History Isnt Science

From SecularHumanism

Historians do not learn, nor do they ever aim to learn, anything useful from history (only about history). - Some historians insist on this, but really, it's only true in the tautological sense that someone applying history isn't acting in their capacity as a historian per se. There is no shortage of historians who are also sociologists or economicists of varying calibers; psychohistorians, of course, cross-list as psychologists.
The suspicion that historians just collect facts and oddities leaving the real work to science is unfair. A similar line or reasoning would say that scientists spend their time recording observation statements and constructing empirical generalizations. Much of the power of science is that it goes beyond the facts rather than simply repeating them. This is why you are correct in thinking that we can learn from science. But a historian exploring lost cultures or constructing ancient buildings from their ruins also goes beyond the facts. Instead of using abstraction and generalization as tools they use the richness and diversity of the facts available as selling points for their models. Both historians and scientists have to make do with incomplete facts. Nonetheless they both manage to present compelling theories. We learn from both. -- ChrisSteinbach

I think I've already addressed your concerns but I'm willing to do so again in detail.

Historians are in the position of experimental physicists. It is not true that they generalize or construct models from their data. It is also untrue that they assign significance to some facts more than others. They do differentiate between historically noteworthy and unworthy facts, but that's all and it isn't good enough! There is no field of science which considers all facts within its purview to have equal value.

If historians were scientists, or acted like scientists, they would make a single model that encompasses all of history. This model would start by explaining the big stuff like why wars happen like clockwork throughout history, and anything fundamental in human society that's changed in the last 10,000 years. Have historians done this? No, they have not. Psychologists like JulianJaynes and LloydDeMause have done this. Political scientists like NoamChomsky have done this. People who are not historians.

What's happened is that historians fucked up, they fucked up so much that people in other fields of science decided to step in and do their job for them. Interdisciplinary work is all to the good but let's be realistic; it's a strongly discouraged abnormality in academia. Academics don't intrude in another field of study, usually considering it beneath them, unless they have decided the people in that other field are hopeless.

As I said, historians are in the position of experimental physicists. The key difference is that experimental physicists are guided by theoretical physicists. The whole of experimental physics functions as input to theoretical physics. And in turn, theoreticians are there to tell experimentalists what facts are noteworthy and exactly how noteworthy they are. Theoreticians are there to tell them that the discovery of the Higgs boson is worth a million times more than adding another couple digits to the exact value of the gravitational constant. Historians do not have a similar arrangement. There is no group of historians that does theoretical history. So in their absence, historians are guided by their own interests and those of the general public. This isn't sufficient because individual interest is unorganized and public interest is uninformed. And that's why historians' achievements have been so paltry when compared with physicists'.

Actually, the situation is a bit more complex. Historians aren't total retards so some of them have recognized the need for theoretical work and model building in history. There have been many attempts to explain why wars occur. And they've all been abysmal failures leading other historians to discourage model-building in the first place. Never have historians arrived at the conclusion that their own knowledge is insufficient to explain history and taken the necessary step of inviting other fields to collaborate. The current arrangement with Jaynes, deMause and Chomsky exists in spite of historians' best efforts. It also exists in spite of psychologists' and political scientists' best efforts! So it's not that historians collect facts and oddities while leaving the real work to science, no it's worse than that. -- rk

Maybe so, but I wouldn't trust Jaynes further than I could throw him. Let the historians improve their own house.

Because they've done such a sterling job over the past few centuries ....
You certainly said a mouthful here Richard. I think I'll have to tackle this in bite-sized chunks.

It is not true that [historians] construct models from their data.

They do! They do! They do! This is important. I'm not sure if you envisage historians as digging up information and running to scientists like a dog running to its master with a bone. Or maybe you have already accepted the historical data you speak as factual. No matter.

It is ironic that I should now have to argue that what historians say should not be taken at face value.

When I spoke of models and theories, I did not have in mind the rules and generalizations abstracted by scientists. I was thinking more of a jigsaw puzzle where most of the pieces have been destroyed or lost. The historian fills in the blanks in the picture using her imagination. What results is almost completely theoretical, even though it contains the known facts in all their detail. Such a model can be used to answer certain questions in a very direct manner.

I don't mean to imply here that the terms 'fact' and 'theory' have set meanings or that you have misused them. I myself only make the distinction for convenience. But in prematurely accepting the output of historians as factual you manage to do both them and yourself a disservice.

Are we still talking straight past each other? --ChrisSteinbach

I see your point. I have indeed been giving history more credit(?) than I should have, in the sense of being unquestioning of 'facts'. However, consider that in other fields, there is no limit to the generality and level of abstraction researchers will study. That's why we have the whole ladder of biology > biochemisty > chemistry > physics approximations > fundamental physics. In contrast, historians limit their research to a very base level. It's very important, just like experimental physics is crucial to physics, but there should be more. -- rk

I think the actual results of these attempts to build big systems out of history (Jaynes, Hegel, Marx,...) may be what discourages new attempts. It's similar to why we see very few of these "big" philosophical systems announced these days.

I don't think any of the people you refer to were historians. Jaynes was a psychologist and Marx was a political economist. I'm pretty sure that Hegel was a philosopher.

You give the impression that at some point in time, historians did try to build large models to explain history. You also give the impression that non-historians have stopped building large models to explain history. My understanding is the first is false and the second is completely false.

The organized study of history as something else than heroic chronicles isn't very old. But there have been numerous attempts to build models and systems out of the gathered material. Models have been successful in creating new and interesting viewpoints. But with bigger models the accuracy gets lower and the model becomes less useful. (AllModelsAreWrongSomeModelsAreUseful?).

In my opinion "science" shouldn't be a Boolean concept. The more something is repeatable and falsifiable, the more "sciency" it is. There are a lot fields, including software-engineering, that have to live with a lot of fuzziness and complexity because there's too many variables involved or they are not repeatable, at least in a way that isolates one or few variables at a time. But researches try their best to study patterns and perhaps some statistical analysis to tease out various factors. Other such fields include psychology, cosmology, and economics. I suspect that part of the reason for a "slow-down" of breakthrough discoveries (per 1860-1930 boom period) is that we already gathered the low-hanging fruit of science-friendly fields such as physics, but must now find ways to tame the messy fields to obtain the next set of breakthroughs. -t

Arguably, SoftwareEngineering isn't really a "science" per se, in the same way that (say) psychology or archaeology are sciences. We do not not discover the nature of the real world via SoftwareEngineering, except for those aspects revealed or caused directly caused by SoftwareEngineering itself. SoftwareEngineering is properly a field of engineering.

I consider engineering to be very similar if not the same as "micro-economics".

Similarly, ComputerScience isn't a "science" but a branch of mathematics. However, all these fields should be governed by scientific principles and standards of rationality, rigour, logic, and application of the ScientificMethod. They are all based in science, but they are scientific rather than sciences. As scientific fields go, SoftwareEngineering is much less fuzzy (and perhaps complex) than some scientific fields. Archaeology, for example, draws conclusions from the barest scraps of evidence -- like deducing the presence musical culture and history from finding a carved scrap of bone that might have been intended to be a percussive instrument, or concluding advanced spirituality from a decorative pattern hewn into a rock. Due to scant evidence, this is accepted within that field as being necessary. SoftwareEngineering is almost entirely tangible; as such it permits applying scientific approaches to almost every conclusion. It is only lack of funding, time, and sometimes agreement over terms and scope that impose limits on this.

Well, okay, I agree that it could be studied with more rigor, such as seeing which function size produces the most profitable company and/or highest user satisfaction. But, in practice it's been too expensive to get the necessary scale. The same could be said of economics: force multiple countries/provinces to use the exact same system except for one or two variables. Similarly, if cosmologists could re-launch thousands of universes and tweak the constants of each one.... But that's not going to happen in reality either. All these will hover in a limbo-land between science and art until it does happen.

See Also: DisciplineEnvy


View edit of October 5, 2009 or FindPage with title or text search