Popper's falsification is useful, but may not be a truly universal test. It seems most ineffective in the case of systems theories, of which complexity research (closely aligned with Turchin's work), systems theory, cybernetics, and evolutionary theories and models are several examples.
Few of these produce reliable forward-looking predictions, most especially for anticipating specific states achieved at specific times. Much of this hinges on simple complexity and measurement error (uncertainty over initial conditions).
What these models do provide, though, are useful models for drawing associations, predicting or describing specific patterns, and establishing the likely existence of specific features or elements of a system.
Darwin predated concepts of genes, chromosomes, DNA, and mutation mechanism. But his notions of variation and inheritance presupposed these and outlined their general characteristics. Later discoveries conformed with, and further refined, these predictions.
Where humanities and social sciences (Turchin's cliodynamics somewhat spans this divide) run into additional complexity is that models and predictions themselves become part of the informational context in which individuals, groups, and societies see, analyse, process, express, and respond to the informational environment.
This is often compared to Heisenberg uncertainty, but this is a category error. Heisenberg is addressing fundamental knowability, where the act of ascertaining system state changes the state itself. Epistemic interactions affect not the state itself, but the informational loop associated with it. The model and derived understanding are part of the sensing, processing, learning, memory, communications, and control interactions --- the model operates within the modeled system, endogenously rather than exogenously.
How to incorporate this distinction into philosophy of science is an interesting problem, though I don't think it is impossible. Pseudoscience lacks not only falsifiability, but generally, consistency and causality. Sound systems theories should be consistent and denote causality.
You're mixing up Heisenberg's uncertainty principle and the observer effect. The former limits the precision at which certain linked properties can be known.
> Historically, the uncertainty principle has been confused with a related effect in physics, called the observer effect, which notes that measurements of certain systems cannot be made without affecting the system, that is, without changing something in a system.
Given that the case I had in mind (On the Media's Bob Garfield) specifically mentioned Heisenberg in his discussion, concerning of political polling, the error if any doesn't originate with me:
BOB GARFIELD What you're saying seems to be describing the Heisenberg effect, where the act of measurement actually affects the experiment itself. You believe that voters are actually behaving influenced in part by what they're reading in the polls and particularly the models.
> The mechanism of the HUP resembles the OE in parts at least.
OE was initially proposed as a mechanism behind the uncertainty principle, but it's actually just an attribute of the wave nature of quantum physics:
> Heisenberg utilized such an observer effect at the quantum level (see below) as a physical "explanation" of quantum uncertainty. It has since become clearer, however, that the uncertainty principle is inherent in the properties of all wave-like systems, and that it arises in quantum mechanics simply due to the matter wave nature of all quantum objects.
To your point, I believe this why is we need to be careful when describing things like many fields of social science as a “science”. It’s more “social educated guesses” than real science.
Social sciences demanding to be taken as concrete as rigid sciences is a problem today I think. We even see them dismissing actual science in some cases not through falsification but through using their own theory to “prove” certain science is “problematic” of heterodox therefore restricting knowledge.
They demand to be taken as seriously, not as concrete, in my experience. Which is fair, we can maybe reach something like what is described in this article eventually - but as Turchin himself put it, it would not exist without History. If we are to devalue social sciences, we are giving up on ever expanding that knowledge.
Should we take religions seriously then? They demand it anyhow.
I do not mean to dismiss the entirety of the humanities or social sciences (or religion for that matter!) I do get concerned however with their motivations and the power they wield over public policy and politics in general. A certain claim to the truth they don’t really have.
I also find it curious that most elite universities and scholars come to more or less the same conclusions regarding subjects that aren’t falsifiable. You’d expect this to be the case in the hard sciences (reproducibility is vital) but not in others. For instance, in religions we can see great divergence. Even in the way economic theory (another thing that isn’t science but often claims to be) has differing conclusions based on the same observations.
I also see today in the application of some of these social sciences that asking to debate or to ask for more evidence is taken as a challenge to it and in fact evidence that the theory is right. See some national best selling books from this year, for example.
At what point do social sciences look more like religion than “science”?
Few of these produce reliable forward-looking predictions, most especially for anticipating specific states achieved at specific times. Much of this hinges on simple complexity and measurement error (uncertainty over initial conditions).
What these models do provide, though, are useful models for drawing associations, predicting or describing specific patterns, and establishing the likely existence of specific features or elements of a system.
Darwin predated concepts of genes, chromosomes, DNA, and mutation mechanism. But his notions of variation and inheritance presupposed these and outlined their general characteristics. Later discoveries conformed with, and further refined, these predictions.
Where humanities and social sciences (Turchin's cliodynamics somewhat spans this divide) run into additional complexity is that models and predictions themselves become part of the informational context in which individuals, groups, and societies see, analyse, process, express, and respond to the informational environment.
This is often compared to Heisenberg uncertainty, but this is a category error. Heisenberg is addressing fundamental knowability, where the act of ascertaining system state changes the state itself. Epistemic interactions affect not the state itself, but the informational loop associated with it. The model and derived understanding are part of the sensing, processing, learning, memory, communications, and control interactions --- the model operates within the modeled system, endogenously rather than exogenously.
How to incorporate this distinction into philosophy of science is an interesting problem, though I don't think it is impossible. Pseudoscience lacks not only falsifiability, but generally, consistency and causality. Sound systems theories should be consistent and denote causality.