Engineers everywhere shake their heads (or shudder) when a recent graduate new hire shows up bursting with knowledge and falls flat on their face when instructed to “do some engineering.” I oversee/mentor a few groups of seniors every semester in a Senior Design course that is intended to mitigate or avert the situation related above. So, I see a lot of painful mistakes. Many of the mistakes are things that I would consider fundamental, that is, for example, using a multimeter to measure the voltage “thru” a component rather than “across” it, or vice versa with current. (By the way, I did this once myself as a kid, before I knew Ohm’s Law and its physical meaning. I burnt-out the current measurement circuit in Dad’s meter trying to measure the voltage at a wall socket.) Short of more hands-on time in the laboratory, I began to ask myself why this was the case.
Abstraction is the concept that allows us to perform high-level tasks without understanding how the underlying process works. Technology is a fabulous example of abstraction at work. For instance, few people can really say that they understand how a computer works. I’m an EE and I have a pretty good idea of how they work; but, I wouldn’t necessarily be comfortable trying to develop a computer from scratch (at the gate or transistor level). So, this brings us back to the multimeter. At one time, the volt meter and ammeter were separate instruments; so, there was more distinction between the measurements. Is this a failure of abstraction or a failure of understanding? It does represent a fundamental failure to understand the physics: the idea that current is the flow of charge and that voltage is the potential (to do work) between charges. But, if no effort is made to stop during the education process to say, “What does this tell us about the physics?” these failures will continue. (Part of the problem is that some of the instructors have a tenuous grasp on the physics as well.) Abstraction remains an important concept, though, permitting sophistication.
The other problems are mass-production and “cheap” information. It used to be that if you wanted to get a job, you needed to know how to do something. (See The Cluetrain Manifesto for more on this sort of thing.) It is rare that one person knows and understands how an entire product is designed and constructed. This is the genius of Henry Ford and friends. It’s efficient for business, but not so good for humanity. Furthermore, we tend to let our knowledge exist in and be cataloged by Google these days. It does not occur to most of us that it might be good to own a hard copy of information that pertains to our knowledge of a product, process, or procedure. Google offers pin-point precision at the expense of “situational awareness,” understanding how the information fits into a greater context. Think of how many great discoveries may slip through the cracks of a web search!
The final culprit is assisted computation. There is a certain amount of insight gained from struggling with a mathematical expression for a physical relationship. There is even more insight in quick approximations. The art of approximation is dying. Dr Donald Milks taught my undergraduate Statics course. He was of the old school. I used to race him in solving example problems in my head using approximations. He almost always won. The few other students who were interested enough dutifully punched away on their calculators. I usually beat them, though. Dr Milks valued good hand-drawn figures, too. Indeed, the act of drawing, too, yields insight about a problem. I miss that.
Technology and abstraction allow us to solve progressively more difficult problems in science and engineering. I embrace that. However, we must take care not to lose sight of the fundamentals. For with time, the most advanced problems may prove intractable when no one understands why we want to know the answer.