A while ago I started realizing that the vocabulary of science is, in general, rarely precise. The terms we use, and in fact even the equations we use, are only convenient handles that allow the human mind to grasp the concepts with which we wrestle. We must use such "conceptual handles" because language and the need to communicate constrains our mind. Fundamentally, we must break the continua and fuzzinesses that compose reality into discrete chunks in order to converse about them; as a result we must think about them in the same way. Computers today are even poorer at grasping continua and more dependant on discretization than are we. (This gap should be in the back of the minds of those that design tomorrow's computers and interfaces.) We must accept this limitation of our thoughts - it is so much easier for us to work with discrete items. The danger comes when we forget that our words and equations are only discrete approximations, and begin to accept our models as reality.
The first seed of this thought came to me while reading "Artificial Life" a few years ago. (Many apologies: I do not have the reference with me, and at two in the morning I can't even remember the author.) The author challenges the notion that things are either alive or not. Rather, the reality is a continuum; a human is more alive than a cricket, which is more alive than a bacterium, which is more alive than a grain of sand. The continuum is based on whatever criteria you choose for "life"- movement, growth, reproduction, intelligence, metabolism, ability to increase information and reduce entropy on a local level, etc. An object that exhibits more of these criteria, or exhibits them more strongly than another, is more "alive". Our traditional, binary, use of the word "life" is very handy in many incidences; a coroner needs to know if a person is dead. But trying to force this use of the term into all situations leads down the dark path- witness the ongoing debate concerning the aliveness of a virus. Biologists disagree. But the debate is solved if we simply admit that our term is inadequate to place all objects in one category or the other. To a biologist who believes viruses are just below the "threshhold", I ask- are rickettsia alive? To one who believes viruses are alive, I would ask the same question about a prion or a self-catalyzing ribozyme.
(Exercise: If you accept the above argument that livingness is a continuum, try to objectively apply that concept to the issue of abortion. You are sure to tie your brain in knots. It's fun!)
Perhaps strangely, the people most resistant to the above concept are, in my
experience, the biologists. In a recent seminar on evolution, I and several
classmates were discussing the early origins of cellular life. The "standard
model" goes something like this: The early Earth had oceans and a strongly
reducing atmosphere which contained CO
Since so many biologists will have trouble accepting the above argument, let us examine a second instance of the same delimma where most biologists accept the argumentation. The difficulty is in defining the boundaries of a "species". As innumerable naturalists have observed, nature does not conveniently pigeonhole every organism on the planet in such a way as to make specific categorization simple. Hybrids, subspecies, and variants abound. To compensate for this, many different concepts such as "biological species", "taxonomic species", and "recognition- differentiated species" have been proposed. At this point in time, many, if not most, biologist recognize the fact that we differentiate species for our own convenience, but that nature does not play that game.
Not to leave out the other disciplines, a very familiar and well-accepted example is present in physics: the wave/particle duality. We have known for many years that neither waves nor particles exist. There is not a single particle nor a single wave in this universe. So why do we use them? They are convenient approximations, just as Newtonian mechanics is a convenient approximation for motion when speeds are low and objects are small. We use the discretized handles because the reality is harder to write about and harder to think about; and the effort of muddling through the more complex reality is often unnecessary. But we should beware the danger of forgetting that our handles are only models.
I'll end this discourse with a few other examples I've come across. I'll add more as they come to me.