by Allen McGrew

Kicking off our Putting Observational Science to the Test series, Dr. Allen McGrew, Associate Professor of Geology at the University of Dayton, provides some insights on Ken Ham’s distinction between “observational” and “historical” science and the implications of this distinction for geological inquiry. 

All science is observational. All science is historical. All science is theoretical. All science is provisional. There is no such thing as “scientific proof.” At its best, science seeks to establish internally consistent systems of explanation (theories) that relate observed reality to natural laws and/or processes. This is not to say that in science “anything goes”; on the contrary, science involves the rigorous and continuous testing and, if necessary, revision of theories in light of new, carefully framed observations. Observations in space have no privilege over observations in time; in fact, the two are inseparable and mutually dependent. Science becomes wholly inoperable if deprived of either.

As he applies it, Ken Ham’s distinction between “observational science” and “historical science” has little to contribute. Consider this thought experiment: Galileo hypothesizes that gravity operates on all objects in the same manner regardless of their mass, density or size.  To test his hypothesis he defines an experiment that involves releasing a marble from the top of the Tower of Pisa at the exact moment that he also drops a lead cannon ball; he posts an observer at the base of the Tower with instructions to observe and record the moment when the two land. Classic observational science, right? Or is it?

Unfortunately, unbeknownst to Galileo, his observer is a keen student of the philosophy of Ken Ham. Having dropped his test objects, Galileo rushes to the base of the tower to ask whether or not the two objects did indeed arrive at the same moment as he predicted. Alas, the observer responds, he can only offer conjectures. Perhaps the two did arrive at the same moment – indeed, that is what his notes and his memory now tell him– but since this event is now in the past, how can he be sure?  Is it not equally possible that he was created but an instant ago with all his memories intact and his notes in hand?

To live in Ken Ham’s world is to live in “the eternal sunshine of the spotless mind” – a world deprived of one of its most distinctive and fundamental dimensions: time. Human memories, written notes, computer databases, tree rings, layers of sediment, fossils, light from a distant galaxy, DNA patterns in different populations, ratios of parent to daughter isotopes in an ancient mineral grain– all are artifacts of the past. Some probe the very recent past, others the very distant past. Some are highly precise, others more approximate.  All are prone to the possibility of misleading disruptions; consequently, rather than relying on any one observation, it is preferable to continually seek consistency tests based on multiple mutually independent observations.  Nevertheless, all are valid modes of probing the past. Without them, reconstructing the development of natural systems through time would be impossible.

As a geologist, reconstructing the evolution of natural systems through time is the very essence of what I do. Virtually all I do is study artifacts of past; if deprived of those artifacts, then my entire science is robbed of any means of probing Earth history. Consider how Ham might respond to this: In 2012, scientists reported the discovery of a Great Basin bristlecone pine in the White Mountains of California with 5065 rings – but how can we be sure the number of rings corresponds to its age? Granted, thousands of replicate analyses indicate that trees at present typically add one ring per year, but maybe it was different in the past. Maybe this tree added 2, 5, 10 or 100 rings per year.  Who can say?  Who was there to observe?

Similarly, geologists have recognized environments in Antarctica where a chronology based on annual increments of snowfall can be extended back over 700,000 years (the EPICA core), but how can we be sure that seasonal snowfall accumulated the same way in the past that it does today? Or what of radiometric age dating? Carefully designed laboratory experiments document rates of decay of diverse radioactive elements, but how do we know those rates were the same millions of years ago? Or what about Galileo’s experiment alluded to above? Just because Galileo makes a discovery about the fundamental nature of gravity today, can we assume that the same principle applies for all time?

What Ham misses is that analogous arguments could also be levied against observations in space.  Just because Galileo demonstrates a fundamental principle of gravity in Pisa, how do we know the same principle applies in Sydney, Australia, or on the Moon, or on Jupiter? Of course, scientists do worry about such questions, and we continually design new tests to probe the limits and range of applicability of any given theory or natural law. Occasionally we adjust or replace widely respected theories in the light of new evidence– witness how Einstein’s relativity theories displaced Newtonian Physics.

The fundamental point is this: science embraces mechanisms for its own self-correction, but it requires a high bar to displace a theory that has for over a century developed a successful, internally consistent track-record of explaining previously mysterious phenomena. Even more impressive is when the same theory predicts and successfully tests the existence of new and unexpected phenomena. It matters not one whit whether the explanations and predictions apply locally or far away, in the present or the near future or the recent or the distant past.

It is only possible to develop and probe the limits of the unknown by starting from the assumption that the unknown is knowable. Conjectural?  Yes, all science (not just historical science) is conjectural; it operates by processes of reasoned extrapolation. However, we also rigorously and systematically test those conjectures against new observations – regardless of whether those observations probe different points in space or different points in time.

To start by assuming the unreliability of a certain class of observation is to surrender before the game has begun. Basically, Ken Ham’s argument boils down to a single foolish syllogism, as irrefutable as it is meaningless: if we start from the assumption that our observations are unreliable, then how can we rely on them? Wow.  He has got us there. Ham’s reasoning reminds me of the young racer I once defeated in a 5th grade track competition: “I would have beaten you,” he informed me after the race, “If you hadn’t have passed me.”