I’ve mentioned in many posts that I have been leisurely reading Popper’s ‘Logic of Scientific Discovery’. The further I go into it, the more I come to see his Vienna roots.
I admit something. I don’t know too much about Popper. I’ve always heard of Popper from second hand admirers among my sociologist friends who see him as one of the great heroes of LSE’s philsophical heritage (also included in that list are Lakatos and Feyrabend). I think perhaps I have misapprehended popper as a mid-late 20thC philosopher of science, the not-so-distant ancestors to the philosophers of science I had come to know in Bristol.
My Kant scholar approach is informed by a very minor understanding of the Vienna philosophers; I see them as having a distinctly Kantian heritage, despite their apparent anti-Kantian character. This route also follows the scholarship of the likes of Ernst Cassirer; who was for one note contemporamous to the great Vienna men. A more contemporary historian pertinent to this issue is Michael Friedman (I should add a tag as to how many references to Friedman this blog has).
I believe that Reichenbach’s notion of the constitutive a priori directly relates to what is understood as Kant’s ‘regulative a priori’ in his body of philosophy of science. As I have also considered; I think that the ‘foundations’ approach is a step forward and the supreme example of this, Carnap’s Aufbau, is a superior attempt at the philosophical foundations of what I think Kant’s desiderata for any future scientific theory must amount to. Did Kant anticipate a Carnap? That’s a question for another post.
For various exegetical reasons, I have maintained that Kant has set a series of conditions for what he called ‘science proper’ which included:
1. Reductionism: all propositions of science must relate to another by means of a higher order system to explain the principles of the lower order. Applied physics requires theoretical physics, theoretical physics requires calculus and calculus in turn requires certain metaphysical/mathematical foundation assumptions to allow the conditions of such to susue.
2. Relata by laws, higher and lower. A revised way of stating the above is to note that Kant (may have) believed that the relata of this higher lower ordering system that justifies higher and lower discourses takes place by explicitly formal laws; these laws must be formal insofar as they actually apply.
3. Systematicity, or the unity of science. These principles must form a system that relates one kind of physical phenomenon with anohter. A general physics must account for all of its constituent parts; matter, heat, sound, energy, motion etc. If we had a theory of planetary motion (such as astrology) that made predictions and claims about a person’s wellbeing and future; it would only be a respectable theory (well, one reason at least) if it had something to say or relate to say, the notion of elements; heat; energy; conservation of mass, and they all in turn were candidates or subsets of a higher formal system; say, a mathematics of change (calculus) or general principles that took place over it as a genera, and yet formed the foundations of the given discourse astrology.
Popper’s vision of science seems distinctly Kantian as he says the following:
1. We must understand scientific principles as universal claims; and even thought there are doubts about the validation of every instance of F=ma (due to the limitations ofthe universe and our own means of testability – namely, we can’t live forever and there is a great dataset of cases that we can no longer verify because they belong in the past). It is better, to crudely state it; to be a universals advocate of the sentences of science and to underlay them as a formal substructure of sentences; than to take the empiricist vision of dealing only with particulars.
2. Particulars cannot take us far enough with induction. Induction has its own epistemological and psychological problems. Which is our motivation to take a more rationalist ‘foundations’ approach.
3. The formal structure of science seems to take great precedence in Popper’s theory. It does not matter that our formal laws cannot map to real observances, or even that real observances cannot map to their porpositional expression. An example of the latter is this: to say ‘I am hurt’ is of very detached relevance to the felt experience of pain. The very fact of your reading this shows that this message is coded in a propositional format. Perhaps the best way to communicate the fundamental inexpressability of those things that propositions relate to would be to make an example of a non-propositional utterance (‘ouch’), or to hit you.
4. The ‘alms ob’ premise. Kant takes to the notion that while we may always have doubts about whether the actual formalisation of science maintains true of the world; we must take to the transcendental assumption that the world as it appears to us is just that; we must take some things for face value, or understand things in the way that we are forced to understand it; because of the brute fact natureo f our understanding to make us think it so. The categories of the understanding such as causation or orientation are things we understand just because they are so; but it is always valid to question them.
Popper takes quite well to what some commentators (I think it’s Bennett and HW Cassirer) call the ‘as if’ premise. If we are to take to certain assumptions or theorisations of science as idealisations or things we may imagine are ‘as if they were true’.
5. The conditions of possibility. Popper holds that the systematisation of science requires what I’ll call a mother premise, namely; that there is one fundamental condition at the top of a scientific theory which acts as a decider of what the other norms are. In a sense, Popper provides the notion of what I sometimes call a metanorm; a norm that decides norms.
I’ve specifically interpreted systematicity to be a metanorm; although Popper does not seem to be directing systematicity; he does make one proviso that relates to it; that we have an ordering langauge that makes sense of instances of particulars by instituting language terms that denote them as type. If we are to pose ‘marsupial’ as a universal we will always have the problem that universals cannot be justified by every instance; but the admission of this objection seems to be treated by popper as some kind of merit of his theory; as it is the same problem that induction has, but induction does not pretend to aspire to universals, because it cant. Some form of idealisation, or a language of types that in turn are ordered by ranking principles; e.g. the Linnaean hierarchical taxonomy.
Popper makes a point of not confusing universal types with sets and classes; perhaps this is something a Kantian theory of science may benefit from. We may have entities that have one class description; but fit multiple classes and sets. Kangeroo is a marsupial >> mammal; but it is also under classses and sets such as ‘australasian’; creatures that keep their children in pouches; ‘cute and cuddly’ or ‘parodied in family guy’ and they can be classified as such without inconsistency.
The relationships with Kantian systematicity and Popper’s vision of science are very intricate and subtle. I’ve just written this post to acknowledge just some of the potential issues and exegetical similarities that they have.
I’m also pleasantly surprised of Popper’s outlook; perhaps I may take his scientific character to be more Viennese than I thought.
6. Demarcation as the condition of science. Like Kant, Popper talks about ‘proper science’ (but not in those words). The problem of science is not of induction (contra Humean thinkers); but demarcation; namely, the separation of science and non-science; science and metaphysics; epistemology qua knowledge and scientific method qua pseudo/philosophical psychology.
Demarcation acts as a methodological primary premise in a similar manner to how Kant’s notion of ‘conditions of possibility’ works.