The Retrodiction Wall

Wednesday

---

Prediction in Science

One of the distinguishing features of science as a system of thought is that it makes testable predictions. The fact that scientific predictions are testable suggests a methodology of testing, and we call the scientific methodology of testing experiment. Hypothesis formation, prediction, experimentation, and resultant modification of the hypothesis (confirmation, disconfirmation, or revision) are all essential elements of the scientific method, which constitutes an escalating spiral of knowledge as the scientific method systematically exposes predictions to experiment and modifies its hypotheses in the light of experimental results, which leads in turn to new predictions.

The escalating spiral of knowledge that science cultivates naturally pushes that knowledge into the future. Sometimes scientific prediction is even formulated in reference to “new facts” or “temporal asymmetries” in order to emphasize that predictions refer to future events that have not yet occurred. In constructing an experiment, we create a few set of facts in the world, and then interpret these facts in the light of our hypothesis. It is this testing of hypotheses by experiment that establishes the concrete relationship of science to the world, and this is also a source of limitation, for experiments are typically designed in order to focus on a single variable and to that end an attempt is made to control for the other variables. (A system of thought that is not limited by the world is not science.)

Alfred North Whitehead captured this artificial feature of scientific experimentation in a clever line that points to the difference between scientific predictions and predictions of a more general character:

“…experiment is nothing else than a mode of cooking the facts for the sake of exemplifying the law. Unfortunately the facts of history, even those of private individual history, are on too large a scale. They surge forward beyond control.”

Alfred North Whitehead, Adventures of Ideas, New York: The Free Press, 1967, Chapter VI, “Foresight,” p. 88

There are limits to prediction, and not only those pointed out by Whitehead. The limits to prediction have been called the prediction wall. Beyond the prediction wall we cannot penetrate.

The Prediction Wall

John Smart has formulated the idea of a prediction wall in his essay, “Considering the Singularity,” as follows:

With increasing anxiety, many of our best thinkers have seen a looming “Prediction Wall” emerge in recent decades. There is a growing inability of human minds to credibly imagine our onrushing future, a future that must apparently include greater-than-human technological sophistication and intelligence. At the same time, we now admit to living in a present populated by growing numbers of interconnected technological systems that no one human being understands. We have awakened to find ourselves in a world of complex and yet amazingly stable technological systems, erected like vast beehives, systems tended to by large swarms of only partially aware human beings, each of which has only a very limited conceptualization of the new technological environment that we have constructed.

Business leaders face the prediction wall acutely in technologically dependent fields (and what enterprise isn’t technologically dependent these days?), where the ten-year business plans of the 1950’s have been replaced with ten-week (quarterly) plans of the 2000’s, and where planning beyond two years in some fields may often be unwise speculation. But perhaps most astonishingly, we are coming to realize that even our traditional seers, the authors of speculative fiction, have failed us in recent decades. In “Science Fiction Without the Future,” 2001, Judith Berman notes that the vast majority of current efforts in this genre have abandoned both foresighted technological critique and any realistic attempt to portray the hyper-accelerated technological world of fifty years hence. It’s as if many of our best minds are giving up and turning to nostalgia as they see the wall of their own conceptualizing limitations rising before them.

Considering the Singularity: A Coming World of Autonomous Intelligence (A.I.) © 2003 by John Smart (This article may be reproduced for noncommercial purposes if it is copied in its entirety, including this notice.)

I would to suggest that there are at least two prediction walls: synchronic and diachronic. The prediction wall formulated above by John Smart is a diachronic prediction wall: it is the onward-rushing pace of events, one following the other, that eventually defeats our ability to see any recognizable order or structure of the future. The kind of prediction wall to which Whitehead alludes is a synchronic prediction wall, in which it is the outward eddies of events in the complexity of the world’s interactions that make it impossible for us to give a complete account of the consequences of any one action. (Cf. Axes of Historiography)

Retrodiction and the Historical Sciences

Science does not live by prediction alone. While some philosophers of science have questioned the scientificity of the historical sciences because they could not make testable (and therefore falsifiable) predictions about the future, it is now widely recognized that the historical sciences don’t make predictions, but they do make retrodictions. A retrodiction is a prediction about the past.

The Oxford Dictionary of Philosophy by Simon Blackburn (p. 330) defines retrodiction thusly:

retrodiction The hypothesis that some event happened in the past, as opposed to the prediction that an event will happen in the future. A successful retrodiction could confirm a theory as much as a successful prediction.

I previously wrote about retrodiction in historical sciences, Of What Use is Philosophy of History in Our Time?, The Puppet Always Wins, and Futurism without predictions.

As with predictions, there is also a limit to retrodiction, and this is the retrodiction wall. Beyond the retrodiction wall we cannot penetrate.

I haven’t been thinking about this idea for long enough to fully understand the ramifications of a retrodiction wall, so I’m not yet clear about whether we can distinction diachronic retrodiction and synchronic retrodiction. Or, rather, it would be better to say that the distinction can certainly be made, but that I cannot think of good contrasting examples of the two at the present time.

Effective History

We can define a span of accessible history that extends from the retrodiction wall in the past to the prediction wall in the future as what I will call effective history (by analogy with effective computability). Effective history can be defined in a way that is closely parallel to effectively computable functions, because all of effective history can be “reached” from the present by means of finite, recursive historical methods of inquiry.

Effective history is not fixed for all time, but expands and contracts as a function of our knowledge. At present, the retrodiction wall is the Big Bang singularity. If anything preceded the Big Bang singularity we are unable to observe it, because the Big Bang itself effectively obliterates any observable signs of any events prior to itself. (Testable theories have been proposed that suggest the possibility of some observable remnant of events prior to the Big Bang, as in conformal cyclic cosmology, but this must at present be regarded as only an early attempt at such a theory.)

Prior to the advent of scientific historiography as we know it today, the retrodiction wall was fixed at the beginning of the historical period narrowly construed as written history, and at times the retrodiction wall has been quite close to the present. When history experiences one of its periodic dark ages that cuts it off from his historical past, little or nothing may be known of a past that once familiar to everyone in a given society.

The emergence of agrarian-ecclesiastical civilization effectively obliterated human history before itself, in a manner parallel to the Big Bang. We know that there were caves that prehistorical peoples visited generation after generation for time out of mind, over tens of thousands of years — much longer than the entire history of agrarian-ecclesiastical civilization, and yet all of this was forgotten as though it had never happened. This long period of prehistory was entirely lost to human memory, and was not recovered again until scientific historiography discovered it through scientific method and empirical evidence, and not through the preservation of human memory, from which prehistory had been eradicated. And this did not occur until after agrarian-ecclesiastical civilization had lapsed and entirely given way to industrial-technological civilization.

We cannot define the limits of the prediction wall as readily as we can define the limits of the retrodiction wall. Predicting the future in terms of overall history has been more problematic than retrodicting the past, and equally subject to ideological and eschatological distortion. The advent of modern science compartmentalized scientific predictions and made them accurate and dependable — but at the cost of largely severing them from overall history, i.e., human history and the events that shape our lives in meaningful ways. We can make predictions about the carbon cycle and plate tectonics, and we are working hard to be able to make accurate predictions about weather and climate, but, for the most part, our accurate predictions about the future dispositions of the continents do not shape our lives in the near- to mid-term future.

I have previously quoted a famous line from Einstein: “As far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality.” We might paraphrase this Einstein line in regard to the relation of mathematics to the world, and say that as far as scientific laws of nature predict events, these events are irrelevant to human history, and in so far as predicted events are relevant to human beings, scientific laws of nature cannot predict them.

Singularities Past and Future

As the term “singularity” is presently employed — as in the technological singularity — the recognition of a retrodiction wall in the past complementary to the prediction wall in the future provides a literal connection between the historiographical use of “singularity” and the use of the term “singularity” in cosmology and astrophysics.

Theorists of the singularity hypothesis place a “singularity” in the future which constitutes an absolute prediction wall beyond which history is so transformed that nothing beyond it is recognizable to us. This future singularity is not the singularity of astrophysics.

If we recognize the actual Big Bang singularity in the past as the retrodiction wall for cosmology — and hence, by extension, for Big History — then an actual singularity of astrophysics is also at the same time an historical singularity.

. . . . .

I have continued my thoughts on the retrodiction wall in Addendum on the Retrodiction Wall.

. . . . .

. . . . .

. . . . .