Back to top

CAUT Bulletin Archives

November 2003

Cost of Underfunding Is Too High

David Hill
Students are seen as the beneficiaries of education - they gain what is needed to perform at a higher level, giving them access to higher paid, more satisfying jobs. But others benefit too. Government and industry gain educated employees. Governments also gain increased tax revenue and society benefits from a better educated workforce. In the 19th century, Bismarck introduced universal elementary education in Germany. The resulting industrial advantage forced Britain to follow suit in order to compete. We now face a similar situation at the post-secondary education level although there are arguments about who should pick up the tab.

"Who should pay?" is a social/ political question. Students will pay through higher taxes over a lifetime because they'll earn more. However, companies couldn't operate without university-educated workers - and companies now pay a great deal less than they used to because their contribution to government tax revenue has dropped dramatically over recent decades. Higher education depends on government support from tax revenues as well as student fees and endowments.

Decreasing government support means higher student fees and cutbacks at universities. Increased fees discourage attendance by qualified people who lack resources, introducing inequality as well as wasting talent. Cutbacks reduce the number of places available, again wasting talent. For example, only 30 per cent of qualified applicants gain access to British Columbia's three major universities.

But underfunding has far more serious consequences because it undermines the very foundation of university teaching. It goes beyond the conventional view that universities provide new knowledge or meet some Enlightenment ideal of the pursuit of "truth," important as these may be.

Human performance can be placed in three broad categories: skill-based, rule-based and knowledge-based.

Skill-based performance is automatic, often muscle-memory-type performance. Touch-typing, riding a bicycle, driving a car, welding, even arithmetic ability, are examples of skill-based performance. Skills are largely independent of conscious control or analysis and hard to describe or teach except by demonstration and practice. Indeed, thinking too much about the process often degrades skill-based performance. Normal life is only possible because we can reduce many activities to this performance level.

Rule-based performance is an ability to recognize a problem situation and apply a solution that has previously been worked out. Using a cookbook to help prepare meals or filling in a tax form are everyday examples. Skills like how to chop food or perform arithmetic are, of course, needed in addition to the rules (recipes, tax guides) involved.

Thus rule-based performance involves codes of practice, troubleshooting manuals and other procedures that can, in principle, be written down, but is restricted to things that are already known well enough to write the necessary instructions. Learning consists of learning the procedures and the problems to which they can be applied. This kind of knowledge is descriptive, and often tells us little about the underlying reality. Thus the Ptolemaic view of the movements of the planets in the solar system, which provided a basis for figuring out how the planets would appear to move in the heavens, was reasonably accurate in predicting how they would move. But it bore little relation to the reality of "masses orbiting the sun under the influence of gravity" and was therefore not much use for anything else.

Skill- and rule-based performance both involve predictable completion times and many jobs require both. The United States Navy's procedures for repairing shipborne computers required a non-functioning module to be thrown overboard and replaced by a new one after a certain time had elapsed. Such a time threshold represented the time needed to apply all the standard diagnostic and remedial techniques laid out in the manuals. After that, something new was required, taking a potentially open-ended time to complete - unacceptable in a battle zone.

This was because time taken for the "something new" is not predictable because it involves stepping outside what is known, using a full understanding of the relevant aspects of reality, not just a superficial description or recipe. The path to a solution is simply unknown and there is no sure path to discovery - it requires "knowledge-based" performance.

Sir Isaac Newton said: "If I have seen further than others, it is because I have stood on the shoulders of giants." His words were appropriate, because he actually stole the quote from the ancient Greeks. But what he meant was that he had been able to discover important new things because those before him had given him so much knowledge to work with. This is the essence of knowledge-based performance - the discovery of new knowledge based on a deep (not just descriptive) understanding, together with the imagination to make useful analogies, see connections and the like.

Einstein asserted that: "Imagination is more important than knowledge" - the perfect complement to Newton's view. You need imagination and techniques for using knowledge creatively. This is what a university teaches.

To do this, university faculty must be directly engaged in research (knowledge-based performance), have appropriate faculty-to-student ratios and a generous time schedule. Just as flying instructors need check rides and practice,
faculty need hands-on research and peer-reviewed publications. They need time for this, and one-on-one contact with students. Financially strangling universities does more than imperil Enlightenment ideals or reduce the flow of new knowledge from university research. It destroys the very foundations of the university mission.

Few university graduates will become academics, enter industrial research labs, join think-tanks, or become creative artists - jobs clearly requiring knowledge-based performance. However, a properly university-educated professional in any career will be able to solve problems and then go on to formalize this new knowledge so that it is available to others as rule-based expertise. And the efforts of the "few" are also important - new academics provide the "seed corn" for the next generation for example.

David Hill is Professor Emeritus of Computer Science at the University of Calgary. His continuing research includes GnuSpeech, an articulatory synthesis system within the Free Software Foundation's Gnu Project. He may be contacted at

The views expressed are those of the author and not necessarily those of CAUT.