Abstract

Science presents us with two sorts of mystery: the mysteries of knowledge and the mysteries of understanding. Why the first occur is easy to understand. Until space craft orbited the moon, the far side of the moon remained a mystery to us simply because we could not see it. We lacked the means of gaining the requisite knowledge. It is less easy to understand why there should be mysteries of understanding. When we have such a wealth of empirical knowledge of quantum phenomena, how can it be that we experience persistent difficulties with understanding quantum mechanics?

Another way of posing the question is to ask, what is the role of theory in physics? What does theory add to the results of experiments? The integral part played by theory in physics and other sciences is often taken to mean that science is irredemably “theory laden”, and that therefore a realist account of its ontology can not be sustained. Scientific statements can not be seen as true or false only by virtue of the nature of the external world. Their truth-value has to be relativised to the theoretical framework in terms of which they are expressed.

I conjecture that a realist account of theory can be given in terms of what could be called “Fregean analysis”, and the accompanying thesis, “the Primacy of Analysis”. This thesis, which can be found in some of Frege’s writings, states that, expressed in terms of senses:

we start with thoughts, and then uncover component senses (for example of names and predicates) by the process of analysis.

This thesis has been criticised by Dummett, who argued that sentences are intrinsically composite. We build a sentence up out of our stock of components (and thereby in parallel build up the thought it expresses out of the corresponding senses of the components). This explains how we can understand sentences which have never been written or spoken before.

The counter to Dummett is to say that while his argument correctly describes fully articulated sentences, where we indeed build the thought from a pre-existing stock of components, what the Primacy of Analysis thesis is trying to describe is how we articulate in language some new state of affairs for which our existing understanding is inadequate. Dummett’s account explains how we grasp the sense of a sentence never expressed before. What is attempted here is to explain how we can express a thought never grasped before.

In extreme cases we are presented by the world with a fact which we can not articulate in words. The process of articulation consists of trying to tease out of the thought its components; the corresponding words then make up a vocabulary in terms of which we can talk about the new fact. Note that it is wrong to speak of the “unarticulated thought”, as if this is something without structure, or with only a partial structure. The thought is objective, and it is intrinsically complex. Grasping it consists in seeing at last what complexity was always in them from the very beginning. The description “unarticulated” refers not to an intrinsic property of the thought, but to our initial relationship to it.

We can not display examples of completely unarticulated thoughts; by definition they are the ones which can not be expressed in language. We can however recognise when thoughts are being expressed in language which is seriously inadequate for the job. Perhaps the best way to deal with this is to use what could be called filler words. These are words simply used to fill in a gap in our understanding, and to point to that gap. We could, for example, use the words “mind” and “… is mental” like this, not to express some understanding, but explicitly to point to the absence of understanding.

The counter-argument to the claim that all scientific discourse is inescapably theory-ridden is that empirical evidence can not only falsify statements within conceptual frameworks, it can destroy the framework as a whole. This rare event can be diagnosed when our whole attempt to describe some new phenomena disintegrates into perplexity and paradox. The classic example of this is provided by the advent of quantum mechanics. It is not merely that the old way of speaking became more difficult to use, or became unfashionable, rather it becomes quite untenable in the light of new facts.

This account of theory can be tested and refined by looking at examples from the history of science. We can distinguish between four types of advance in articulation.

The account of theory which emerges from these considerations is therefore the following. Experiment gives us initially unanalysed facts. We then have to articulate them, distill out of them the conceptual vocabulary (primarily a vocabulary of incomplete expressions) before we can say anything, true or false about them. This work is not pure creation on our part, but discovery of what was already there through the act of creation. This realist interpretation is supported by the proven ability of experimental facts, not just to falsify specific theories within a conceptual structure but also to cause the failure of the structure as a whole.

(Click here to go to the full essay.)