Computer Science being in the College of Business (and Computer Science), I want to begin this essay with a meditation on the nature of business transactions. Almost all voluntary business dealings of whatever form consist in an exchange between two parties, each offering something of value to the other party and receiving something of (slightly) greater perceived value. The transaction is generally normalized in monetary terms, the buyer offering money and the seller offering some other product or service in exchange for that money, which the seller can then turn around and become the buyer in another translaction with another party. In a free market, each person and each company produces one or more (possibly intangible, such as labor) products and sells those products to willing buyers for money.

It is the nature of a business to buy the raw materials of its product, add some form of value to them, then resell the finished product. The value added almost always involves a labor component, and the labor is generally also bought by paying wages to the persons selling their labor. The employer in this relationship is only interested in obtaining a marketable product at a cost below what it can be sold for; apart from government agencies, the employer has no interest in the person nor welfare of the employee except as it contributes to the profitability of his business (which it does, so there cannot be a complete disinterest).

Computer science professionals contribute value to the economy by selling their expert labor, usually to an employer who specifies what programs they will write, but sometimes independently as a consultant in the same kind of relationship with his clients. The employer or customer cares only for the successful development of the product, and only incidentally for the personal performance of the individuals working on it. Obviously, if a person is not contributing to the product, then that person will soon be looking for another job, but it is actually more important to the employer to deliver finished product to his customers at an acceptable cost than to worry over who worked how much on it. Most employers manage the risk of poor performance in individual programmers and other employees by putting them into teams which average out the variations. Collaboration thus contributes to the success of the product by (to some degree) disregarding the success or failure of the individual contributions. Some employers explicitly tell their team members, "We will not let you fail." They mean the product is crucial, and the individual and start-up failures of particular people in particular incidents is secondary. They do not mean that the employee who consistently fails to contribute value to the company, or that costs the company more in support (by other employees filling in) than they are contributing, will continue to collect a salary. This form of risk management by collaboration is widespread in the industry because it works, even if the employer is unaware of the reason for the benefits of collaboration. Team players help the company succeed, and loners often do not. Every good manager knows this. But the product is still paramount: it matters not a bit how well the team works together if they produce no product, and a divisive and contentious team that ships product is better than a cohesive and happy team that does not.

The university is also an employer creating a product by aggregating the labor of its employees (the faculty and staff) to add value to its raw materials. The product that the university sells is educated persons, people able to enter the workforce as productive members of society, people who can collaborate in those teams to produce marketable products for those employers to sell. Such people will be able to get and hold jobs, and parents and students are willing to pay the university to provide that kind of education. That is the product that the university has to sell, and it is that product the employees -- notably the faculty -- are expected to work together to produce.

To be productive members of society, to be good employees of their future employers, the graduating students of any university -- and SBU in particular -- need to be able to collaborate, but more than that, they need the specific skills their employer wants to pay for. The collaborative skills are mostly taught at an early age (see Robert Fulghum's classic All I Really Need to Know I Learned in Kindergarten), but SBU does have courses designed to develop collaborative skills, and some of them are taught in the Department of Computer and Information Science. Nevertheless, the most important part of the curriculum is the computer science part -- indeed, that's our name. Students who get out without knowing how to build functioning programs, or who do not understand data structures and code structures and networks and client-server interaction and all the other technologies taught in any respectable computer science curriculum, will not be able to earn their salary from any employer interested in computer science graduates, no matter how well they collaborate.

Unlike the employment situation where these skills will be put to use, the individual written products the students develop in the course of their education are not the purpose of their activity. The written products (exams, term papers, problem sets, software) will be discarded; it is the students themselves who are the product of the university, and it is the students that we try to ensure will be successfully produced. The sole function of the intermediate written products is to measure the progress of the student development, and thus also to reward and motivate that progress. Collaboration by students on the intermediate written products is counter-productive, because collaboration hides the individual person's contribution in favor of the collaborated product.

When I teach a course, I carefully design the written materials the students are expected to turn in, and the grades assigned for those materials, to maximize the discrimination between individual students' command of the subject matter. When several students share a single assignment, I am unable to determine from the paper turned in which of them know the material and which of them are getting a free ride. That prevents me from performing the job my employer is paying me to do, and ultimately prevents the university from delivering the product it is being paid for. This university rightfully rewards such sociopathic behavior on the part of the students with strong disincentives, and if necessary removes the unfit products from the product line to protect the quality of the rest of them.

Tom Pittman

Rev. 2002 October 29