PHILOSOPHY OF LANGUAGE QUNE S TWO DOGMAS OF EMPIRICISM LECTURE PROFESSOR JULIE YOO Why We Want an A/S Distinction The Two Projects of the Two Dogmas The Significance of Quine s Two Dogmas Negative Project: Discrediting the Two Dogmas Positive Project: Holism The First Dogma: The A/S Distinction Kant: Analyticity as Conceptual Containment Frege: Analyticity as Laws of Logic and Definition Verificationist Account of Analyticity The Second Dogma: Reductionism Atomism and Holism Empiricism Without the Dogmas Scepticism About Meaning? Quine s Two Dogmas of Empiricism Page 1 of 9
WHY WE WANT AN A/S DISTINCTION There s a reason why we want there to be an A/S distinction. Science is a matter of acquiring true synthetic statements: these would include laws of nature and relevant particular claims about the nature of our physical world. It makes scientific inquiry strictly a matter of discovery rather than arbitrary definition-mongering. Non-Scientific Example Here s an egregious instance of definition-mongering. Suppose you re looking for your car keys and you get your friend to help. Your friend shouts: FOUND YOUR CAR KEYS! but then she produces an old pair of socks. If your friend backs up her discovery by appealing to the following definition, then your friend will have made a correct discovery: car keys = dfn socks But surely this is cheating. What makes your search for car keys a meaningful enterprise is the assumption that car keys has a stable meaning across its multiple uses. Scientific Example This is even more pressing in the case of scientific discovery: THERE IS LIFE ON THE MOON! When we ask the scientist to back up this claim, we expect her to dish out some empirical evidence that will make it appropriate for our old meaning of life to apply to certain creatures on the Moon. But suppose the scientist appeals to the following definition: life = dfn (old definition) + rocks existing on the Moon Then the proclamation is true. But this decision to change the meaning of the word life hardly counts as growth in scientific knowledge. In order for there to be growth, our terms must have a stable meaning that we can apply and reapply in novel situations. In other words, our scientific terms must have a meaning component, that component that remains the same over their different applications, and an empirical component, that component that determines whether the sentences in which they appear are true or false depending upon how the world is. These components represent the A/S distinction. THE TWO PROJECTS OF TWO DOGMAS The Significance of Quine s Two Dogmas Quine s Two Dogmas of Empiricism, is one of the most influential papers in 20 th century analytic philosophy. There are several reasons for its enormous impact: Quine s Two Dogmas of Empiricism Page 2 of 9
! It overturned the once highly popular theory of verification.! It advanced an anti-foundationalist and anti-reductionist view of knowledge.! It integrated the pursuits of philosophical knowledge and scientific knowledge.! It called into question the very existence of meaning. Each one of these reasons themselves have other repercussions. Scepticism about meaning, for example, calls into question what kind of agenda and method is available to philosophy as well as the possibility of linguistic translation and the possibility of an intentional psychology. The paper engages in two projects, one negative and the other positive: Negative Project To expose and discredit the two dogmas of empiricism. First Dogma: Each empirical statement is made up of two components, a meaning component and a factual component. The meaning component is what corresponds to the intension or Fregean thought it expresses, and the factual component corresponds to the set of confirming or disconfirming observations. These components represent the A/S distinction. Second Dogma: Each meaningful statement has its own finite discrete fund of observations that confirm or disconfirm it. This assumes that sentences of a language are atomistic and ultimately reducible to discrete sets of observations. The two dogmas of empiricism are related. According to atomism, the truth or falsity of a sentence in one s language is to have no bearing upon the truth or falsity upon a logically unrelated sentence in one s language. But this kind of hermetic insularity can be conferred only if sentences can have clear and stable meanings, i.e., only if there are analyticities. Conversely, there can be analyticities only if (among other things) there are certain sentences that are not subject to observational confirmation or disconfirmation. Positive Project This is to finish the empiricist project purified of its dogmas. Here, Quine proposes an holistic theory about the confirmation and disconfirmation of scientific theories, and tries to demonstrate what empiricism is really like once it has been purged of its dogmas. Quine s Two Dogmas of Empiricism Page 3 of 9
THE FIRST DOGMA: THE A/S DISTINCTION To demonstrate the untenability of the first dogma, Quine paper proceeds in two stages: the first attempts to show that there is no precise characterization of analyticity, and the second attempts to construct an overall general argument against the possibility of analyticity. Positivists defined such statements as one that are true (or false) strictly in virtue of the meanings of the words in the sentence. S is analytic IFF S is true solely in virtue of the meanings of the words in S. This seems straight-forward enough, not too obscure or controversial. But if we want to get to the bottom of analyticity, which is what Quine is urging us to do, we need to look closely at this initial definition of analyticity. Notice that the definition contains the word meaning. As a part of a definition of analyticity, it cannot be invoked. The reason is that the notion of meaning itself is defined in terms of analyticity, and so cannot be used in a non-circular analysis of analyticity: X means Y if it s analytic that Y follows from X. This means we need to look elsewhere to ground the definition of analyticity. Is there a ground? Quine considers two proposals, one given by Kant and the given by Frege. Quine spends most of his time going over Frege s proposed explanation of analyticity, and we will too, but we will begin with a quick review of Kant s proposal. Kant: Analyticity As Conceptual Containment Kant attempted to explain analyticity in terms that seem quite natural: All sisters are siblings is analytic because the concept of being unmarried is contained within the concept of being a bachelor. So the idea here is that we explain analyticity in terms of conceptual containment. But while this may be a suggestive metaphor, we can t rest a serious analysis upon a metaphor. Kant s second attempt defines analyticity in terms of the contradictory status of the denial of an analytic claim: All bachelors are unmarried is analytic because its denial, All bachelors are married, is contradictory. But to make the analysis legitimate, we need to define a contradiction. And such a notion already presupposes analyticity. So this proposal won t do. Frege: Laws of Logic and Definition The notion of a logical truth can be defined strictly syntactically. This means that it can be defined without having to appeal to what any of the terms means (really?). So it can be a good candidate for explaining analyticity in non-circular terms. Can it furnish the needed explanation? S is analytic IFF S is a law of logic. A law of logic, or what Quine calls a logical truth, is a truth that is guaranteed by the logical terms (truth-functional connectives and quantifiers) appearing in the sentence: If then ; and ; or ; not ; all ; some. Their truth does not depend upon how Quine s Two Dogmas of Empiricism Page 4 of 9
we fill in the blanks, which are the non-logical components of the sentence. For instance, the following sentences are necessarily true, simply because of the way in which the logical connectives are involved in the sentence. a. If P, then P. If it s raining, then it s raining. If cats sing opera, then cats sing opera. b. Either P or not-p. Either it s raining or it s not raining. Either cats sing opera or cats don t sing opera. c. An un-x is not X. An unmarried man is not married. Unmusical cats are not musical. d. An X is an X. A female sibling is a sibling. An opera-singing cat is a cat. There are a potentially infinite number of logical laws (the axiomatic laws plus all their theorems). They are utterly boring and uninformative, and sensible people rarely go around saying things like this (except philosophers). A moment s reflection, however, will lead you to realize that this current explanation of analyticity is too narrow. Although logical truths may be analytically true, not all analytic truths are logically true. A sister is a female sibling has the logical structure of: Y is X. (Y = sister; X = female sibling) But sentences with this structure are not logically true. Substitutions of X and Y can generate sentences that are not necessarily true or even true at all. So we need to supplement our original definition of analyticity. Correct Definitions = Synonymy: Since logical laws alone don t get us very far, maybe we can appeal to the notion of a logical truth plus definition: analytic sentences are those truths that can be turned into logical truths by substituting terms with their definitions. S is analytic IFF S is a law of logic or S is derivable from laws of logic and definitions as premises. A standard definition of sister is female sibling. 1. A female sibling is a sibling. (logical truth) 2. A sister is a female sibling. (definition) 3. A sister is a sibling. (analytic truth from (1) and (2)) Surely not just any definition will do. How are we to determine the correct definitions? Maybe definitions are correct if the definiens and the definiendum are synonymous: Quine s Two Dogmas of Empiricism Page 5 of 9
sister = definition female sibling IFF sister is synonymous with female sibling. But what is it for terms to be synonymous? We can t take this notion for granted. It too must be explained in terms that don t ultimately appeal to analyticity. Synonymy and Substitution: Perhaps we can explain synonymy in terms of the Principle of Substitution (see Lecture 2 on Frege). According to this principle, Two expressions S and S* are intersubstitutable salva veritate IFF the replacement of S with S* in a sentence S (or S * with S in a sentence S *) always retains the same truth-value. We can use this principle to give us an analysis of synonymy: Two expressions S and S * are synonymous IFF S and S * are intersubstitutable salva veritate. This definition of synonymy, however, is too weak because synonymy requires intensional equivalence. Intersubstitutivity admits extensional equivalence, and so the definition above will erroneously will count all extensionally equivalent co-referring expressions as synonyms. But surely the following co-referring expressions are not synonyms: cordate (creature with a heart) and renate (creature with a kidney) water and H2O Synonyms: Substitution and Necessity: One way of trying to fix this is by appealing to necessary intersubstitution salva veritate. Two expressions S and S * are synonymous IFF S and S * are necessarily intersubstitutable salva veritate. But now we have to explain what it means to say, Necessarily P. Quine claims that this could me nothing more than: Necessarily P IFF P is analytic. And we have come full circle, again. Quine considers other candidates and finds them all wanting. He concludes that there is no adequate explanation of analyticity. VERIFICATIONIST ACCOUNT OF ANALYTICITY (AGAIN) We have gone over proposals suggested by Kant and by Frege and have found them both wanting. However, the Verificationist Theory of Meaning may hold out some hope. Quine s Two Dogmas of Empiricism Page 6 of 9
Remember that according to this theory, the meaning of a sentence consists in the method by which it can be empirically confirmed of disconfirmed. With this account of meaning, we can proceed to give an account of synonymy: S and S * are synonymous IFF their truth conditions are verifiable under exactly the same circumstances. With this grounding of synonymy, we can finally give an account of analyticity: S is analytic IFF S is derivable from the laws of logic and substitution of synonyms. In fact, we can simply cut to the chase and state that analytic sentences are those sentences that are confirmed no matter what the empirical facts turn out may be. Bachelors are unmarried will be true even if all bachelors were wiped out. As Quine has famously said, they are confirmed come what may, and immune to revision. In other words, analytic sentences are simply beyond the reach of empirical findings. They are beyond empirical reproach. THE DOGMA OF REDUCTIONISM Atomism and Holism This view of analyticity makes two assumptions: one, that all empirical sentences are reducible into sentences about sensory experiences, and two, that scientific confirmation is atomistic. Both assumptions, however, are deeply mistaken. The first assumption is the dogma of reductionism : The notion lingers that to each statement, or each synthetic statement, there is associated a unique range of possible sensory events such that the occurrence of any of them would add to the likelihood of truth of the statement, and that there is associated also another unique range of possible sensory events whose occurrence would detract form that likely hood. The dogma of reductionism survives in the supposition that each statement, taken in isolation from its fellows, can admit of confirmation or information at all. (p. 49) If some sentences are going to be treated as beyond empirical reproach, then they will have to be little individual packets of meanings that don t get affected in the even of any empirical discovery. In other words, this view of confirmation assumes that each sentence is confirmed has a meaning independently of its relationship to other sentences in a theory or a language, and that if can be disconfirmed on an individual basis. But this assumption, as Quine as argued, is naïve and false. But Quine s anti-reductionism leads him to claim: our statements about the external world face the tribunal of sense experience no individually but only as a corporate body. Quine s Two Dogmas of Empiricism Page 7 of 9
It s not individual sentences that are confirmed or disconfirmed. It is the theory as a whole. This is because hypotheses make sense only against the background of auxiliary hypotheses other chunks of the theory to which the hypothesis in question belongs. And each of these auxiliary hypotheses themselves make sense against the background of other hypotheses auxiliary for them, and so on. As a result, if there is what appears to be disconfirming evidence, it isn t only the hypothesis in question that may get revised but some other part of the theory. Any sentence, even It is raining in Toledo, can be held come what may. And conversely laws of logic can be subject to revision or rejection should it be expedient for the scientific theory to do so. In fact, several have suggested that we jettison the law of excluded middle to simplify strange findings in quantum mechanics. Empiricism Without the Dogmas Quine s argument against the A/S distinction forces us to reconsider the accuracy of a widely held picture of scientific and intellectual progress, which in philosophical circles, is known as scientific realism. According to this view, there is a world out there, which we attempt to describe by cataloging the things it contains and specifying how those things causally affect each other. A scientific theory is a body of statements that expresses our most rigorously tested beliefs concerning this catalogue of existing things and the relations they bear to each other. We construct, revise, and discard scientific theories as we accumulate knowledge by acquiring more and more true beliefs about the world. By acquiring more and more true beliefs, we get closer and closer to an accurate picture of what the world is like. But this view of the scientific enterprise gets its support from postulating a sharp distinction between the analytic and synthetic. Take a real case, the case of whales changing their status from fish to mammals. You have already had the privilege of thinking about the two ways of explaining this change: Empirical Discovery: The change in status is strictly a matter of empirical discovery. Science is a matter of describing mind-independent ontological categories. These are what philosophers call natural kinds. Being a mammal or being a fish represent some of the basic natural kinds of biology. These mind-independent facts about the world that there are natural kinds such as being a mammal or being a fish, as well as the fact that whales have lungs, bear their young in the womb, and so on are what force us to revise our previous conception of whales. Semantic Shift: Science is not a matter of cataloging mind-independent ontological kinds and their relations, not because we can know that there are no such things. Rather, the point is that even if there are such things, science is at best only in the business of organizing and making sense of the experiences we have of the world. So it is up to us whether we want to put whales in the category of fish rather than mammal, and it just so happens that we chose to place whales under the category of mammal to make our biological taxonomy simpler. Quine s Two Dogmas of Empiricism Page 8 of 9
There is not just one way of changing a conceptual scheme. Our experiences underdetermine what kinds of changes are mandated in light of the experience; we could have gone with whales as fish if we wanted, and the scheme that catalogues whales as fish is no less an accurate way of describing the world. The empirical discovery conception rests upon there being an A/S distinction. But in the absence of this distinction, we can t assume that scientific progress [growth in scientific knowledge] is a function of more and greater empirical discoveries, for empirical discoveries sometimes force us to alter the meanings we assign to our words, which then shapes the way in which we go about understanding, explaining, and describing the world. All that science gives us is a conceptual scheme, a framework that give us natural kinds with which to organize our experiences. But these natural kinds are not immutable (and if immutability is counted as among their essential features, then there are no natural kinds), since recalcitrant experience may force us to alter them should pragmatic considerations make that an expedient decision. SKEPTICISM ABOUT MEANING? We can trace the ever-globalizing trend in determining the basic unit of linguistic meaning: first it was individual terms (Locke and Hume), then it was individual sentences (Frege, Russell, Verificationists), and with Quine, we come to the entire language as the basic unit of empirical significance, i.e., the basic unit of meaning. But this way of putting things as a bit misleading, because Quine is actually a sceptic about meaning tout court. So strictly speaking, there is no such thing as the basic unit of meaning because there is no such thing as meaning! We can t underestimate the importance of Quine s conclusion if it is true. Grice and Strawson, early critics of Quine s Two Dogmas, explained the consequence this way: If talk of sentence synonymy is meaningless, then it seems that talk of sentences having a meaning at all must be meaningless too. For if it made sense to talk of a sentence having a meaning or meaning something, then presumably it would make sense to ask, What does it mean? And if it made sense to ask, What does it mean? of a sentence, then sentence-synonymy could be roughly defined as follows: Two sentences are synonymous if and only if any true answer to the question, What does it mean? asked of one of then is a true answer to the same question, asked of the other If we are to give up the notion of sentencesynonymy as senseless, we must give up the notion of sentence-significance (of a sentence having meaning) as senseless too. But then perhaps we might as well give up the notion of sense. Quine s assault upon analyticity is, then, an assault upon meaning. How do we communicate if meaning isn t what s being transmitted in our vocalizations, scribbles, and hand gestures? Quine s Two Dogmas of Empiricism Page 9 of 9