You are here: Articles --> 2005 -->
Editorial: Binary thinking
Vous êtes ici : Essais --> 2005 --> Editorial: Binary thinking
by Geoff Hart
Previously published as: Hart, G.J. 2005. Editorial: Binary thinking. the Exchange 12(3):2, 8–9. http://www.stcsig.org/sc/newsletter/html/2005-3.htm#editorial
It's often said that there are two types of people in the world: those who divide everything into two categories, and those who don't. Scientists and those of us who write about them and their endeavors tend, by our nature, to fall into the former group. Perhaps we adopt this pattern as a result of our ongoing experience with science and technology. After all, the more we study science, the more we see how the universe comes in convenient pairs:
Even the basis for scientific "proof", the statistical test of probability, divides neatly in two: something is either true or false—never true to some extent—depending on which side of an arbitrary statistical divide it falls on, and many scientists frown on reporting results as nearly significant or somewhat significant. Yet when it comes to the really interesting issues of science, this binary thinking can distract us from important discoveries, and can be profoundly misleading. A few examples:
Scientists and philosophers have debated the definition of life for as long as there have been scientists and philosophers, and for long before either profession bore a formal name. Every definition that has been proposed thus far has been deeply flawed. For example:
If you accept the notion that life evolved out of non-living matter, then we clearly have a problem: At the far extreme, we have a solution of miscellaneous organic chemicals that nobody would characterize as living; at the other extreme we have ourselves, clearly alive—at least by our own definition. Somewhere in between these two extremes, self-organizing, energy-binding, environmentally responsive matter crossed an invisible threshold and became alive. The lack of success in finding a universally accepted definition of life suggests to me that the location of that threshold depends more on personal preference than on any objective criteria.
The solution to the problem may thus be so simple that it has been overlooked by those struggling to define a binary, alive versus dead threshold: Our attempts to clearly distinguish between the living and the non-living misses the real point. Wouldn't it be more useful to ask the question of how alive something is rather than dividing a non-binary world neatly in two? Using alive and not alive as two extremes of a broad spectrum is a much more interesting exercise.
An even more vexing problem arises when we attempt to define sentience. Clearly, as the ones who are doing the defining, we consider ourselves sentient. As a result of this chauvinism, science has a long tradition of defining sentience based on purely human characteristics, so that the world falls into two groups: sentient humans and everything else. Even the mind–body problem (whether consciousness exists independent of the body or is indissociably tied to it) becomes a binary issue. Historically, the definition of sentience has become progressively more anthropocentric and ever-narrower as our steadily improving knowledge of animal behavior has eroded many of the distinctions between us and our fellow animals.
But this approach flies in the face of what each of us knows from firsthand observation. Any pet owner knows that their dog, cat, or parrot—and even less clearly sentient creatures such as reptiles and amphibians—shows all the hallmarks of intelligence. Our pets have personalities, moods, and fears, just as we do. Arguing whether they have these aspects to the same degree that we do is a mug's game. Again, the issue comes down to a matter of degree, not an absolute either/or distinction. Defining extremes of sentient and not is more interesting for narrowing the scope of the discussion than for assessing the sentience of any organism.
Carolus Linnaeus, the Swedish taxonomist, carried the urge to classify things to its logical maximum. Not satisfied with groups of two, he created the whole Kingdom/Phylum/Class/Order/Family/Genus/Species system so well known to biologists. This system, which is based largely on anatomical similarities (including reproductive structures) works wonderfully for grouping living organisms. Since "form follows function", organisms with similar forms typically resemble each other in their ecological niches too. Unfortunately, Linnaeus completed his work before genetics became a science, and thus created a system that worked far less well in terms of meeting the needs of evolutionary biologists. The problem with Linnaean taxonomy is that through the process of adaptation to their environment, organisms that are wholly unrelated in an evolutionary sense can end up with very similar characteristics; think of flying insects and birds, for example. (Please note that this particular example should not be extended to a criticism of Linnaean taxonomy, which does distinguish clearly between these and other winged organisms.)
To solve the problem encountered by the evolutionary biologists, Kevin de Queiroz and J. Gauthier proposed the Phylocode system of taxonomy, which groups organisms based on their evolutionary history. (For a discussion of the problem they set out to solve, see <http://www.ohiou.edu/phylocode/>.)
To geneticists and evolutionary biologists, Phylocode is a wonderful tool for organizing organisms in a useful way. For classical taxonomists and most field biologists, of course, it's a direct challenge to the system that has served them well for centuries. Predictably, scientific binary thinkers being true to their nature, the debate over the relative merits of Linnaeus and Phylocode has been conducted with all the formal dignity and grace of a barroom brawl. Of course, someone like me, a pragmatist with no emotional stake in either system, feels obliged to ask the obvious question: why not do both? The Linnaean system has continued to meet the needs of field biologists for more than a century, and there's little reason to discard a system that remains so broadly useful. At the same time, the Phylocode system meets a serious need of evolutionary biology, and to support this work, it should be welcomed with open arms—or at least tested to see whether it's really as good as its proponents claim. And in the meantime, scientists who straddle both fields of research could simply add the Phylocode nomenclature to their existing Linnaean taxonomies, thereby doubling the utility of their classifications, while achieving the equally salutary effect of annoying both groups of binary partisans.
There are two problems with binary thinking. The first is simple reality: far more of the universe is made up of things that are continuous in nature, with no neat distinctions and many intermediate values between any two points, than is made up things with discrete, binary characteristics. Magnetic and electrical fields may be inherently bipolar, but both types of field have an infinite range of magnitudes. The portion of the electromagnetic spectrum we can see (visible light) is a great example of how this works: the classic cartoon rainbow has only seven colors (red, orange, yellow, green, blue, indigo, and violet, with neat binary divisions between any two adjacent colors), but a closer look reveals an infinite range of intermediate shades. Truth itself is very similar, with most of life's "truths" coming in a bewildering shades of grey rather than simple black and white. The second and more serious problem is that, as the debates over sentience and taxonomy demonstrate, binary thinking divides the universe into us and them, opposing camps who can only agree on the need to fight until one camp declares victory. Lost amidst the melee is the potential gain that comes from understanding the value of both sides in the debate and using each side's tools whenever they're most effective.
As scientific communicators, our responsibility must be to avoid the traps of binary thinking. Life, the universe, and everything are far more interesting and complex than binary logic acknowledges, and we do our readers a tremendous disservice when we oversimplify that reality. We may seem to have the choice of communicating in binary mode—or not—but that dichotomy too is a false one. A skilled communicator adopts different solutions for different problems, or perhaps for different aspects of the same problem. This suggests that we should use binary thinking where it makes the most sense, while still remembering to supplement it with a more nuanced approach should that prove more appropriate.
One of my favorite old jokes illustrates the problem of binary thinking and the benefits to be gained from learning when to temper it with that more nuanced approach: An engineer, a physicist, and a statistician were hunting moose in Canada, and after a short walk through the marshes they spotted a huge moose. The physicist raised his gun and fired at the moose, but missed; the splash from the bullet striking water revealed that the bullet had landed 3 metres to the right of the moose. The engineer, realizing that there was a substantial breeze that the scientist had failed to account for in his theoretical model of the moose, aimed to the left of the moose and fired. Coincidentally, the wind died at that precise moment and his bullet also missed—with a splash of water exactly 3 metres to the left of the moose. The statistician immediately jumped up and down screaming, "We got him! We got him!"
In science, as well as in moose hunting, the really interesting things usually lie somewhere between the obvious extremes.
My essays on scientific communication have now been collected in the following book:
Hart, G. 2011. Exchanges: 10 years of essays on scientific communication. Diaskeuasis Publishing, Pointe-Claire, Que. Printed version, 242 p.; eBook in PDF format, 327 p.
©2004–2014 Geoffrey Hart. All rights reserved