What makes a technology civic?

For the past few weeks, students involved with C4FCM have been hopping between meetings, brainstorming ideas for new technologies and applications for existing technologies that can promote civic engagement and community building. A laundry list of desirable traits for new technologies with civic potential is emerging. New technologies deployed within a community should be inexpensive, scalable, and easily accessible on the part of the end user. A technology should not require the implementation of an extensive infrastructure, and should aim to make use of existing technologies with which members of most communities are already familiar. Some students are in favor of internal networking technologies that defy the logic of centralization and are not reliant on external networks or communications infrastructure. In addition, a technology should be adaptable so that different communities can tailor it to their specific needs. As a bonus, the technology should help to bridge the digital divide and should be suitable for use by those who are semi-literate or illiterate.

Few expect a technology with civic potential to exemplify all the characteristics highlighted in the wish list above. A quick search for technologies that aspire to some of these standards did, however, yield interesting results. Healthline, for example, is an information access system hosted on a server and designed for use by community health workers in the developing world. The system can be accessed at any time, toll-free, using a fixed line or cellular phone. Community practitioners can call Healthline and retrieve medical information through a spoken conversation with the system in their native language. While the Healthline project does not address issues of community building and civic engagement per se, it serves as a useful model for analysis and to consider the obstacles that might arise when deploying a new technology within a community.

Healthline is a collaborative research project involving Carnegie Mellon University, the Aga Khan University, and the Health and Nutrition Development Society. I was able to speak with Jahanzeb Sherwani, a final-year PhD student in the Computer Science Department at CMU who has developed Healthline and implemented a prototype of the system in Pakistan. He likes to think of the system as an “audio-wiki,” a centralized telephone-based system to which people can call in and receive and update shared information. “The system is much like the web, where anyone can get online and access and share information. Instead of a PC, web browser and keyboard, however, you have a phone and your voice,” he says.

According to this simplified description of Healthline, the system exemplifies many of the ideals for new technologies outlined above. Since it makes use of existing technologies and infrastructure–i.e., phone networks–it is accessible to end users, both literate and otherwise, and helps bridge the digital divide. Moreover, it is community specific owing to its emphasis on using local languages. Sherwani also points out that the system is scalable since servers for the system in a particular country need only be equipped with an adequate number of phone lines to enable access from across the nation.

Sherwani is excited that the prototype for the Healthline application succeeded in its specific goal–medical information access by community health workers. But he acknowledges that voice-based technologies aimed at disseminating information and empowering communities may run into some significant obstacles.

First of all, there’s the software problem. Healthline uses Microsoft Speech Server (MSS) 2007 Beta straight out of the box, without any modifications, reducing the need for much early programming innovation. To make the software specific to different languages as well as contextually appropriate, the Healthline team hand codes the pronunciation of words the system should be able to recognize using the phonetics of US English. This method saves the cost of having ‘language packs’ developed for specific languages by companies that work on speech recognition solutions. In addition , the system can be taught vocabulary that is specific to the context in which it is being deployed (for example, Healthline emphasizes medical terminology, but an environmentalist group using the same application could focus on the vocabulary needed to describe climate change).

That said, the process of hand coding and relying on phonetics to utilize English-language packs has its drawbacks. The extra work adds over a month to the timeframe of project implementation. It also requires the coding skills of computer scientists and the input of a speech expert who is familiar with the phonetics of the required language as well as one of the MSS supported languages. In other words, Healthline and similar models do not successfully bridge the digital divide, as IT experts are required at the implementation stage to put the system in place.

Sherwani also emphasizes that a system like Healthline must be developed in light of extensive local research before it is deployed in a community. “The product has to be customized,” he says. “Different countries and cultures respond differently to interfaces. We need to figure out the basics of interface design so that we can design appropriate interfaces for a variety of specific needs. The fact is, telephony will help the majority of the world’s population interface with any kind of information technology. Before that can happen, though, new interfaces that can facilitate information entry and sharing have to be designed.”

Sherwani’s comments suggest that appropriate interface design may be the key element in successfully deploying a new technology in a specific community. His critique of the constraints posed by using MSS also suggests that streamlined, user-friendly software that can be utilized by non-experts will be the first step towards empowering communities. Even though Healthline is more invested in disseminating medical information in the developing world, as a model in the broadest outline, it serves as a useful starting point from which to conceive of technologies appropriate to the task of fostering civic engagement.