The Half-Life of Facts

by Samuel Arbesman

  Ultimately, the facts of language—in this case, the prescriptive rules that we learn as schoolchildren—are those that are based on our own language experiences. Just as we view technological innovation, facts about dinosaurs, and even the acceptability of different types of Lego pieces with a perspective formed in the crucible of childhood, the same thing happens with language.

  Whatever sort of knowledge we have, at this point it should be abundantly clear that we are far from perfect when it comes to having our own up-to-date facts in our minds. But must it be so? Or can we adapt to all of this change? Can we recognize all the biases that are part of each of us and prevent our knowledge from being updated only once a generation or so?

  • • •

  THERE are already institutions that deal with the far reaches of the fact timescale, trying to help us deal with change. On the fast end of facts, there is a company started a few years ago called Ambient Devices. A spin-off from the MIT Media Lab, Ambient Devices has created a number of informational appliances for rapidly changing information, including market and weather data. For example, they offer an orb that can be placed on a desktop that glows red when the market is down and green when the market is up. They offer the Ambient Umbrella, whose handle glows when rain is in the forecast, so it won’t be forgotten. These gadgets provide a way for us to be kept abreast, in a vaguely useful way, of the changing facts around us. But these sorts of devices are only good for quickly changing facts.

  On the other end of the spectrum there is the Long Now Foundation, which is geared toward fostering long-term thinking and awareness. They get people to think in terms of millennia, and are even constructing a clock out in the Texas desert. Designed to operate for ten thousand years. They also have a sporadically used news-clipping service that highlights articles that might be relevant centuries from now, as a way to keep on top of really big and important changes.

  But what of everything else? What of all the facts that change at that intermediate rate, along the timeframe of years, or of a single generation? This is the timescale of the mesofact, and the one on which most of us fail to keep up-to-date with our facts, the one where cognitive failures reign.

  There is a straightforward, though not always easy, solution to dealing with changing facts: constant education and the omnipresence of information. How each of us implements these rather difficult solutions is certainly a personal choice, but the following are a number of suggestions.

  For example, there is a Web site called Worldometers.27 It is part of the Real Time Statistics Project and acts as a clearinghouse for the sorts of estimated real-time counters we have likely seen, such as the National Debt Clock near Times Square in New York City, which counts the U.S. debt. Worldometers aggregates counters from numerous organizations for such quantities as the current world population and the number of member countries in the United Nations, new books published this year, HIV/AIDS–infected individuals, the amount of coal left on Earth, and the number of species that have gone extinct this year. While many are clearly estimates, they are carefully curated and provide a measure of our current factual awareness of the world.

  But what else can we do? It’s a lot easier to choose a few things to focus on rather than drink from the unfiltered firehose of ever-changing facts. How do we balance this?

  Perhaps some form of informational triage is in order. Or failing that, we can periodically require a radical reexamination of what we believe. How should we do this? Let’s take the example of prices, facts that are always around us.

  Most of us have an intuitive sense of the cost of everyday goods. When we hear that corn costs a dollar for twelve ears, we think this sounds low, and we will likely buy it. On the other hand, some of us might not be aware of certain prices. I’m lactose intolerant, so I have very little sense of the price of a gallon of milk. Frequently checking the Consumer Price Index instead of looking at the prices of individual items allows us to gauge price changes in a more consistent fashion and to put current shifts into historical context. For example, most changes to the CPI have only occurred in recent years, while prices for consumer products were actually quite stable for centuries prior to the twentieth century.

  Similarly, I don’t know the current cost of gold at all, except that it’s rather high. However, I remember following it carefully when I was in middle school, looking every day at the Money and Investing section of the Wall Street Journal and seeing if it was on its way up or down.

  While this is one solution to making sure we aren’t caught unawares by facts, looking up prices on a daily basis seems onerous for most of us. And going even further and putting all of it into a historical context, while interesting, is a lot to ask. It would be great if many people used the Web site MeasuringWorth.com,28 which has data on the purchasing power of the British pound going all the way back to 1245; somehow, however, I don’t see this site becoming particularly popular anytime soon.

  Of course, if someone is an expert on a topic, he is required to keep abreast of related information. Gold traders have to follow its price. Medical facts, as mentioned many times, change and are overturned very often. In order for physicians to deal with this continually vibrating knowledge, they are regularly required to take courses called continuing medical education. Every few years they’re required to get up-to-date on what is considered state-of-the-art. Doctors used to listen to audiotapes and then take a fill-in-the-bubble test. Now they can listen to CDs or even stream the lessons online, and then take an online exam. As we continue to be bombarded by information in our everyday lives, we must take this lesson from experts and actively learn all the time; otherwise we are doomed to be saddled with outdated mesofacts.

  It’s much harder to do this when one’s livelihood doesn’t depend, for example, on medical practices or the price of corn or gold. It’s hard enough to have the newest knowledge in one’s own field, but dealing with knowledge that’s outside of one’s area of expertise is even harder. Unless we want to make it our jobs to figure out how to invest, just use index funds and don’t bother focusing too carefully on individual stocks.

  Perhaps the same advice can be used for knowledge. Unless it’s one’s job to keep abreast of a certain field of knowledge, simply use the informational equivalent of index funds. But what are informational index funds? They are publications and Web sites that aggregate changing knowledge all in a single place. These include magazines, blogs, and the “What’s News” column in the Wall Street Journal, among other sources.

  While informational index funds can help, reading omnivorously is still important, and we have already been given some help with this. The Atlantic has begun running a series called Media Diet,29 which asks influential thinkers what they read and how they get their facts and news. These influential people, from Gay Talese and the newspapers he carefully reads to David Brooks and the blogs he frequents, give their informational diets to help guide others.

  But when it comes to being aware of facts, there’s actually an even better solution: Stop memorizing things and just give up. That sounds terrible, but it’s not. Our individual memories can be outsourced to the cloud. Specifically, rather than relying on memorizing often out-of-date facts, and still usually only half-remembering them, embrace the idea that we have the Internet at our disposal, with search engines at our fingertips that enable us to search for any fact we need anytime.

  This is already happening.30 A recent paper in the journal Science finds that people are coming to rely more and more on search engines rather than their own memory. When the study was released, many people fretted about this and how it is hurting our brains and making us dumber. While this is certainly a common argument,31 I took away the opposite conclusion. Paradoxically, by not relying on our own memories, we become more likely to be up-to-date in our facts, because the newest knowledge is more likely to be online than in our own heads. Medicine has exploited this idea through
a constantly updated online medical reference32 site called UpToDate; looking something up guarantees the most current information.

  Of course, it’s good to keep a certain state of affairs about the world in our minds, but the more we look things up, the more likely we are to not be caught unaware when we encounter a new and startling idea.

  • • •

  EVER-CHANGING facts are all around us. But there is often a disconnect between the state of knowledge around us and what we hold in our minds. This disconnect can sometimes be quite large, and it can be due to a wide variety of reasons, from factual inertia to generational knowledge. Happily there are ways of doing our best to avoid being so surprised when we encounter a new bit of information. Some of this involves simply recognizing the systematic ways in which we update our own personal store of knowledge, or by using technological tools to help us stay a bit more up-to-date. But we can also be aided by something more general, to which this book has hopefully acted as a guide: recognizing the regularities in how knowledge changes around us.

  CHAPTER 10

  At the Edge of What We Know

  AN Italian Franciscan monk named Luca Pacioli wrote a book on mathematics in the late fifteenth century. This textbook, Summa de arithmetica, geometria, proportioni et proportionalita, was one of the earliest books published after the advent of the printing press. While focusing on algebra and other aspects of math, there also was a significant section on accounting.

  Pacioli chronicled double-entry bookkeeping—entering each entry twice, both as a debit and a credit, in order to reduce errors—which was the first time this method had been codified in print. This error-checking methodology,1 which was being used by Italian merchants, finally could be spread far and wide.

  Mary Poovey, an English professor at New York University and the author of the monograph History of the Modern Fact, has argued that the modern conception of the fact,2 with its notion of objective reality that often goes hand in hand with a certain quantifiable quality, was first seen during the Middle Ages. Specifically, Poovey identified the “invention” of the fact with the advent of double-entry bookkeeping. Only with the introduction of this methodology in the fifteenth century or so, Poovey argued, did humans become accustomed to thinking in terms of bits of information in quite this way.

  Of course, this doesn’t quite ring true. As humans we have been chronicling pieces of knowledge for millennia, if sometimes with a bit less objective mathematical truth than that found in a ledger. We can see numerous examples as far back as the ancient Greeks. Anaximander, a philosopher from Miletus in Greece, who lived in the late sixth century BCE, detailed a number of facts about the origins of human beings:3 Initially fish grew out of hot water and earth, and then humans eventually grew inside them until they were capable of surviving on their own. This is one of the reasons that Anaximander discouraged the eating of fish.

  Or we can look to Empedocles, another pre-Socratic philosopher, who wrote of the fact that while water and wine mix together, water and oil do not. Or Anaxagoras, who wrote much about nature and included a discussion of weather in one of his works, where he noted that thunder was a clash of clouds and that lightning is due to the friction of the clouds, like how sparks are formed. Shooting stars are sparks shaken from the air. While the ancient philosophers worked to order their world, they were more than a little hit-or-miss.

  The world was ready for new types of knowledge by the late Middle Ages. As the Middle Ages and its accounting systems gave way to the Renaissance, which in turn laid the foundations for the Scientific Revolution, facts were given a new sort of prominence.

  As the scientific method was being codified, and our surroundings were subjected to experimental rigor the likes of which the world had never seen, facts were generated and overturned at an ever-increasing pace. Finally, the testable scientific fact had arrived. This is the critical insight of the Scientific Revolution: Science requires an idea to be refutable.4 It is not good enough for a concept to seem compelling; it must have the potential for a new fact to come along and render it false. As we have seen over and over, this not only can happen, but often is the rule rather than the exception. Kathryn Schulz, in Being Wrong, notes:

  This is the pivotal insight of the Scientific Revolution:5 that the advancement of knowledge depends on current theories collapsing in the face of new insights and discoveries. In this model of progress, errors do not lead us away from the truth. Instead, they edge us incrementally toward it.

  Ever since the end of the Middle Ages and the Scientific Revolution—in addition to the Industrial Revolution, which saw ever-changing technological facts—we have lived in this new and exciting realm of knowledge. But what is next for facts, for their creation and displacement? Since the sixteenth century, we seem to have been surrounded by knowledge that is changing faster and faster. Will this continue to occur? Are the half-lives of facts themselves shortening, and must we learn to adapt more and more rapidly? Or are we living in some temporal sweet spot for facts, the turning point after which their changes in at least some domains might actually be slowing down?

  • • •

  THROUGHOUT this book I have explored various trends that seem to be speeding up, whether it’s technological innovation, scientific understanding, or even how information wings its way around the globe. But will we ever reach some sort of plateau? Or are we forever doomed to contend with everything around us changing faster and faster?

  As I mentioned in chapter 6, the innovative pace of cities is ever quickening. In order to continue growing, cities seem to actually require periodic drastic innovative reboots—from advanced sewage systems to building methods that allow for skyscrapers—that are happening more quickly. For the first time in human history these rapid changes are occurring multiple times in a single generation, and they don’t seem to be slowing down.

  This parallels work that has examined the rate at which great ideas come about and are integrated into society. In 1971, a team of researchers compiled a list6 of the sixty-two greatest advances in the social sciences since 1900, in an effort to understand how these concepts were generated and what their properties were. Most important, they examined the delay until the impacts of these major advances became manifest. They found that it takes about ten years for them to reach penetration, but critically, this time has been shrinking in recent years. No doubt, if a similar study were done again, we might find even shorter delays.

  Recent climate change has even changed facts around the globe that we thought were far from mesofacts. There are certain island countries in the Pacific Ocean that will likely vanish in the next few decades due to rising sea levels, raising bizarre questions related to statehood: Do submerged islands, even though part of the ocean, remain nation-states?7 The Maldives and Kiribati are in this strange place concerning their statehood, and not because of conquest or political reasons. These countries’ physical existences might be changing, a state of affairs we have never had to grapple with before.

  It would seem that, in these cases, new knowledge is being created more rapidly, spreading faster, and causing changes in things that we had not even realized were in the realm of mesofacts.

  On the other hand, there seem to be examples where the change in facts is slowing down. While Moore’s Law has had an incredible run that has lasted decades, and it has incorporated successive generations of technology, there are many who feel8 that it only has a couple of decades or so more to go. At that point, in the near future, we will start to bump up against limits imposed by physics, such as the size of atoms, which will ultimately limit how many components we can cram onto a circuit.

  The same sort of limits could be argued to hold with transportation speeds. We have had an astonishing sequence of technologies that have allowed us to go faster and faster, but an exponential pace doesn’t seem sustainable forever. While going to the moon for our lunch break sounds wonderful
, it just isn’t likely.

  This does not mean that technological change stops adhering to mathematical rules. But in the long term they might very well be adhering to a logistic curve, with an eventual slowing toward a limit. We are simply in the fast-changing portion in the middle right now, so it is hard to see the eventual slowdown. That being said, as humans we are very good at being pessimistic and underestimating our ability for continued innovation. Even though each individual technology might reach its limits, a new one comes along so often to innovate around these limits that the change around us might not be slowing down for a long time to come.

  But what of scientific knowledge? While we are nowhere near the end of science—the sum of what we don’t know is staggering—we might very well be in a logistic curve of ever-changing knowledge as well, rather than one of exponential growth. One of the reasons I believe this could be true is simple: demographics. It seems unlikely that the rapid population growth will continue growing faster and faster. Whenever a country has become industrialized, its development has gone hand in hand9 with a drop in birth rate. Therefore, as the world as a whole advances technologically, population will cease to grow at the frenetic pace of previous decades and centuries. Combined with energy constraints—we are nowhere near our limits, but our energy resources are certainly not unbounded—exponential knowledge growth cannot continue forever. On the other hand, as computational power advances, computer-aided scientific discovery could push this slowdown far off into the future.