The Keynesian world view: deterministic uniformity in the macro world versus the uncertainty of an individual unit in the micro world
It may be necessary for us to take a roundabout journey in order to shed light on the Keynesian 'parts-whole' question from an interdisciplinary point of view.
Let us begin by looking at Rudolf Carnap's Philosophical Foundations of Physics (1966). As for the deterministic view of science in the nineteenth century, he states:
Nineteenth century physicists and many philosophers as well took for granted that, behind all the macrolaws, with their inescapable errors of measurement, are microlaws that are exact and deterministic ... The behavior of molecules must depend on something. It cannot be arbitrary and haphazard. The basic laws of physics must be deterministic.
(Carnap, 1966, p. 281)
In fact, certain scholars conceive of nineteenth-century natural science, in particular physics, as a typical model and reconstruct moral science on the basis of it. They would probably share the confidence of nineteenth-century physicists described above. It seems to me that such confidence forms the background to their adopting an atomic hypothesis in the field of moral sciences. I think it is not unreasonable to say that a striking example of that attitude is neoclassical economic theory as it emerged from the 'marginal revolution', particularly the Walrasian scheme of general economic equilibrium in its complete logical consistency.
Compared with the history of natural sciences, Keynes in TP might occupy the intermediate place between nineteenth-century classical physics and twentieth-century modern physics. The reason for this is that, from the standpoint of a theory of knowledge, Keynes is interested in the uncertainty involved in the behaviour of individuals - 'the twilight, as I may so say, of Probability' (Locke's words as quoted by Keynes) - and also in the doubtful and inconclusive feature of their judgement. In short, he was exclusively concerned with the rational but indeterministic character of individual consciousness. In broad terms, it may be said that Keynes rooted 'the principle of uncertainty' in the ground of moral science by proposing a new logic of probability.
It may not, however, be unreasonable to think that Keynes himself came to hold doubts about his own standpoint concerning 'science and universe' once the implications of the 'doctrine of relativity' are considered. However, he would not enter a discussion of this issue, and his doubt remains confined to a footnote of TP: '[i]s this interpretation of the principle of the uniformity of nature affected by the doctrine of relativity?' (TP, p. 276n; emphasis added).
It would seem that what 'this interpretation of the principle of the uniformity of nature' entails, is the proposition that 'it is in respect of such positions in time or space that "nature" is supposed' (TP, p. 252). Briefly put, it can be traced back to the view of nature in classical mechanics, according to which, given the position and momentum of a certain elementary body in space at a point of time, we can exactly predict the position (and momentum) of it at any future point of time by relying on some fundamental laws connecting those bodies.
When, following Cambridge philosophers like Bertrand Russell, Ludwig Wittgenstein and George Edward Moore, Keynes endeavours to picture the 'universe', he assumes, I think, the world of Newtonian classical physics. If so, his interpretation of 'the uniformity of nature', or rather of 'the atomic uniformity' must be radically influenced by the emergence of modern physics, as he himself anticipated with uneasy feeling.
However, when Keynes comes to grips with the human world, he is mainly concerned with the uncertainty of human conduct. This kind of uncertainty originates from rational but inconclusive judgement arrived at by virtue of probable inference in decision-making processes. And it is by reference to the properties of probable arguments that Keynes endeavours to secure the foundation of a new type of probability.
To sum up, a conspicuous feature of Keynes's thought is that there exists a unidirectional and deterministic uniformity (or regularity) in the macroeconomic world. This uniformity is explicitly based on the principle of effective demand, the core of which is the multiplier mechanism. By contrast, we meet an ambiguous and uncertain multiplicity within the domain of individuals' activities. Those activities cannot be reduced to the scheme of rational behaviour assumed by neoclassical economists, except within a given narrow field.14 In short, Keynes's thoughts seem to show an ambivalent, dual characteristic. In so far as Keynes's ideas as a system are concerned, a deep gulf still lies between the 'whole' of the system and its component 'parts'.
Our next step consists in discussing the views of eminent scientists and scholars in order to find a way to critically reorient Keynes's theory so as to highlight his unsettled problematique.