by David Arnold (edited by PJD)

Peter Denning introduced the panel (Fran Allen, David Arnold, and Peter Freeman). We don’t have formal presentations because we want to have a dialog with the people who attended.

PF is pleased with recent developments. CS is broadening. A substantial response to the invitation for an IT Dean’s meeting (held later in the day) indicates that a movement may be brewing; about two dozen schools will make presentations at the meeting and another dozen observers will be present. Many of these schools are not institutions in the sense originally intended by CRA, but the CRA Board unanimously favors taking a broad view of CRA membership. The CSTB is about to undertake a study of the intellectual content of CS research.

FA continued. She said more about the CSTB study for NSF to identify the core content of our field. Who are we? As in the past, physicists, chemists, biologists, and others question whether there is anything to computer science other than infrastructure. CSTB hopes, with this study, to put that question to rest. Everyone else will be able to clearly identify who is in our field of research.

DBA said he helps keep ITPI focused on international perspectives -- professional values in a global industry.

Jim Napolitano RPI: RPI offers Degrees from Computer Science and Computer Engineering, which are not the same as IS. They convene an industrial advisory group that helps them produce the needed graduates. CS should not be seen as training.

Jacob of Dalhousie University in Nova Scotia. IT is too broad – e.g. too widely used in statistics gathering. Before we use specialist names we should define their meaning rather closely. We should resist giving new titles and confusing everyone further. It’s not about territorial battles. If they think we are about infrastructure they have a half truth, but a limited view of our field. We need to clarify with others.

Peter Freeman. The reality is that IT worker is a common term and hence we can’t ignore its existence.

Bill Chu, North Carolina Charlotte. Our university defining an IT college by bringing together Engineering, CS, Math and business to form a committee. IT is in four areas: foundation (CS), Information Management, Technology, and Applications. The program is still being developed.

Rob Kling, Indiana, Founding Professoer in Informatics at Indiana University. CS does not control the bounds of the field. That is set by a variety of demands from industry. Public access to the internet has probably been the largest destabiliser of the definitions. The field could be the union of all ACM SIGs or equivalent. One limiting factor has been the CS view that computing is derived from Turing computability. The internet cannot be understood that way. The Internet is about social behavior. Social communications are an important aspect of all this. As long as the foundations are anchored in logic – human behaviour will be outside the core and HCI is excluded. Statistics is a good example that pays attention to human behavior. The question becomes, what boundaries are useful and interesting -- e.g. Webmasters need to understand computability, but also design and human factors for their own jobs.

Peter Freeman. The definition of CS used by Rob is rather narrow. There are much wider definitions which allow HCI and Human factors and are much broader. A more important issue is the organisational impact of how we define ourselves. The same CS&E core shows up differently in a science school and an engineering school.

Rob. Why I raised the issue of human behaviour, is that HCI has not swept the field. Neither has SE. The point is to see where the core will be stretched. How will the human behaviour side be accommodated within an essential core?

Fran Allen: People are the most important part. How many from industry? Let me give my perspective (an industry one) on why I think this initiative is important and why I signed up. It seemed to me that the field was disintegrating because many of its constituent groups were going their own ways. At the same time, paradoxically, our technology and many of our needs were being driven by collaboration and a convergence on computing, communications, and information. The Internet and Web give us the ability to address much larger and more complex problems and see more of the world. Yet, despite all these common interests, the field had not gelled. It is necessary to be proactive and build an umbrella to bring everyone together. We can capitalize on the fact that the driving forces are collaboration and seeking of solutions farther up the value chain and closer to people’s needs and the customer. The workforce is becoming much more diverse in terms of fields of specialty, ethnicity, and gender. We all need to work with a much broader understanding of the customer environment. Not only have the needs of the workforce have changed dramatically but so have educational expectations. Certification and distance learning are becoming increasingly needed. Colleagues are becoming certified in many different aspects. An umbrella around the field will give us a framework.

DBA was struck by a discussion in a panel yesterday on minorities and women. Bill Wulf and Oscar Garcia said that only 20% of those entering CS eventually graduate. Why so much attrition? One factor is that we don’t have an identity. Other fields do. It’s not very clear what it means to be in our field.

Andrew Hume AT&T labs resonates with this. I am a systems researcher. I try and apply things to information flows in industry. Not widely recognised, but even a small push of the envelope brings a huge impact on implementation complexity.

Fran Allen. One more example. She had a recent opportunity to look across the company. Labels were much wider than you might expect. E.g., A certified IT architect – someone to put together a solution, which could be certified and tested. Fram had not even recognised the existence of the group of professionals until this exercise.

Frank Freidman. Fran hit the nail on the head. We at Temple have an information systems program -- which has moved to the business school. The real focus of that program was on people needs and client needs which brings a very different focus to the CS program. CS needs that as part of the wider educational lessons. Losing it is not good. What are the needs that need to be satisfied for industry (e.g., net-centric computing , client server computing, component based computing). Many students are working in those areas and not at the lower level detail. The Internet has exploded the range of tasks, careers and exploded the CS basis. What are the new things we now need? How can you deal with these without understanding distributed systems?

Almost all departments seem to cover COM/CORBA type courses.

Peter Freeman. Ten years ago there were probably only a fifth the number of IT workers as today. But the number of CS graduates has not increased ten-fold.. Professional society memberships have fallen in the same period. How do we change that? It might happen that we put together academic programs that define the field? We’re probably not going to define the profession academically and probably not through the professional societies. We can contribute. The IT worker study has contributed for example. We can produce white papers, that crisply define things and influence the policy makers, HR managers etc. The prime question is what can we do as academics?

Oscar Garcia (Wright State). A long time ago I gave a talk called "educating the amphibian". The amphibian is now the IT worker who goes through a variety of stages of development. He now asks what do you think is research in IT? Many answers now refer to the Web, networking, business, and e-commerce. Membership in societies has been flat. The societies are perennially interested in broadening their appeal but thus far have not succeeded. We continue to deal with the same issues. The challenge is enormous and old.

Dennis Kafura (Iowa State). I completely agree with Oscar. I began to worry when I found graduates are more up on the content of Wired than CACM. But I am impressed in the way ACM is continuing to expand its areas.

XXX. These comments suggest that we ought to redesign the curriculum from scratch. Processes are important to teach. Curriculum 2001 should be revolutionary not evolutionary.

Jacob (Dalhousie). The problem is that the rest of the world is changing faster than academia. This makes academia appear irrelevant. Three faculties now teach e-commerce. Systems evolve so much faster than degrees can. Professors like to continue to do what they like to do and they are good at it. This is the realism of constraints.

Peter Freeman. We live in an academic equivalent of Silicon Valley startup. We need to live in Internet time. Our departments operate in glacial time.

David Arnold. The requirements are two way. Academia needs to be responsive, but Industry needs to accept that their requirements are often short term and graduateness is important aspect. Training in the ability to think. More flexibility needed in other processes -- e.g. tenure, research conference refereeing.

Frank Friedman (Temple). It seems that this meeting is calling for more involvement in wider definition of CS educational inputs to influence other departments on campus. The web and the Internet have changed many aspects of this. Ten years ago companies looked to CS for problem solving skills and they looked to CS for new devices and tools. Today’s companies find the tools they need on the Web, not in CS departments. The picture is upside down. The call for broadening is not new -- but the pressure for broadening is.

Fran Allen. The NSF tends to fund a lot of collaborative projects. Collaboration is often required. I think there is a big push on this from every direction.

Rob Kling. I want to think about what a CS graduate as opposed to the IT professional. CS graduates can cope with web technology. Issue is the Internet in every home. Our users are now everyone and not just other engineers. A different issue is that of the computer scientist who has no idea of how to assess whether a tool that has been developed will be successfully used. Social scientists are also not equipped to run the experiment. We don’t raise people in CS who talk to people.

XXX. I agree that Academia moves slowly and industry fast. The other night at dinner my neighbour said that few colleagues could use power point. Social scientists and the arts people use the Internet in a broader sense. The Internet changes the view of what distance learning is for these people. We are less good in CS at seeing the wider possibilities.

Rob Kling . Is the ACM moving fast enough? One problem is that ACM has actually driven away some important groups. For example, IS people have their own societies and curricula. Most IS departments are separate from CS, and are often in different schools. IS departments have a lot of human factors research, but few CS departments do. The CS/IS partnership is not strong.