The Half-Life of Knowledge

Free essays and podcast episodes are published every Tuesday, and paid subscribers receive an additional essay and episode on Thursdays.


Scientometrics is a field of study focused on the measurement, organization, and analysis of published works related to science.

This field has become increasingly vital as the industrialization of science—especially with the dawn of formal research and development practices and techniques, but perhaps even more so with the dawn of computer-amplified R&D methods—and has thus expanded fairly substantially since 1978 when the Scientometrics academic journal was founded.

One practical application of this field is the ability to understand which research methods are widespread, which researchers are doing the most influential work, and which realms of inquiry are receiving the most attention.

Another is documenting the sometimes slow, sometimes sudden shifts in our understanding of things across various fields.

This is interesting information, but also useful in that it can help us understand how rapidly information we have about the world can be replaced with newer, better information—at times without us realizing anything’s changed.

The Half-Life of Knowledge is a phrase attributed to economist Fritz Machlup, who studied, among other things, the economic value of knowledge back in the 1960s.

Machlup, and others of his day, were trying to figure out how much the acquisition of certain types of knowledge were worth. To do so, he needed to understand how long the knowledge we paid to acquire would remain relevant.

His and similar work was focused on engineering as a baseline profession, and they discovered that while an engineering degree acquired in 1930 was useful for about 35 years before most of the fundamentals a person learned while earning that degree were no longer true, because of new discoveries and the development of better techniques, the half-life of an engineering degree earned in 1960 was only relevant for about 10 years.

In 2002, the president of the National Academy of Engineering, William Wolf, said that the half-life of engineering knowledge was somewhere between seven and two-and-a-half years. More recent estimates have supported the lower-end of that range.

Following a 1966 paper in which he supported the earlier 10-year half-life estimate for an engineering degree, then IEEE Fellow and President of the University of South Carolina, Thomas Jones, decided to work out what it would take for an engineering undergraduate to stay current in their field, over time.

His conclusion was that a typical student invested about 40 hours a week of study, across 120 weeks, to earn their degree—which is about 4,800 hours. He estimated that, because of the loss of relevant knowledge that would occur even during that initial learning period, a student would need to invest five hours a week, 48-weeks a year for the rest of their lives, to understand what’s changed in technology, math, and science.

In other words: over the course of a 40-year career, a typical engineer would need to invest 9,600 hours (and never forget anything they learned, previously) just to keep up with the basics of the field. Which is the equivalent of having to earn two additional undergraduate degrees.

In the years since this paper, the so-called “knowledge decay” at the root this half-life effect has become more rapid in some fields: especially those in which humanity has moved leaps and bounds in terms of our overall understanding and capabilities.

For the latter-half of the 20th century, that’s meant engineering, computer-related knowledge, manufacturing and materials, and the like, have all attained shorter half-lives, while some others may have remained comparably stagnant—though importantly, there are few, if any fields that can be said to be holding still, with nothing changing and no new knowledge emerging beyond what existing practitioners already understand.

This concept supports the assertion that lifelong learning isn’t simply nice to have or only necessary in some trades, but instead probably something we should have baked-in to our educational system from the beginning.

Rather than school being a phase—a period of time when we pick up some skills and knowledge before segueing into the real, adult world—it should be a continuum we’re always on, and which we dip in and out of throughout our lives. Something that’s built into the superstructure of everything else we do.

Building this expectation into our careers, but also the way our economies and educational systems are structured, would allow more people to remain professionally relevant, longer. It would also provide the basis for learning new skills and trades as others disappear due to technological and social evolution, but also as individuals change and grow and decide to try their hand at something new.

This is a useful way of thinking in terms of our professions, but also our understanding of the world as a whole.

Consider that it’s not only engineering knowledge that evolves and grows over time, but also culture, norms, ideologies, and social mores.

There’s evidence that it becomes more difficult to change our minds and accept new paradigms as we age, but that we’re more capable of feeling at home in any time period, no matter how different it might be from the one in which we grew up, if we’re able to maintain a sense of place and context within even relatively unfamiliar settings.

It’s possible to be more comfortable with change, then, but remaining so typically requires that what’s happening around us doesn’t seem startling, shocking, and sudden, instead making sense within the context of what we’ve been learning along the way, during our five hours (or however much we choose to invest) of focused learning each week.

This doesn’t mean we’ll all agree with each other about everything if we all know more about more things, and it doesn’t mean that all changes, in our professions or the cultural landscape, will be inherently and universally good and positive and long-lasting.

It does imply, though, that just as lifelong learning might allow us to stay current within a practical, marketable field, so too might it allow us to stay current in our understanding of ourselves, other people, and the societies of which we’re a part.


Enjoying Brain Lenses? You might also enjoy my news analysis podcast, Let’s Know Things.

There’s also a podcast version of Brain Lenses, available at brainlenses.com or wherever you get your podcasts.