants imagined Internet as a centralised system hierarchically organised around a big Central Computer. These two major models were tightly correlated with the kind of functions that participants expected from Internet. The perception of Internet as a communication device led to a very distributed model very similar to a telephone network.
The perception of Internet as a tool for information seeking leads mostly to centralised models built around a big central computer which contains all the information. According to the centralised model, John, who is based in London and wishes to get information about the French National Library, has first to contact the British Central Computer which connects him to the World Central Computer which in turn connects him to the French Central Computer which connects him to the French National Library. To communicate with his friend Paul, who lives in Paris, John sends a message which has to go through all these central computers to reach Paul. As a participant explained : « My request goes through the first central computer – the nearest from my home – which links to lots of people. Then my request is analysed by the general central computer, which collects all the web sites. The sites themselves are on the personal computers. The central computer is a switchboard which knows all addresses». In the distributed telephone model, computers are freely interconnected. This model is based on a strong analogy with the telecommunications and supports other beliefs, like the belief that it is the one who calls who pays for the communication. A modem and a phone line are sufficient to get connected.
The analogy drawn with the telephone network enables people to import practices and concepts related to telephones to the communication via Internet : for example, the future customers view the forums on Internet like phone meetings between three or more people. The construction of representations of illness, technology, social organisations, has an impact on how people will approach these issues but also and foremost how they will act upon them. In our Internet study we discovered that according to what model a user was entertaining, they would make different mistakes in their first interactions with the web. Similarly, Farmer described different attitudes to medication if a disease was “sent” or developed by the patient.
Farmer P. Infections and Inequalities. University of California Press, Berkeley, 1999.
Gentner, D. & Stevens, A.L. Mental Models. London : LEA, 1983.
Norman, D.A The invisible computer. MIT Press, Boston, 1998.
Rogers, E.M Diffusion of innovations. Free Press, New York, 1995 (first edition : 1962).
Sperber, D. La contagion des idées. Odile Jacob, Paris, 1996.