This is part of the “virtual book club” discussing readings for the symposium at the University of Missouri on October 7: Moving Negotiation Theory from the Tower of Babel Toward a World of Mutual Understanding.
Rishi Batra suggested reading Samuel Arbesman’s book, The Half Life of Facts: Why Everything We Know Has an Expiration Date (2013). Here’s his description of it.
The author posits that facts in each discipline “decay” and become out of date at predictable rates. This analysis is not specific to negotiation, but it is important to consider for our field. If facts become out of date in other fields, what facts have become so for us? For summaries of the thesis, you can read this essay in the Boston Globe, or book reviews in Slate or the Wall Street Journal, or this interview of the author in The Economist.
John: Arbesman’s basic thesis is that knowledge changes, sometimes so slowly that we don’t recognize the changes. As a result, he argues that we become complacent about believing things without careful thought.
In the Boston Globe essay, Arbesman writes: “When people think of knowledge, they generally think of two sorts of facts: facts that don’t change, like the height of Mount Everest or the capital of the United States, and facts that fluctuate constantly, like the temperature or the stock market close.
“But in between there is a third kind: facts that change slowly. These are facts which we tend to view as fixed, but which shift over the course of a lifetime. For example: What is Earth’s population? I remember learning 6 billion, and some of you might even have learned 5 billion. Well, it turns out it’s about 6.8 billion.”
(A small quibble: the height of Mt. Everest presumably is a slow-changing fact as it probably changes due to natural forces, though this isn’t obvious without precise equipment.)
Arbesman coined the term “meso-facts” for slow-changing facts and he uses the familiar analogy of the frog in a pot of water that is heated to boil so slowly that the frog doesn’t jump out because it doesn’t notice the change.
Some erroneous “facts” are fixed in our minds because they have been generally accepted for a long time. The WSJ piece cites the example that, in 1870, due to misplacement of a decimal point when transcribing his notes, a German chemist mistakenly wrote that spinach has a huge amount of iron – 35 mg per serving – but the actual amount was 3.5 mg. This erroneous finding was repeated until 1937 but by that time, “the legend of spinach’s nutritional power had already taken hold, one reason that studio executives chose it as the source of Popeye‘s vaunted strength.”
Arbesman cites Daniel Kahneman’s work on “theory-induced blindness” reflecting the fact that scientists are prone to intellectual inertia. He illustrates this with the story of 19th Century physicians who ignored research about the benefits of washing their hands and continued to treat patients with unwashed hands for a period of time.
This is consistent with Thomas Kuhn’s theory of the structure of scientific revolutions in which a “dominant paradigm” persists in a scientific community until substantial “anomalies” arise that do not fit real-life experience very well and a new, revolutionary paradigm that works better replaces the dominant paradigm.
“Knowledge, then, is less a canon than a consensus in a state of constant disruption,” as described in the WSJ article.
In the DR field, the trial rate is an example of a meso-fact. People often claim that only 1-2 percent of cases go to trial. This is based on Prof. Marc Galanter’s “vanishing trial” report which found that the federal trial rate dropped from 11.5% in 1962 to 1.8% in 2002. He previously noted that the federal trial rate was 15.2% in 1940. In addition, the trial rates in different federal district courts varies tremendously. But the 1-2 percent figure regularly is bandied about and accepted as an enduring truth. (This doesn’t address the fact that the trial rates in state courts often are substantially higher than in federal court or that “trials” vary widely, but don’t get me started . . .)
Some of the widely accepted wisdom in our field is that cases like Brown v. Board of Education should be tried and not settled. Although this is a philosophical belief, it is based on the assumption that Brown led to school desegregation in the South and general advances in civil rights. However, in The Hollow Hope: Can Courts Bring About Social Change?, Gerald Rosenberg presents an extensive empirical analysis to debunk these factual claims about the Brown decision as well as other court decisions that are widely assumed to have significantly affected social policy.
This “fact” about Brown leads many people in our field to believe, without careful thought, that it is inappropriate to settle cases involving major social policy issues. People making this assumption should consider Jeffrey Seul’s article, Settling Significant Cases, 79 Wash. L. Rev. 881 (2004).
The belief about inappropriateness of settling significant cases is a “social fact” by virtue of being generally taken for granted as the right way to think about the issue.
Under the sociological theory of “institutionalism,” such “arrangements are reproduced because individuals often cannot even conceive of appropriate alternatives (or because they regard as unrealistic alternatives they can imagine).” Paul J. DiMaggio & Walter W. Powell, Introduction, in The New Institutionalism in Organizational Analysis 1, 10-11 (Walter W. Powell & Paul J. DiMaggio eds., 1991). This is akin to Kahneman’s concept of “theory-induced blindness” and Kuhn’s description of the durability of dominant paradigms despite the problems with them.
In this way, the acceptance of the two-model structure of negotiation theory with adversarial and cooperative models (albeit with various labels) is a social fact given its widespread taken-for-granted acceptance in our field.
What prompted you to suggest reading Arbesman’s ideas, Rishi?
And let me turn your question back to you. What facts or beliefs in our field do you think are out of date or just plain wrong?
Rishi: John, thanks for engaging me in this discussion, and for taking The Half Life of Facts so seriously as a text. I know that it isn’t a traditional dispute resolution text, but I thought it was appropriate for a symposium where we are hoping to confront our preconceptions about what we know — and what we think we know.
It is often uncomfortable to challenge our beliefs about anything we know, and particularly things we think we know with certainty. One of the many things that I hope readers of the book or the articles you cite come away with is a kind of comfort with the idea that much of the things we think we know are actually either already out of date, or soon will be.
Mr. Arbesman shows through extensive examples from fields across physics, math, psychology, and even religion that this overturning of information is not random, but rather predictable, and can be modeled in many cases. Through the use of scientometrics (the quantitative study of science), he finds regular patterns in the pace of discovery, and indeed in the pace of the decay of knowledge, i.e. how quickly facts get overturned.
For a symposium that helps us move away from confusion about theory and towards mutual understanding, recognizing that some of our theory, or even much of it, could have an “expiration date” is a helpful paradigm. And to recognize that this phenomenon isn’t due to some sort of flaw in our thinking, but part of the natural process of all academic endeavors, eases the way, at least for me, to being more open and even skeptical of old beliefs.
Interestingly enough, John, two of the facts that you cite in your review are good examples of meso-facts that are due for updating. First, you mention the story of Popeye and spinach that was introduced and then replicated after a decimal point error. However, it turns out that that story of how Popeye came to eat spinach is itself a meso-fact that turns out not to be true.
Spinach isn’t a good source of iron, and people did think it was so in the past. However, the original overestimate was not due to a decimal point error, but due to faulty methods that were updated by the time the first Popeye strips came about. The decimal point story came much later, in an unsourced publication in the 80’s, which was then copied and passed down. After the decimal point story was debunked, Arbesman graciously acknowledged his error, while noting that this is exactly the type of process that forms the core of overturning old knowledge that he writes about in his book.
The other meso-fact you mention is the height of Everest, as an example of a slow changing fact due to natural forces. Arbesman cites the height of Everest in the book as a meso-fact, not only because of natural forces, but also because of, as you mention, better measurement techniques. As our measurements get better, we are better able to figure out how much the height is both rising (due to tectonic plates) and falling (due to erosion and glacial melt), and, we also learn that Everest itself is moving at around 6 centimeters a year, making its very location a meso-fact as well.
I bring up the point about measurement, in particular, because it leads to another area where I see The Half Life of Facts being very relevant. Arbesman touches on the measurement of error in his book, particularly in the measurement of p-values in science, which can tell us if the result of an experiment is statistically significant or not.
It turns out that due to often unintentional manipulations of data and experiments, many results are reported as statistically significant, even when future replication turns out to be unavailing. Often the lack of replication does not come to light for many years, if ever, partly due to the “file drawer problem” where unsuccessful studies or replications get put in file drawers instead of being published.
This has led to what has become known as the replication crisis in the sciences. It turns out that many scientific studies, and perhaps a majority of them, are false . This has most famously come to light in the psychological sciences, where well established and cited results on, say, the priming effect or ego depletion turn out to be due to experimenter bias or statistical artifacts.
Several “replication experiments” are going on right now to try to confirm some of the most robust theories in psychology, with the initial findings showing that less than half have been able to be replicated. And while the lack of replication does impact psychology, recent efforts to replicate biological science results have also been unsuccessful, and commercial laboratories show that only 30-40% of results from academic labs can be reproduced in commercial laboratories.
This can be seen as particularly worrying for the field of negotiation and dispute resolution, where many of our beliefs rely on an understanding of psychology and biology.
I have seen dispute resolution experts talk about the priming effect and how mediators and negotiators should use it to craft their openings. This makes sense in a world where the priming effect has been seen as a true and replicated psychological “fact,” as it was up until just a few years ago. Should this matter as much now?
Moving to another fact about negotiation that we take as true, there have been many studies on the impact of “anchoring” and how this impacts the way negotiators should approach making the first offer.
I know of no studies that show that anchoring doesn’t work, but if suddenly next year we find that anchoring doesn’t work very well, or only in certain situations, should that change a technique that I have seen work for me?
As negotiators, we seem to “know” a lot about what makes people happy in dispute resolution (procedural justice, opportunity to be heard, a result that is “fair”) but many of these theories seem to be the kind of facts that are ripe for overturning.
Another area I see application of Arbesman’s work to our own is in the area of frameworks. Framework thinking is a good way to organize and understand a complex area, and negotiation is no exception. You mention the two model cooperative / adversarial (or integrative / distributive, win-lose / win-win) framework of negotiation as an example of a social fact in our field, and I agree that it is, and I teach it to my students to help understand what is going on in a negotiation.
However, I wonder (and this conference may clarify) how separate these two models really are. As your own research shows, John, negotiators vary across a range of variables that don’t fit neatly in to this existing framework, and can negotiate in ways that defy existing models.
Negotiation frameworks of this type abound in the literature. Another piece I recommended for the symposium is that of the Seven Elements of Negotiation, which is written by Bruce Patton but (I think) first proposed by Roger Fisher. At least a generation of law students have been taught about the seven elements at certain law schools, and I use it myself to give students a way to prepare for their negotiations. Certainly, they seem to find it helpful (as compared to having no framework for preparation at all), but I do wonder if this is an example of theory-based blindness.
Does this framework truly capture all of the important ways to think about preparing for a negotiation (probably not), or does it need to be modified in some way to capture either more elements, or be thrown out all together?
Just as in physics, where there is a search for a “unified field theory” of different forces (what Arbesman would call a “fact phase transition”), I also wonder if it is time for more of a “unified field theory” of frameworks, where we combine some or all of the other frameworks that have been proposed. How do the “seven elements” combine with the “Three Tensions” (Mnookin, et. al.), the “5 Core Concerns” (Fisher and Shapiro), the “three conversations” (Stone, et. al.), or the “process, relationship, substance” frameworks? Can we add your framework in here as well, John?
Also as with the Everest example, technological growth facilitates changes in facts, since technological change and scientific and other knowledge go hand in hand. From a negotiation perspective, I believe that new technologies will both enhance and challenge our understanding of negotiation truths. As we are able to negotiate in virtual environments, for example, we can pull apart the impacts of self-identity vs. perceived identity in a negotiation. Our understanding of neurology and brain science facilitated through better brain scanning will help us actually see some of the internal processes that go in to decision-making in negotiation.
In addition, technology facilitates our ability to negotiate both synchronously (through video) and asynchronously (through email and text) around the world. I suspect that negotiating through these new modalities will both confirm some our understanding of what good negotiation practice is, but also require new techniques and facts about what the best practices should be. And, of course, as technology allows us to interact more and more with other cultures around the world, new facts about cross-cultural negotiation will come to light.
So, to get to your last question John, while I have outlined some areas of concern above, of course, I don’t know for sure which of our facts are wrong yet. As Arbesman would say, that is the point. However, I do know that some of them must be. And depending on the half-life of our particularly young field (any thoughts on how long our half-life might be?), much more of our knowledge will be overturned sooner or later.
I hope some of that happens at our symposium. If it does, instead of being unhappy about being wrong, I will take comfort in the fact that this is part of a regular, and in fact, systematic process that is part of the advancement of all human knowledge.
John: Thanks for your very thoughtful comments, Rishi.
Your suggestion about different frameworks anticipates a post I plan to upload at the end of this series. I will compile a number of frameworks, including the ones you mentioned and many others. I think of this as like putting pieces of a jigsaw puzzle on a table. While it would be nice to put them together into a single picture – or unified field theory – that may be too much to expect. But perhaps we can make some other useful combinations.
Your post cautions us that the shapes of some of the jigsaw pieces may change over time and, indeed, we may need to discard some that are wrong or misleading.