Richard Creath

Every Dogma has its day

This paper is a reexamination of “Two Dogmas” in the light of Quines ongoing debate with Carnap over analyticity. It shows, first, that ‘analytic’ is a technical term within Carnaps epistemology. As such it is intelligible, and Carnaps position can meet Quines objections. Second, it shows that the core of Quines objection is that he (Quine) has an alternative epistemology to advance, one which appears to make no room for analyticity. Finally, the paper shows that Quines alternative epistemology is itself open to very serious objections. Quine is not thereby refuted, but neither can Carnaps analyticity be dismissed as dogma.

The Unimportance of Semantics

Philosophers often divide Carnaps work into syntactic, semantic, and later periods, but this disguises the importance of his early syntactical writing. In Logical Syntax Carnap is a thoroughgoing conventionalist and pragmatist. Once we see that, it is easier to see as well that these views were retained throughout the rest of his life, that the breaks between periods are not as important as the continuities, and that our understanding of such Carnapian notions as analyticity and probability needs reevaluation.

Carnap’s intellectual development

ORIGINS The young Carnap is not easy to classify. He was neither really a scientist nor a proper philosopher. Among scientists he felt temperamentally at home, but he regretted the slovenliness of the enterprise. The officers at headquarters, he thought, needed to bring some order to operations on the front. Unfortunately these “officers” – the neo-Kantian philosophers whose lectures he attended in Jena and Freiburg before the First World War – seemed too unsure of the terrain to guide the scientific sappers in their spadework of intellectual trench warfare. But Carnap did not reject the neo-Kantian tradition he grew up in. He assimilated a good deal of it. The impulse for the revival of Kant in mid-nineteenth-century Germany had originally come from the natural sciences rather than philosophy, particularly from the great physicist and physiologist Hermann von Helmholtz. He and the philosophers who followed his lead had wanted to complete the job of eliminating metaphysics that Kant, in their view, had left unfinished. Helmholtz’s physiology of perception, they thought, could render the transcendental aesthetic metaphysically harmless. Though they held that the subjective feelings of spatiality and temporality are built into our perceptual system, just as Kant had argued, this did not mean that the geometry governing the perceived world was put there by human perception. We have no idea, Helmholtz said, whether physical space is Euclidean or non-Euclidean; this is an empirical question like any other, to be settled by going out and looking. So much, Helmholtz had thought, for Kant’s best example of supposedly synthetic a priori knowledge.

Carnap and Husserl

INTRODUCTION From a contemporary vantage point, the conjunction may appear puzzling. What could Carnap - anti-metaphysical logician, student and “legitimate successor” of Frege - possibly have in common with the founder of transcendental phenomenology? Yet, as Michael Dummett has observed, the German philosophy student of 1903 likely regarded Husserl and Frege as mathematician-philosophers remarkably similar in interests and outlook (Dummett, 1993, 26). To be sure, subsequent developments pushed apart the incipient programs of phenomenology and analytic philosophy. Husserl turned to “transcendental subjectivity” in 1905-1907, whereas analytic philosophy around 1930 took a “linguistic turn” precisely to distinguish its methods from those placing cognitive reliance upon intuition or individual subjectivity. Then again, the 1927 publication of Heideggers Being and Time inexorably changed perceptions of phenomenology as having acquired an expressly “existential” and “ontological” orientation, preempting and obscuring its original Husserlian impulses towards logic and the foundations of mathematics. Epitomizing this history in a memorable metaphor, Dummett notes that the respective influences of Frege, the “grandfather of analytic philosophy,” and Husserl, a patriarch of “continental philosophy,” run through twentieth-century philosophy like the Rhine and Danube, mighty rivers rising close together, briefly running parallel but then diverging to widely separate seas. Extending the metaphor a bit further, Carnap is a central current flowing into one of these seas, while Heidegger is the torrent surging into the other.

Tolerance and logicism: logical syntax and the philosophy of mathematics

The logical empiricism of the Vienna Circle presented itself as the reconciliation of a thoroughgoing empiricism as regards substantive knowledge with the certainty and necessity of mathematics. According to empiricism, sense experience is the only source of substantive knowledge of facts, but sense experience might always have been different from what it actually was. Moreover, anything we extrapolate from what we experience may be falsified by further experiences. Knowledge grounded in sense experience is thus knowledge of contingencies, and except perhaps for knowledge of what is immediately experienced (“Red here now”), it is revisable on the basis of further experiences. In contrast, knowledge of mathematics appears to be unrevisably certain substantive knowledge of necessary truths. The contingencies of our sense experience thus seem to afford no basis for knowledge of mathematics. In a word, empiricism must be false, because mathematics is substantive a priori knowledge. The logical empiricist approach to this dilemma is to deny that knowledge of mathematics is substantive knowledge. The apriority and necessity of mathematics is the apriority and necessity of logic. Logic is in its turn grounded in tacit conventions for the use of certain symbols, symbols that do not themselves stand for anything, like the signs for negation and conjunction. As Hans Hahn put it: [Logic] only deals with the way we talk about objects ; logic first comes into being through language. And the certainty and universal validity of a proposition of logic . . . flows precisely from this, that it says nothing about any objects . . .We learn – by training, as I should like to put it – to assign the designation “red” to some of these objects, and we make an agreement to assign the designation “not red” to any others. On the basis of this agreement we can now state the following proposition with absolute certainty: None of these objects is assigned both the designation “red” and the designation “not red,” which is usually expressed briefly as follows: No object is both red and not red.” (Hahn, 1933/1987, 29)

Carnap’s Move to Semantics: Gains and Losses

In 1931 Walter Sellar and Robert Yeatman published a delightfully silly history of England entitled 1066 and All That 2, as they said, “comprising, all the parts you can remember including one hundred and three good things, five bad kings, and two genuine dates”.3 History, they tell us, is not what you think; it is what you can remember. So their history is simplified and garbled, and the moral point is put front and center: every development is described as a good thing or a bad thing, a good king or a bad king. What makes 1066... comic is the cleverness of its insight into what confusions people actually have and the antic candor in giving us the moral point without wasting any time on dates, motivations, or any other such confusing historical details.

Was Carnap a Complete Verificationist in the Aufbau

It is argued that Carnap was not a complete verificationist in the Aufbau despite the widespread view that he was. That doctrine would be intrinsic to constructionalism only if either of two additional assumptions are made, and there is no reason to believe that Carnap made these assumptions. Further, in the Aufbau Carnap did not demand verifiability independently of constructionalism, and his clear rejection of verifiability in Pseudoproblems counts heavily against his ever having accepted it in the Aufbau.

The Gentle Strength of Tolerance: The Logical Syntax of Language and Carnap’s Philosophical Programme

Before Rudolf Carnap wrote The Logical Syntax of Language he was an important philosopher, a very important one. After he wrote it, he was a great one. And tolerance is his great idea; it is at the very centre of his philosophic programme from this moment on. With most philosophers a programme is a wellspring of ideas that occasionally crystallizes into a book. In this case, the order is reversed: the book produced the idea and the idea generated the programme.

The Role of History in Science

The case often made by scientists (and philosophers) against history and the history of science in particular is clear. Insofar as a field of study is historical as opposed to law-based, it is trivial. Insofar as a field attends to the past of science as opposed to current scientific issues, its efforts are derivative and, by diverting attention from acquiring new knowledge, deplorable. This case would be devastating if true, but it has almost everything almost exactly wrong. The study of history and the study of laws are not mutually exclusive, but unavoidably linked. Neither can be pursued without the other. Much the same can be said of the history of science. The history of science is neither a distraction from “real” science nor even merely a help to science. Rather, the history of science is an essential part of each science. Seeing that this is so requires a broader understanding of both history and science.

Are Dinosaurs Extinct

It is widely believed that empiricism, though once dominant, is now extinct. This turns out to be mistaken because of incorrect assumption about the initial dominance of logical empiricism and about the content and variety of logical empiricist views. In fact, prominent contemporary philosophers (Quine and Kuhn) who are thought to have demolished logical empiricism are shown to exhibit central views of the logical empiricists rather than having overthrown them.