Editorial: What's in a word?

by Geoff Hart

Previously published as: Hart, G. 2008. Editorial: What's in a word? the Exchange 15(1): 2, 9–10.

In a previous editorial (the Exchange 14(1), February 2007) I wondered "what's in a name?" This musing having now fermented on my mental compost heap long enough to become reasonably good intellectual fertilizer, it seems like a reasonable time to return to those musings and take them in a slightly different direction. Specifically: Like all good technical communicators, scientists use words very precisely, to unequivocally communicate specific concepts. In theory, the goal is to present information objectively, free of all the subjective trappings that so bedevil clear communication. In practice, this approach does indeed go a long way towards achieving at least the appearance of objectivity, but that appearance can be deceiving. Consider, for example, how even a seemingly straightforward word such as species can vary in its meaning.

The simplest definition—indeed, the one most of us learn in grade school—is that species comprise groups of organisms that can interbreed successfully; that is, they can produce viable offspring capable of perpetuating the species. This is certainly a useful definition both to laymen and to scientists who study plant breeding because it reminds both groups that sunflowers and watermelons are very different species and cannot produce offspring no matter how many storks might intervene. (Of course, that ignores the wonders of modern genetics technology, which allow the transfer of genes between even such seemingly disparate organisms as fireflies and bean plants. More on that in a moment.) This initial attempt at a definition suggests to us that horses and donkeys are also different species, though here the waters grow a bit murkier: even though they can interbreed and produce offspring (mules), their offspring are sterile and cannot themselves reproduce. So possibly they're different species, at least from the perspective of this definition. Yet no one would declare a man and a woman who cannot produce children together different species simply because they are infertile as a couple. Clearly, this definition of species needs further work, whether your purpose is to inform the breeding of new organisms—or only to be able to find your beans should you drop them in a dark kitchen.

If you travel a bit farther along the road of scientific knowledge, you'll inevitably encounter the subject of taxonomy, most famously in the form that classifies organisms into the nested boxes of domains, kingdoms, phyla, classes, orders, families, genera, and species. Classical taxonomy, which Linnaeus originated early in the game of defining species (and indeed, early in the game of modern science), worked primarily on the basis of similarities in physical structure, on the very plausible logic that similar structures within a group of organisms probably had similar underpinnings; a rose is a rose, after all, except when it's a Rose. Unfortunately, Linneaus new nothing of genetics and evolution, and thus did not know that structural differences might mask closely related organisms nor that structural similarities might mask widely unrelated organisms. The characteristics used by Linnaean taxonomists to classify organisms gradually became increasingly obscure, as can be seen in the debates among paleontologists over the degree to which subtle differences in bone morphology could define whether two bones from our most ancient ancestors belong to the same species, and in the related debates over the consequences for human evolution. Since no one can test whether these differences are truly meaningful (the organisms from which the bones were drawn being long extinct), clearly more information is needed.

What about genetics? Now we seem to be on more promising ground. It's clear that the complete set of genes possessed by a group of related organisms (their genome) provides an unmistakable clue to the identity of a species. Organisms that diverged long ago will be more genetically distinct than organisms that diverged more recently, and—returning for a moment to our interbreeding definition of species—organisms whose genes are sufficiently different will be unable to breed because the mixture of genes won't work successfully together. Obscure though genes are, at least in comparison with gross morphological differences such as having wings or fins rather than arms that a Linnaean taxonomist might focus on, this approach has the additional merit of appealing to those of us who still feel that the ability of organisms to breed is the touchstone for whether two organisms belong to the same species; as I noted in my previous editorial, sometimes the public consensus on the meaning of a name can be as important as the scientific reason for choosing that name. Unfortunately, genetics alone can mislead us. A chimpanzee and a human being are commonly said to be 99% identical in terms of our respective genomes, yet nobody would consider us to be the same species. (Even if you don't notice the other obvious differences, the chimp's extra pair of chromosomes is a dead giveaway that things aren't so simple.) As a result, that number is deceiving. Whether you raise that chimpanzee from the moment of birth in its original jungle home, or in a modern suburban home, a chimpanzee it will remain—not a human, even if raised with the best intentions and the most sincere efforts to turn our hairy Eliza Doolittle into a proper Englishwoman.

Clearly, genetics alone aren't sufficient to solve the species question without considerable caution: if 99% of the chimp's genes are shared with humans, we have only a 1% chance of choosing the few key genes that distinguish between us if we don't know what we're looking for. And knowing what to look for might not be so easy as it seems. Though some characteristics of an organism such as human blood types are clearly determined by one or a few genes, most complex traits are determined by a moderate to large number of genes that not only interact with each other, but also interact with their environment. Picking a chimp out of a crowd is child's play. But genetically identical organisms can differ radically in appearance under the right circumstances. Rooted cuttings from the same individual plant, for example, may have entirely different phenotypes (e.g., appearances) if they are grown in sufficiently different environments. Worse yet, modern genetic engineering allows us to promiscuously swap genes between organisms located about as far apart on the evolutionary tree as we can imagine. In 100 or 1000 years, will we be able to easily define which species these chimeric organisms truly belong to?

Most egregiously, cultural factors come into play as well. Should one wish to set aside a specific subgroup of humanity for nefarious purposes, it would clearly be useful if you could define that group as a different species. But if the two resulting groups have compatible reproductive organs and are capable of interbreeding, they're clearly the same species. And if they share more of their genomes than chimps and humans—enough so that all the chromosomes match up precisely—they're also clearly the same species. Thus, we must invent a new word that accomplishes the same goals as species without failing the test of allowing us to discriminate. The word that solves this problem is race, and its definition is every bit as subjective as the most demanding discriminator might wish, since the goal of choosing this word is to seek until we find one or more criteria, no matter how absurd, that allow us to define the other group as different.

This last example sums up the previous examples by revealing how clearly a word definition can fail the test of objectivity: in each case, the seemingly objective word choice is objective only within a specific context. For the amateur and professional naturalist alike, the simple Linnaean species taxonomy is all we need to identify that fascinating plant we just touched so we can figure out how to treat the resulting rash. For both amateur and professional plant breeders, knowing that sunflowers and watermelons are different species helps suppress the temptation to grow 6-foot-tall plants whose dangling watermelons relentlessly track the daily course of the sun. For professional gene twiddlers (thankfully, there are as yet few amateurs), being able to define a species based on its genotype allows them to carefully control their studies of the functions of individual genes and how those functions differ among species.

From our perspective as scientific communicators, that's the more useful and interesting insight: not that we humans are inherently subjective beings, but rather that we use words to accomplish specific goals within specific contexts. Understanding that subtle point is what lets us choose the best word for the job. It also (as in my example of species) helps us to recognize that words may acquire different connotations in different contexts, and, recognizing those differences, helps us to provide the additional words required to warn readers not to apply the wrong context when they attempt to understand what we're trying to say.

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.