The Department of Computer Science at the University of Toronto is considering changes to its requirements regarding the number of courses students must complete in various sub-disciplines of computer science in order to obtain an M.Sc. or Ph.D. While I am in moderate opposition to these breadth requirements, I present here an argument partially in favour of breadth; this argument is entirely an academic exercise for the sake of challenging my existing views (okay, it’s also fun) and is presented as a single-sided argument.
We, as a species, appear to have a collective case of obsessive compulsive disorder. We have been curators of curios, archivers of art, and hoarders of information, no matter how mundane. In some cultures, überhoarders — learned individuals — were held in high esteem. Think Archimedes, Leonardo Da Vinci, and Ben Franklin. By the mid-20th century, our stash of knowledge was growing too fast for any one individual to keep pace. Due at least in part to the circumstances of the time, the number of individuals holding post-secondary degrees and PhDs grew. The renaissance man gave way to the specialist.
The 20th century gave rise to the Internet and the World Wide Web. Information became more accessible and more people gained access to a vast and steadily growing digital library of facts, misinformation, and cats interacting with invisible items. The number of people entering post-secondary and post-graduate studies continued to grow during the first decade of the 21st century, fueled by worry of an economic downturn. The depth of human knowledge continued to increase while individuals continued to specialize.
In The Wealth of Nations, Adam Smith argued that the division of labour was to be well-regarded. Intellectual specialization seems to be an entirely different beast. Certainly, many important discoveries and ideas have come from the minds of specialists but, at some level, we seem to recognize the importance of interdisciplinary thinking at the social scales: cross-pollination between different fields of study occurs when researchers find common interests. But what about at the human scale? But what of the renaissance man within whom a serendipitous commingling of facts from the broad span of accrued human knowledge can give birth to a wondrous idea? Have we strangled our intellectual and scientific progress by specialization with our pack-rat attitude towards information?
Our accumulation of information (and the occasional bit of knowledge and shred of wisdom) result in the apparent need to specialize to advance any particular field. But when everyone is specialized, it matters little who put two ideas together. One computer scientist is about as good as a hundred, productivity aside. Save for cases of true genius that can see something new where others see nothing, each scientist is fungible; someone will eventually put the same ideas to use ((The ability to act upon those ideas is another matter.)). In fact, this may lead to redundant research as dozens of specialists all arrive at the same idea.
We seem to already have a solution to this information overload problem. Although the amount of information is overwhelming, as search engines became better at locating information we were looking for, sometimes uncannily so with technology like Google Suggest, it has become less important that we be able to recall information so much as to know that it exists and be able to find it. Like using a calculator, use of the Internet for information retrieval can be a crutch. Likewise, when properly used, they can be used to increase productivity and reduce cognitive load to allow more thoughts and tunes from the radio to pop into our minds.
To become an “expert” in a field, perhaps it is now unnecessary to remember every last detail of every major paper published in the area. A knowledge of the important papers is still necessary, but with the ability to retrieve them at a whim, the need for total recall is decreased. This frees up time to read more and perhaps even become an expert in multiple fields.
Even if we cannot become experts in multiple fields, it may be enough to just be aware of major discoveries and ideas from outside the areas of our expertise. We appear to do no harm on that front by forcing breadth requirements on students; Certainly, these breadth requirements can result in more well-rounded individuals or, at the very least, computer scientists versed in its various sub-disciplines. It appears, then, that breadth requirements are laudable. But in my opinion, our (Department of Computer Science’s) definition of breadth continues to remain narrow since we rarely look outside our own department for breadth.
By now, it may seem as though the department needs to decide between producing well-rounded individuals and better researchers through relevant training. But aren’t these two goals satisfied by increasing students’ breadth of knowledge? Scientists that are aware of more ideas have the potential to generate unique and exciting ideas. Taking this to the extreme, we see that students will probably tend towards the same courses to satisfy breadth requirements (out of interest or to reduce workload); instead, incoming students could be allowed to pick one area of specialization and the University can pick one or more other areas at random for the student to assimilate new information. What a menagerie of students would emerge! And each of them a polymath. Would this lead to a renaissance man renaissance?
Again, this is a simplistic academic exercise. Forcing students into classes that do not capture their interest is a waste of students’ time; we need only look into the literature concerning motivation and learning. Admittance into graduate school is an acknowledgement that an individual can work independently to some degree and are capable of rational thought. Considering that probably only a few people enter a Ph.D. program with visions of diving into pools of money, Ph.D. students, if not M.Sc. students, ought to be able to take whatever and however many courses they wish.