The Ghostly Machine

Stephen L. Talbott

From In Context #4 (Fall, 2000)

“Since World War II,” writes Nicholas Metropolis, "the discoveries that have changed the world were not made so much in lofty halls of theoretical physics as in the less-noticed labs of engineering and experimental physics" (quoted in Dennett 1995, p. 186). And it does appear that the balance between theoretical and experimental science, between pure and applied science, and, more broadly, between science and technology, is shifting.

One of the most prestigious sciences today is genetic engineering — and we hear no great outcry against this popular label from researchers in the field. In physics and astronomy, it has required some of the most grandiose and sophisticated engineering feats of our time to construct the crucial experimental and data-generating machinery. Moreover, the blurring line between research and commerce testifies to the scientist's increasing focus on profitable applications.

Biological Mechanisms

Actually, it's not clear that the distinction between science and technology ever was a very sturdy one. Certainly it didn't count for much in the mind of Francis Bacon, one of the influential progenitors of the scientific revolution:

Not only did [Bacon] maintain that knowledge was to be valued for the power it gives man over nature; but he practically made success in this aim a part of his definition of knowledge. The key words he uses to distinguish the knowledge he exalts from the knowledge pursued by the Schoolmen are “fruit” and “operation.” In other words, not only “science” but knowledge itself, that is, the only knowledge that is not mere trifling, is, for him—technology. Knowledge (for which Bacon, when he wrote in Latin, of course used the word scientia) is that which enables us to make nature do our bidding. (Barfield 1965, p. 56)

But if the distinction between science and technology has always been tenuous, now it is under positive assault. The influential philosopher and Darwinian theorist, Daniel Dennett, argues that biology as a whole “is not just like engineering; it is engineering. It is the study of functional mechanisms, their design, construction, and operation” (Dennett 1995, p. 228). And even in cognitive science the effort to understand the human mind is now widely equated with the engineering project aimed at the creation of functioning artificial intelligences.

In fact, one scientific discipline after another has been transformed by computer modeling and simulation. In meteorology, for example, the preeminent way to seek greater understanding of weather patterns is to construct computer models that can simulate those patterns. The assumption: if your model works, you probably understand the things you have modeled. That is, the model itself somehow captures and displays whatever it is we might aim at with our understanding.

This is a strange assumption to make, given the questions raised by twentieth-century physics about the relation between models and reality. But, in any case, it is a rather natural assumption when you are convinced that the world, like a model, simply is a machine.

There is no doubt that the single-minded determination to produce technologies — things that work — can result in many devices that work wonderfully well indeed. Nor is there doubt that science derives a great deal of its authority in our society from its ability to create these mechanisms. But what exactly do we mean by the “workability” of a mechanism?

Machines as Abstractions

During the industrial era, the machine was often viewed as an oppressive material weight, dragging us down. It may have dragged us down, but less because of its materiality than because of its disregard of our materiality, our embodiment. What oppressed us was the peculiar way the machine functioned — the way it disrupted and narrowed the material context of our work, cutting us off from any genuine conversation with the sensuous presence of the world. Our mechanized actions were stripped of their expressive, gestural qualities and finally reduced to the quantitative abstractions yielded by time and motion studies. The artisan gave way to the assembly line worker.

Another striking thing about a machine is that, so long as it is in good working order, it runs by itself without our participation. And if we narrow our vision to certain aspects of this functioning, we find that the principles of operation are fully disclosed upon the surface of the machine.

By this I do not mean that the principles must be simple, but rather that they are surveyable without remainder by analyzing the “external” relations of all the parts. Once we understand the functional idea of a machine, we can completely specify its mechanical behavior without appealing to any sort of being or entelechy animating it from within (Steiner 1998, p. 272, fn. 32).

Moreover, when we turn our attention away from the qualitative, material presence of the machine to focus instead on the abstract, functional relations between the parts, we find that the relations can be quantified and formalized. Our latter-day achievements in this regard (thanks to Alan Turing, Alonzo Church, John von Neumann, and others) are nothing less than stunning, and the transition from the industrial age to the information age occurred when the formalizations (like the software of a computer) began to detach themselves from their mechanical substrate.

So now the abstract essence of the machine finally floats altogether free of the material world. And while it is true that the thoughts originally articulated into our machines as their functional ideas came from us, increasingly the mechanized world, as it sublimates into abstraction, wields our thoughts. Abstractions do not require a conscious thinker. Chains of formal logic or mathematics march along by themselves with all the coercive rigidity and necessity of the mechanisms from which we abstract them. And the human mind wielded by them moves in lockstep.

Anyone who takes the mind seriously in its own terms is often scorned for believing there is “a ghost in the machine.” But it is the machine itself that has dematerialized. Those who pursue computer models of the mind are the true believers in ghosts; the mind to which they entrust themselves is the mere ghost of a machine.

From Abstraction to Imagination

When we speak of a science whose glory is the fact that it “works,” what we mainly have in mind is its ability to construct mechanisms whose primary function is to sustain a set of abstract relationships. These relationships can be mapped usefully to our tasks in the world, as when we apply spreadsheets to the management of a business.

But it is always necessary to remember that the abstract relationships borne by the machine are, by themselves, lifeless, capable only of sucking the life out of our tasks. It is we who must find a way to restore them to life. If we are content to run a business strictly according to the dictates of a spreadsheet, we will discover that our priorities have become inverted: instead of applying financial discipline in order to achieve some worthwhile end in society, we will allow the bottom line of the spreadsheet to become an end in itself. Any higher achievement requires us to do the hard work of restoring the bare, abstract relations of the spreadsheet to the fullest possible context of meaning.

The difference between a science convergent upon engineering (in the narrow sense of the discipline) and a science pursuing the full potentials of understanding has everything to do with this recovery of context. It is not a matter of rejecting the abstract movement of thought, but rather of marrying it to the counter movement that grasps wholes, images, unifying ideas. This latter is the work of imagination, and we have tried to bring it to the fore in this issue of In Context by presenting a qualitative study of the skunk cabbage.

More needs saying about understanding in relation to mechanistic principles. But, meanwhile, the foregoing might be taken as a partial commentary on one of our more recent efforts to characterize the task of The Nature Institute:

The Nature Institute is dedicated to pursuing a science of nature rather than of mechanisms assumed to lie behind nature. This is a qualitative science, contextual and holistic in spirit, and ethically informed in immediate practice rather than in afterthought. The Nature Institute also promotes humane uses of technology rather than mechanical uses of humans.

References

Barfield, Owen (1965). Saving the Appearances. New York: Harcourt, Brace and World.

Dennett, Daniel C. (1995). Darwin's Dangerous Idea: Evolution and the Meanings of Life. New York: Simon and Schuster.

Steiner, Rudolf (1998). Goethean Science. Spring Valley, N.Y.: Mercury Press