From WouterBeek.com

Contents 1 Introduction 2 Logical positivism 3 Popper 4 Quine 5 Kuhn 6 References

Introduction

In his penetrating 'Two Dogmas of Empiricism', Quine sought to replace the traditional notions of reduction and verification by yet another monistic notion, namely that of empiricism. The difference between the logical positivistic stance and Quine's does therefore not – as the commonly held conviction has come to falsely depict – mark a change from an absolutist view towards a more relativistic one. (This predicament will be reserved for the work of – among others – Kuhn and Feyerabend.) But the logical positivism-Quine axis runs along an entirely different dimension of multifarious ways of truth ascription.^[1]

Logical positivism

For the logical positivists, the total sphere of language was assumed to be known in advance. If I say something, then my statement thus uttered reaches throughout the whole of language, and - therefore - throughout the whole of reality (language and world are – Tractarian style – one). I therefore know, whenever I know a statement, which other statements are in accord or in conflict (as the case may be) with that very statement. So if I say P, then I automatically know this to be antagonistic to ~P, and this is not due to any form of further analysis of the world, but simply due to the character of our grammar. In a similar vein, generalized statements – whenever they are made – reach throughout the world, and are – again, due to grammar itself – linked up with all the particular statements that are in accord or conflict with it. So (Ax)Px is in accord with Pa and in disarray with ~Pb. Remark that this hold true regardless of the semantics of the symbols used, and this is exactly what was meant when I said that the interrelation among statements was an entirely grammatical one.

Popper

Popper, however radical his deviation from the logical positivist scheme might have been, never abandoned the notion of an overarching language that reaches throughout the total space of possible situations in the world. One might be reminded of the explicit negligence towards the concept of language one constantly seems to face in the work of Popper. But this explicit debunking of the potential problems that language might bestow upon the possession of a philosophically adequate image of the world, only serve to strengthen the recognition of the fact that for Popper language and thought so intricately run into one another, that the whole notion of semantical deception never surfaces in the first place.^[2] So although the point is never argued for nor against by Popper himself, it is clear that his contribution to the philosophy of science consist mainly of the insight that the universal quantifier be interchanged for the existential one (thus turning positive generalizations into negative particularizations). But what I would like to emphasize here, is that this very change in quantification presupposes that language is of exactly the same nature as it was in the logical positivistic movement (otherwise the transposition would not have been possible in the first place).

Quine

According to Quine, we cannot test a statement in isolation. So the notion an an experiment, i.e. what an experiment amounts to, changes in a most radical way here. This means that the meaning of a sentence is not to be identified in its direct relation to reality. In other words, the semantic structure of the sentence does not reflect – at least not in a direct and unproblematic manner – the way in which the world is structured. There is a dissonance of structures here. The meaning of a sentence changes accordingly, and is now only quite loosely related to the world.

Quine's position is often summarized as stating 'that the notions of apriori and aposteriori do not exist'. But this most completely misses the point that he in fact tries to covey. For the point is not that the notion of either aprioricity, nor that of aposterioricity, is in any way misstated. The point is that the set of sentences thus defined – as those that hold true regardless of the world, respectively those that hold true regardless of grammar – are just denoting the empty set. This should not be taken in a wrong sense though, as to suggest that these notions do not apply to sentences at all. For in fact they do apply, along the board. We only have no criterion whatsoever to delineate the one component from the other. For in order to do that, we would have to possess a 'birds eye view' of the world, i.e. an epistemological stance towards the world that is not mediated by language.^[3] But even though the distinction between these two components cannot be made, they nevertheless do exist. Otherwise the entire notion of empiricism would become fundamentally flawed, and this is what relativism – which is thus clearly to be distinguished from Quine's position – eventually amounts to.

The Quinian point is then that we cannot delineate world from grammar. Because language is no longer assumed to be in observable sync with the world, even empirical observations can no longer be taken to be in sync with the world either. For what I might think of as an empirical observation, e.g. that this is the blood of Jesus Christ, might – actually – turn out to be quite wrong, e.g. it actually is red whine (and just that). My failure, in this example, is explained as due to the way in which my sense perceptions are embedded within the total fabric of my grammar.

But even though language does not show us, in any way whatsoever, which components of my thoughts are held in place by the solidity of reality, and which parts are strings that run through and onto that reality, as the threads of human rationality and human grammar; in other words, even though language masks the empirical component, this component must still be there, somewhere. And this is not a leap of faith, but just the definition of what 'empiricism' means. (And this is the main reason why – again – the notion of relativism does not occur to Quine, and it shouldn't occur to anyone else, at least not at this point.) But still language is taken as one, i.e. as a monisticly structured whole.

Kuhn

In Kuhn there is a remarkable shift that cannot be explained as the next step on the weary road characterized by the successive theoretical refinements of Carnap, Popper, and Quine. For although the way in which an experiment was allowed to take place, or – in other words – the way in which observation statements were considered to be supportive and/or refutative for hypothetical or general statements; so even though various differences for establishing the truth of a scientific statement within language among the above mentioned philosophers of science existed, the notion of language itself remained unquestioned. But now, in the work of Kuhn, and as a late exponent of the linguistic turn, the view on language as exhibiting a totally known monistic system no longer holds.

The claim is not that whenever I utter a sentence P, I do not necessarily therewith reach throughout the whole of a certain language. So I do not contradict that by positing the general statement (Ax)Px, all of the Pa, Pb, etc. are postulated to stand in a certain relation. But my claim, in accordance with Kuhn, is that not everything that stands in a certain affirmative or negative relation to (Ax)Px is of the form (λx)(Px v ~Px).

A scientist says: “The hypothesis with regard to x, belonging to a certain domain D, is of the form (Ax)Px.”, meaning that a thing's characteristic description might be determined by checking against the occurrence of a certain property P, as defined over D. But then another scientist might say exactly the same thing, and yet mean something entirely different by it, by implicating a different domain of discourse (e.g. 'phlogiston' instead of 'atoms'). And what does that mean? The notions of P, x, etc. work relative to an interpreted model. But since the conception of how the domain looks like is itself the subject of scientific investigation and scientific-revolutionary overthrow, no set of symbols can be taken as constituting the rationally justified and normatively vindicated total sphere of grammar.

Whenever different hypotheses (or one hypothesis and one observation) reside within different grammatical spheres, there is no way in which both of them might be interrelated. “Well, except by construing a new grammatical sphere in which both of them occur and are interrelated.”, you might say. And this is true, but in that case the interrelation process of the two statements is no longer logically given, but – instead – arbitrarily construed.

References

1. ↑ Please remark that we shall, for simplicities' sake, equate the outcomes of an experiment to the singular observation statements in which they are uttered. Although this connection is not unproblematic, it would run too deep into the lattice of theoretical complexity.
2. ↑ It is no coincidence therefore that Popper uses the citation of Kant that most poignantly repudiates the ideas of the originator of the true linguistic turn, Johann Georg Hamann:

   “I for my part hold the very opposite opinion, and I assert that whenever a dispute has raged for any length of time, especially in philosophy, there was, at the bottom of it, never a problem about mere words, but always a genuine problem about things.”
   ['Preface to the first edition, 1934.' In: The Logic of Scientific Discovery, p. xv]

3. ↑ It is evident that we do not have such a point of view. And this is not to be equated to the dreary and pointless remark that 'all thinking is linguistic', for it is perfectly clear that even animals – who are for convenience's sake assumed to not possess any form of language – exhibit quite intricate cognitive activities that become manifest in tandem to their quite intricate forms of behaviour. But the point should be taken far less wobbly, namely as the pure fact that although certain forms of thought might be non-linguistic, this doesn't help us in any epistemological way, since the cogwheels of rationality and truth ascription do not operate in a non-linguistic environment.