Karl Popper is undoubtedly one of the leading philosophers of science in the post modern world. His influence, however, is not restricted to this particular discipline. Bryan Magee says, “The range of Popper’s intellectual influence, unapproached by that of any English-speaking philosopher now living, extends from members of government to art historians.”[1]

                  The importance of Popper, especially viewed from the vantage point of a theist, is that he attempts to reject two major philosophical theses--Positivism and Relativism. The former precludes metaphysics and thereby Biblical theism altogether; while the latter eliminates any possibility of possessing a knowledge of having any “ultimate truth.”


                  This paper is only one of three papers dealing with Popper’s attack on Positivism. Popper’s relationship to positivism has often been misapprehended and therefore misrepresented. Because Popper dealt with many of the same categories as did the positivists, they often allied him closely to themselves, and insisted that he was not so much different from them as he himself claimed. However, Otto Neurath, who was a member of the Vienna Circle, nicknamed Popper “the Official Opposition.”[2] It was Popper’s German work, Logik der Forschung (1935) that contains “the chief of what have since become the generally accepted arguments against logical positivism.”[3]

                  It is the general purpose of the other two papers to discuss more specifically Popper’s criticisms of logical positivism. But it is widely known that logical positivism in its classical form is a thing of the past. Copleston states that “logical positivism in an explicitly and clearly defined form has passed into history.” John Passmore speaks of the sect as having “disintegrated.”[4] Therefore, this paper will deal more particularly with some of the historical developments that have occurred through the logical positivist movement.


                  The purpose of the survey is not to minimize Popper’s criticisms of logical positivism, but to recognize that logical positivism[5] has made criticisms of itself resulting in reformulations of its original theses. Clearly, criticisms of those theses which have long since been abandoned are not adequate in themselves. The changes have been many and frequent within logical positivism. Kraft says that because of rapid development, many changes of views were involved and that “it was to be expected that several of the original, oversimplified doctrines would in time be superseded.”[6] It is unfortunate that it was almost twenty-five years before the English translation (1959) of Popper’s Logik der Forschung was published. In the preface to this English edition he speaks more to the inadequacies of language analysis as a solution to the genuine philosophical problems--whether it be ordinary language analysis, or attempts at constructing ideal languages.[7]


                  The history of logical positivism is quite difficult to survey. For the views of its proponents have in some cases changed radically. Also, there have often been fundamental disagreements among its own members. Bergman points out that this lack of agreement results greatly from the fact that the writers who are called logical positivists “derive intellectually and, in most cases, biographically from one of two centers, the Cambridge School of Analysis and the Vienna Circle.”[8] Hunnex also points out these differences and cautions the critics of analytic philosophy not to refer


to the whole as “positivism.” He says,


. . . to speak of all analytic philosophy as ‘positivism’ is misleading and ignores the important differences between ideal language analysis and ordinary language analysis on the one hand as well as important differences between the Cambridge and Oxford analysis on the other hand.[9]


                  It is not within the scope of this paper to discuss all of the developments and variations in logical positivism. But the one thesis most central to logical positivism--and the thesis acutely attacked by Popper--is the thesis of “verification.” This paper will deal not so much with Popper’s criticism of the verification principle but with the major historical developments this principle has been forced to take. Hempel says, “The concept of testability, which is to render precise the vague notion of being based--or rather baseable--on experience, has undergone several modifications which reflect an increasingly refined analysis of the structure of empirical knowledge.”[10]




The Background of The Verification Principle


                  The validity of metaphysics was poignantly raised by Kant as well as Hume and other philosophers. For them, metaphysics was a land that would leave human understanding in the state of contradiction.[11] The problem with metaphysics is that it did not produce any real, ultimate solutions to problems that had been raised for centuries. Kant asked, “If it [metaphysics] be science, how is it that it cannot, like other sciences, obtain universal and lasting recognition?”[12] He said that even the ignorant could deliver a final verdict in this area of metaphysics for “in this domain there is as yet no standard weight and measure to distinguish sound knowledge from shallow talk.”[13] It was Kant’s attempt to establish some system of weights and measures for metaphysics that was a hallmark in modern philosophy. However, his system, too, failed to “obtain universal and lasting recognition.”


                  The logical positivists adopted several of Kant’s technical terms, among which were analytic, synthetic, a priori and a posteriori. “Kant used them in an effort to reform metaphysics, but the positivists used them in an effort to eliminate it.”[14] The positivists took analytic statements to be necessarily true, for their denials would be contradictions. The truth of these statements was established a priori. Examples of these were to be seen in logic and mathematics, (eg., 2+3=5) and in grammatical constructions that were true by virtue of definition of terms (eg., all bachelors are unmarried men). Analytic statements were seen as statements that do not convey factual information about the world, for their truth is determined without reference to it.


                  For the positivists, there was only one other class of statements that were capable of being true--synthetic statements. In contradistinction with analytic statements, synthetic statements (1) could be either true or false; (2) were to be judged as to their truth or falsity by an a posteriori evaluation; and (3) conveyed factual information.


                  Aided by the Logical Atomism of Wittgenstein and Russell, the positivists uncovered a type of statement that appeared to be synthetic but one that actually conveyed no information about the states of affairs in the world. This kind of statement they termed “pseudo-synthetic.” For the positivists, it was the verification principle that determined whether a statement was “synthetic” or “pseudo-synthetic.” If there were no means by which a statement could be verified, then that statement was viewed as a “pseudo-statement.”


                  With the additional aid of Wittgenstein, the Logical Positivists introduced a new, third category. For pseudo-statements were to be seen as being neither true nor false, but “meaningless.” This is one of the points on which Wittgenstein had great influence upon the Vienna Circle, even though he was not a member of it. In his Tractatus (1921), he presented the idea that metaphysics arises because philosophers attempt to say what cannot be said; they “do not understand ‘the logic of our language’.”[15] For Wittgenstein, any attempt to go beyond “the limits of language” results in statements that are neither true nor false, but meaningless. It is easy to see why the Vienna Circle readily accepted Wittgenstein’s Tractatus as closely allied with their own intentions.”[16]


                  For the Vienna Circle, then, if a statement could not be verified empirically, it was “factually meaningless.” Their conclusion was that if philosophy was to be considered a genuine branch of knowledge, it must divorce itself from metaphysics. Their unique contribution on this point, according to Ayer, was that they made the impossibility of metaphysics “depend not upon the nature of what could be known but upon the nature of what could be said.”

The Verification Theory of Meaning


                  As has been shown above, the intention of the logical positivists was to set forth some criterion which could determine which statements were meaningful and which were meaningless. According to them, any synthetic statement which could not be verified by experience was not a candidate for being a meaningful proposition. This was a development from the earlier empiricists such as Mill and Spencer. According to their “traditional empiricism” even the statements of mathematics and logic must have an empirical basis. But the Vienna Circle combined empiricism with the insight that mathematics and logic were a priori, i.e., not based in experience. Kraft says,


This constitutes a fundamental revision of empiricism, the earlier claim of empiricism to derive all knowledge and science from experience as the sole ground of validity is thus abandoned. The empiricist thesis now restricts itself to factual knowledge. All synthetic judgments are to be validated by experience; there is no other way of establishing their validity.[17]


                  The criterion of demarcation for the logical positivists then was the Verification Theory of Meaning. This principle is technically not a criterion of truth, but a criterion of meaning.[18] The principle was popularized in English philosophy by A.J. Ayer in the first edition of the book, Language, Truth and Logic (1936) in the form: “The meaning of a proposition is the method of its verification.” Any statement not logically capable of being verified by experience was asserted to be meaningless by this criterion. Thus, it appeared as though all metaphysical statements were, in Hume’s terminology, worthy only to be “committed to the flames.”




                  As it was originally formulated, the verification principle was very restrictive. In the early days of the Vienna Circle the concept was presented that complete verification was required. This meant that, when seen in this highly restricted sense, complete verification was possible only to the observer himself. Any statements about the remote past or the distant future had to be classified as meaningless. The problem of complete verification became quite intense over the nature and status of the so-called “protocol statements.”


Protocol and Elementary Statements


                  For verification to occur, there must be some adequate basis upon which to make a verification. To the early Logical Positivists, it seemed that some apodeictic foundation must be laid for verifying observations. Such a foundation was initially seen in the protocol statements. At first, these were viewed as sentences which themselves were in need of no confirmation.[19] Rather, they were seen as statements which “provided a touchstone by reference to which all other statements could be empirically verified.”[20] The early view was that it was through protocol statements that language, as it were, made contact with fact and that they alone were verified directly: all other empirical statements were verified indirectly through them.[21] All meaningful statements, then, were required to be capable of reduction to these protocol statements. Establishing such a reduction system was Carnap’s major task in his theory of constitution.[22] Kraft refers to this goal of the logical positivists when he says,


Just as the Vienna Circle faced the clarification of the content of empirical concepts through their reduction to data of experience as a fundamental task of empiricism, so they took upon themselves the further task of clarifying the content and validity of empirical propositions through their reduction to elementary propositions.[23]


                  It was Wittgenstein once again that had a good deal of influence concerning this issue. He had accepted Russell’s division between simple and compound propositions, and had set forth the general approach to “atomic” and “molecular” propositions. It was Wittgenstein’s new insight that “the truth-value of compound propositions depends exclusively on the truth-values of the simple propositions which are their components; the former are ‘truth-functions’ of the latter.”[24] Therefore, it was the truth-value of the simple (atomic) propositions that alone mattered. The task was then undertaken to find the atomic propositions and to determine them on the basis of their logical form. Wittgenstein identified them with the term “elementary propositions.” These propositions were considered to be ones which could be immediately compared with reality, i.e., with the data of experience. For Wittgenstein, all empirical (and therefore meaningful) propositions must be capable of being reduced to elementary propositions, unless they were already such.


                  Since these elementary or atomic propositions were considered to be assertions about experiences, the Vienna Circle assumed that they must be found among the “protocol sentences.” Kraft says “Protocol sentences are supposed to describe the simplest knowable states of affairs, and not to contain any sentences obtained by interpretation of the given. They are supposed, therefore, to designate the immediately given.”[25]


                  But there was no unanimity concerning what statements satisfied these requirements and several members of the Circle disagreed over the concept of the “given.” And the status of these statements being absolutely certain was quickly challenged even by members of the Circle itself. Neurath claimed that even protocol sentences could be declared invalid under certain circumstances and that such sentences were never without interpretive content.[26] Neurath insisted that no statements could really be compared with the “given” at all but that statements could only be compared with one another. Carnap was influenced by Neurath to adopt a similar position on this issue. They held that protocol statements really had no privileged status at all among empirical statements since it was impossible to compare any statement directly with reality.


                  Both Neurath and Carnap challenged the former Wittgensteinian concept of elementary statements. Rather than being statements of certainty that referred to “private incommunicable experiences”--thus raising the problem of solipsism--these statements referred to public physical events which were to be intersubjectively verified. There statements were seen now as “physical statements.”[27] The result was that the truth of these co-called protocol statements was open to question in the same way as was the truth of other physical statements. They were all hypothetical in nature. One can readily understand the disastrous effect this move had on the original and strict version of the verification principle: no synthetic statement whatsoever could meet the requirements of the principle! As Joergensen says, “. . .not even protocol sentences, by means of which the truth of all other reality sentences was to be tested, were capable of being verified in the strict sense of the word so that they become absolutely certain.”[28]


                  Since the protocol statements themselves were shown not to be certain or unique, the formerly held view of a reduction of all empirical statements to them also mandated an alteration. Carnap stated that “There are no absolutely basic sentences in the construction of science.”[29] Instead, he claimed that any reduction to a so-called “terminal point” was a matter of decision or convention--wherever one wished to make a halt. This change in the status of protocol statements had a great significance for epistemology. Kraft says that this incisive turn “meant again the elimination of a last remnant of absolutism from epistemology.”[30]


Other Difficulties with Complete Verification


                  The demonstration that even singular statements could not be verified according to the verification principle was perhaps the crowning blow to its continued use in that form. But the principle was attacked for other major reasons as well. Several of these major difficulties will be discussed briefly.


                  Remembering that the positivists include only two classes of cognitively meaningful statements that could be true --the analytic and synthetic--a first major difficulty with the principle was that it could be classed as neither. The statements: “The meaning of a proposition is the method of its verification” was certainly not analytic, and it could not be empirically verified, meaning that it could not be treated as synthetic. This meant that the criterion itself was meaningless. This seemed to be quite a logical embarrassment.


                  Several of the Logical Positivists (eg. Carnap and Ayer) responded by saying that the principle was not really a proposition (or statement), but rather a “proposal” or a “recommendation.” Hence, the principle would be of a “different logical status” than analytic or synthetic statements and would not be subject to its own criterion.[31] But this response greatly weakened the original formulation of the principle.[32] Passmore points out that “now it appeared that the metaphysician could escape their criticisms simply by refusing to accept their recommendation.”[33] The original purpose of the positivists to universally eliminate metaphysics from meaningful discussion is no longer possible to fulfil. Charlesworth is correct when he says that “the verification principle has no real prescriptive force, nor any power to ‘eliminate’ metaphysical or any other kind of propositions.”[34]


                  A second major criticism of the verifiability criterion--one that was stressed by Popper--was that the criterion not only excluded metaphysics, it also excluded any universal statements of science. The laws of science do not, by their very nature, admit of complete verification. Passmore relates that “there is no set of experiences such that having these experiences is equivalent to the truth of a scientific law.”[35]


                  A third major difficulty of the meaning criterion was that it could not escape the problem of solipsism. Complete verification was taken to be possible only for an individual through the so-called protocol statements. The problem involved attempting to make the transition from one’s private experiences to the experiences of others and to the public world. But there was, in principle, “no way of determining whether a proposition is verifiable--or unverifiable--for anyone except myself.”[36] This is one of the major problems that caused Carnap and Neurath to view the protocol statements as being “intersubjective,” describing physical events rather than private experiences.


                  The end result of these and other problems was that it was recognized by the positivists that the criterion of meaning as it was initially formulate and intended was far too strict. No statement could, in fact, meet its requirements satisfactorily. Therefore, several attempts have been made which present the principle in a “weaker” form. Ayer admits, however, that although attempts have been made to give some weaker version of the verification principle a thoroughly precise expression, “the results have not been altogether satisfactory.”[37] (37)




                  Clearly, the “strong” version of the meaning criterion had its difficulties. It could not allow, when consistently applied, statements about the remote past or distant future as being meaningful. Neither would it allow certain generally accepted statements as being meaningful. Under the originally formulated criterion, even the statement “There is molten metal at the earth’s core” could not be meaningful because it could not actually be verified. It should be pointed out, however, that although such criticisms were leveled against the positivists on these points, it seems doubtful that they actually intended for the criterion to be interpreted in such a way. Blanshard points out that these criticisms were probably directed to a “straw man.” He says, “There seems to be no good evidence that any responsible positivist, the original solipsists apart, ever limited the meaningful to what was technically and practically verifiable by me.”[38]


                  The positivists attempted then to “clarify” the criterion by saying that statements could be meaningful if they were verifiable in principle. This view was adopted by Schlick in 1932 and was termed by him “consistent empiricism.” Blanshard states the view in the following way: “ If a statement is to be meaningful, I must be able to verify it, not necessarily in practice, nor even in my lifetime, nor even in the race’s lifetime, but a least conceivably in theory.”[39]


                  Ayer uses as an example the statement: “There are mountains on the farther side of the moon.” This statement, at the time of his writing, was unverifiable from a practical standpoint. But he says, “I do know what observations would decide it for me, if, as is theoretically conceivable, I were in a position to make them.”[40] In making this adjustment, the positivists felt as though they had escaped a criticism, but that they were still successful at eliminating metaphysics. For, as they said, a metaphysical statement in the traditional sense is incapable of verification even in principle.


                  However, this clarified criterion of verification in principle was soon shown to be problematic itself. First of all, it encountered problems with such simple and common sensical words such as “all,” “some,” and “exists.” One of the major criticisms Popper issued against the positivists was that only an infinite amount of observations--a “practical” impossibility--could logically justify any universal assertion. This being the case, no scientific law or universal assertion could meet the requirements of the “clarified” criterion. For such universals could not even be verified in principle. Diamond states,


This was especially embarrassing for the positivists. The application of their most distinctive principle resulted in the denial of meaning to well-established scientific laws. This meant that, if valid, their work would undercut the prestige of science. This was a case of the positivists sawing off the limb they were sitting on, because the appeal of their program was heavily dependent on this prestige.[41]


                  This problem of verifying universals was foreseen by such men as Wittgenstein and Schlick. As a background to help understand how they treated this issue, it should be made clear that they viewed universals in two different senses: (1) in a restricted sense in which one might speak of “all the people in Vienna,” and (2) in an unrestricted sense such as the laws of nature. Wittgenstein and Schlick recognized only the restricted category as containing genuine cognitive propositions. For they were aware that only in a restricted sense could such statements be verified (and therefore meaningful). How did they handle the second category? They viewed the laws of nature as not really asserting anything about the empirical world. They were viewed as nothing but syntactic or methodological rules.[42] Thus, they attempted to avoid the problem of verifying universals in a traditional sense. Their attempt was not adequate, however, so the problem and Popper’s criticism remain.


                  Two other major problems with “verification in principle” are mentioned by Hempel. First, it cannot adequately deal with a conjunction of two sentences, one meeting the criterion and the other not, with the word “or.” Such a construction would allow meaningfulness to the entire conjunction when one of the constituent statements would clearly not meet the criterion.[43]


                  Another problem is that the criterion would lead to a major logical inconsistency. Logically, the denial of a true statement must be false. But if the criterion were applied to an existential but true empirical statement such as : “There exists at least one thing with the property P,” its denial would be meaningless. For the only way to make the denial meaningful would involve a universal sentence such as: “Nothing has the property P.” Thus, if the criterion were valid one would have to deny the fundamental logical principle that if a statement is true or false, its denial is false or true respectively. Or one would have to deny that “(x) P(x)” is logically equivalent to the negation of “(Ex) P (x).” Hempel concludes and says that “the criterion, which has disqualified itself on several other counts, does not warrant such drastic measures for its preservation; hence, it has to be abandoned.”[44]





                  Since the concept of verifiability was shown to have insurmountable difficulties serving as a criterion of meaning, some felt that a criterion of falsifiability would be successful. Recognition for suggesting this criterion is often attributed to Sir Karl Popper. Hempel says,


The idea of using theoretical falsifiability by observational evidence as the ‘criterion of demarcation’ separating empirical science from mathematics and logic on the one hand and from metaphysics on the other is due to K. Popper.[45]


Thus, falsifiability was taken by many to be a criterion of meaning with the same general intent as that of verifiability. Many have presented Popper as setting forth falsification as a criterion of meaning.[46] As late as 1950, and being a little less assuming, Hempel says, “Whether Popper would subscribe to the proposed restatements of the falsifiability criterion, I do not know.”[47]


                  It is clear, however, that Popper has been wrongly credited or unjustly blamed (depending upon one’s position) for setting forth falsifiability as some kind of sequel to the verification theory of meaning. Magee states,,,, “Not only was Popper not putting forward a criterion of meaning; he has always held that to do so is a major philosophical error.”[48] Popper himself explicitly attacks this misinterpretation of his position in a footnote in The Logic of Scientific Discovery (1959). After saying that “not the verifiability but the falsifiability of a system is to be taken as a criterion of demarcation,” Popper notes:


Note that I suggest falsifiability as a criterion of demarcation, but not of meaning. Note, moreover, that I have already (in section 4) sharply criticized the use of the idea of meaning as a criterion of demarcation, and that I attack the dogma of meaning again, even more sharply, in section 9. It is therefore a sheer myth (though any number of refutations of my theory have been based on this myth) that I ever proposed falsifiability as a criterion of meaning.[49]


                  It is true, however, that if falsification is taken as a criterion of meaning that it fails on analogous grounds as did the verification criterion.[50] With the necessary conclusion that both verification and falsification were too strict as a criterion of meaning, a “weaker criterion” had to be chosen.[51]         




                  Since it was shown to be too problematic for conclusive verification or falsification to occur, particularly with regard to universal statements, a more liberal criterion of meaning was sought. In the first edition of Language, Truth, and Logic (1936), A.J. Ayer presented such a criterion. In this work, he made a distinction between “strong” and “weak” verifiability. In the strong sense a proposition could be considered meaningful “if and only if its truth could be conclusively established in experience.” A proposition was verifiable in the weak sense “if it was possible for experience to render it probable,”[52] or “if some possible sense-experience would be relevant to the determination of its truth or falsehood.”[53] But since this statement was itself ambiguous, Ayer issued a second version of it. It might be stated: “…a statement is verifiable, and consequently meaningful, if some observation-statement can be deduced from it in conjunction with certain other premises, without being deducible from these other premises alone.[54]


                  But as Hempel alleges, this version of Ayer’s criterion is highly inadequate. For instance, the following hypothetical syllogism would meet its requirements:


                  Statement: “The Absolute is Perfect”

                  A conjoined premise: “If the Absolute is perfect, then this apple is red”

                  Observation statement: This apple is red”


Note that the observation statement is not deducible from the conjoined premise alone, but that it can be deduced from the conjunction of both the initial statement and the conjoined premise. This would, according to Ayer’s criterion, make the statement “The Absolute is Perfect” a meaningful statement--a highly undesired conclusion for Ayer![55] Ayer admits in the second edition of Language, Truth, and Logic (1946) that this criterion is “far too liberal, since it allows meaning to any statement whatsoever.”[56]


                  Therefore, Ayer proposed an emendation of his criterion in that second edition of Language, Truth, and Logic. This version of the criterion attempted to stipulate more precisely what kind of statements were allowed (or rather disallowed) to be included as conjoined premises. He declares:

. . . a statement is indirectly verifiable if . . . first, that in conjunction with certain other premises it entails one or more directly verifiable statements[57] which are not deducible from these other premises alone; and secondly, that these other premises do not include any statement that is not either analytic, or directly verifiable, or capable of being independently established as indirectly verifiable.[58]


Thus, Ayer was attempting to preclude the use of such “unverifiable” conditionals such as: “If the Absolute is Perfect, then . . .” because such a conditional would supposedly not meet this new criterion of being capable of being “independently established as [even] indirectly verifiable.”

                  But Hempel argues that the conjoined statement: “If the Absolute is Perfect, then . . . [this apple is red]” cannot be torn apart in such a way. For the conjunction S.N.--where S satisfies Ayer’s criterion and N is a statement like “The Absolute is Perfect”--must still be viewed together as a meaningful statement. In relaying Hempel’s argument, Ashby says, “. . .if S is meaningful, the S and N will be meaningful, whatever statement N may be.”[59] If this be the case, then Ayer’s new emendation is not successful in precluding metaphysical (and theological) statements from meaningfulness. The criterion, from the positivists viewpoint, is still much too permissive.


                  The “permissiveness” of Ayer’s partial or indirect confirmation criterion can also be attacked (or rather used) even by theologians. For even though talk about--which would itself be meaningless according to Ayer’s intentions--could not be “verified,” statements about Him could be “confirmed.” As Diamond expresses it:

. . .statements. . .like ‘God divided the waters of the Red Sea,’ cannot be conclusively verified by observations, but very few theologians would concede the point that there are no observations that could confirm the truth of their statements. If the demand for conclusive verification and falsification is             abandoned, many theologians could specify observations that count for the truth of their statements.[60]


Therefore, according to Ayer’s own criterion, evidence can be presented that confirms even the existence of technically “non observable reality.” But God talk is not totally unique on this point. Physics itself is forced to rely upon it. For instance, any statement about subatomic particles is a statement which “cannot be observed directly, but we regard certain types of behavior of drops of oil between charged plates as evidence that confirms the existence of electrons.”[61]




                  The search for a general criterion of meaning has been shown to occur on unfertile soil. In each case, it has proved to either too restrictive (i.e., not allowing even accepted scientific statements to be meaningful), or too permissive (i.e., allowing almost any statement to be meaningful). Many philosophers of the positivistic tradition saw the futility in their search for a general criterion of meaning. The question expressed by Blanshard seemed penetrating especially to those of the British schools: “What is the point of a test if that test itself must be brought to the test of an inspection of actual meaning, which is thereby conceded to be the court of last appeal?”[62] For this reason, and also greatly because of Wittgenstein’s turn in his Philosophical Investigations (1953), the British analytic movement has largely been away from the search for a general criterion of meaning. Instead, they have been more interested in how ordinary language is actually used.

                  However, in America, those of the positivistic tradition have not relinquished their search for a general criterion of meaning so easily.[63] The roots for this new approach to a meaning criterion began quite early. It was Carnap who, in his two articles entitles “Testability and Meaning” (1936-37), set forth a major critique and revision of the earlier criterion of meaning. He, as well as Ayer,[64] admits that the definition of meaningfulness in terms of verification is too narrow. In this epochal work Carnap began speaking of “confirmation” rather than “verification.” He says:

If by verification is meant a definitive and final established truth, then no (synthetic) sentence is ever verifiable, as we shall see. We can only confirm a sentence more and more. Therefore we shall speak of the problem of confirmation rather than the problem of verification.[65]


                  Although both the published works of Carnap and Ayer began speaking of “conformability” rather than verifiability, the two works were founded on divergent methodologies. The work of Ayer basically attempted to provide a criterion of meaning through deductive relationships to observation sentences within a natural language; whereas Carnap sought “to regard as cognitively meaningful all and only those statements that can be expressed in a formalized empiricist language.”[66]


                  Carl Hempel, in his article “The Empiricist Criterion of Meaning” (1950), adopts a position closely associated with that of Carnap. In referring to the former attempts at prescribing a criterion of meaning, he says:

. . . it is useless to continue the search for an adequate criterion of testability in terms of deductive relationships to observation sentences. . . . as long as we try to set up a criterion of testability for individual sentences in a natural language, in terms of logical relationship to observation sentences, the result will be either too restrictive or too inclusive, or both.[67]


Hempel then continues and states,


The predicament would not arise, of course, in an artificial language whose vocabulary and grammar were chosen as to preclude altogether the possibility of forming sentences of any kind which the empiricist meaning criterion is intended to rule out.[68]


This translatability criterion of meaning can be expressed: “A sentence has cognitive meaning if and only if it is translatable into an empiricist language.”[69]


                  This empiricist language was to be an “ideal” language which contained two kinds of components. First, it included the customary signs of symbolic logic; and secondly, its vocabulary (i.e., its subjects and predicates) was to be one that referred to observable things or observable characteristics of things.[70] It was a “physical” or “physicalistic” language that is also referred to as a “thing-language.”[71] According to Hempel, this criterion “allows cognitive import to a sentence only if its constitutive empirical terms are explicitly definable by means of observation predicates.[72]


                  Ashby affirms, however, that these “observation predicates” could well be problematic.[73] For if such “observation predicates” are used to designate properties that are in principle observable, but ones that have never actually been observed (eg., “magnetic,” “electrically charged”), how can one know that such an unobserved property is really observable?


                  Although Hempel says that “a language. . .is empiricist if all its sentences are expressible, with the help of the usual logical locutions, in terms of observable characteristics of physical objects,”[74] he admits that there are certain problems with constructing such a language.[75] For many terms, even in the physical sciences, are not strictly definable in terms of observation predicates. Terms such as “magnetic” or “electrically charged” are not characteristics that are actually observed; hence, a thing-language in the strict sense is not possible for these cases.


                  But Carnap, in his “Testability and Meaning” articles, suggested a system of “reduction” by which such characteristics (i.e., “empirical constructs”) could be expressed in an empiricist language. The task of setting forth a system of reduction had been central in his earlier theory of constitution.[76] But this subsequent thesis of reduction, though similar to the earlier, had several fundamental modifications. One of these was that “contrary to constitution, reduction is defined in the formal mode of speech, and the definition is expressed exactly by means of logistic symbols.”[77] A second alteration was that it changed the function of definability. Kraft explains this when he says the following:


Carnap’s constitution system was dominated by the positivist-empiricist principle that every empirical concept of science is reducible to, and accordingly definable by, concepts expressing relations between experiences. . . . This thesis is not subjected to a basis restriction. Reducibility is still maintained, but unqualified definability and thereby replaceability by relations of experiences cannot be claimed any more.[78]


Rather than “unqualified definability,” Carnap’s reduction sentences now take the form of “partial or conditional definitions.” These reduction sentences “provide a means for the precise formulation of what is commonly referred to as operational definitions.”[79] Thus empirical constructs (e.g. “magnetic,” “electrically charged”) are made translatable into an empiricist language by means of reduction sentences--sentences that can be used to specify the rules under which a particular definition might be applied. The effect of this translatability criterion is that it “qualifies a sentence as cognitively meaningful if its non-logical constituents refer, directly or in certain specified indirect ways, to observables.”[80]


                  Such a physicalistic language provides the basis for a unity of science. The language is chosen because it is intercensal, intersubjective and universal.[81] But even with Carnap’s reduction sentences, the construction of such a language is not without major problems. For instance, translatability of psychological sentences into a thing-language must be premised on the truth of logical behaviorism,[82] which is a highly uncertain thesis. Also, it must be admitted that the thesis of physicalism is one that can never be proved until the reduction of all concepts in the social and natural sciences is made, “which means, of course, never.”[83]




                  It has been the purpose of this paper to trace some of the major forms in which the empiricist’s criterion of meaning has been presented. This paper has shown, as also Brand Blanshard observes, that “The formulations have been so various that one cannot talk responsibly of the verifiability theory; one must specify one or other of its forms.”[84] It will not suffice then to criticize the verification principle and its subsequent forms in some monolithic fashion. Even more would it be irresponsible to excoriate some earlier version of the principle alone--a version long since abandoned by the positivists themselves--without apparent cognizance of its subsequent reformulations.


                  Yet in all the presented versions of a meaning criterion, at least two major characteristics seem to be common to all. The first is an avid attempt at eliminating metaphysics from meaningful discourse; and the second, a correlative to the first, is that the only experience that can assure us of fact is sense experience. The former seems to be the major intention, while the latter is the major premise.


                  Such a premise and intention need to be potently challenged with an adequate alternative. But reality itself has proved historically to be unkind to both.




Ashby, R.W. “Verfiability Principle” S.v. in Encyclopedia of Philosophy, Vol. 8, pp. 240-247.

Ayer, A.J. Language, Truth and Logic. 2nd edition, NY: Dover Pub., 1946.

Ayer, A.J. “The Vienna Circle” in The Revolution in Philosophy. London: Macmillan & Co., 1963.  Reprinted from 1956.

Ayer, A.J. “Editor’s Introduction” in A.J. Ayer, editor. Logical Positivism. NY: The Free Press,1959. (Abbreviated “Introduction”)  

Bergman, Gustav. The Metaphysics of Logical Positivism. Madison, WI: University of Wisconsin Press, 1967.

Blanshard, Brand. Reason and Analysis. LaSalle, IL: Open Court Pub.Co., 1962.

Charlesworth, Maxwell J. Philosophy and Linguistic Analysis. Pittsburgh, PA: Duquesne University Press, 1959.

Diamond, Malcolm L. Contemporary Philosophy and Religious Thought. NY: McGraw-Hill Co., 1974.

Feigl, Herbert. “Logical Positivism After Thirty-Five Years.” Philosophy Today 8 (Winter 1964):228-45.

Hempel, Carl G. “The Empiricist Criterion of Meaning.” in A.J. Ayer, editor, Logical Positivism. 108-129. Reprint Vol. 4 of Revue Internationale de Philosphie (1950).

Hunnex, Milton D. Philosphies and Philosophers, revised edition. San Francisco: Chandler Pub. Co., 1961.

Joergensen, Joergen. The Development of Logical Empiricism. Chicago, IL: The University of Chicago Press, 1951.

Kant, Immanuel. Prolegomena to Any Future Metaphysics. Indianapolis, IN: The Liberal Arts Press, 1950. Originally published in 1783.

Kockelmans, Joseph J., editor. Philosophy of Science: The Historical Background. NY: The Free Press, 1968.

Kraft, Victor. The Vienna Circle. NY: Philosophical Library, 1953.

Magee, Bryan. Modern British Philosophy. London: Secker and Warburg, 1971.

Magee. Karl Popper. NY: The Viking Press, 1973.

Martin, Norman. “Rudolf Carnap.” S.v. in The Encyclopedia of Philosophy. Vol. 2, pp. 25-33.

Passmore, John. A Hundred Years of Philosophy. NY: Basic Books, Inc., 1957.

Passmore. “Logical Positivism.” S.v. in The Encyclopedia of Philosophy. Vol. 5, pp. 52-57.              

Popper, Karl. The Logic of Scientific Discovery. 2nd English edition. NY: Harper and Row, 1968. E.T. 1959.

Popper. Conjectures and Refutations. NY: Basic Books, 1963.



                                                                                                                                                Dr. James D. Strauss

                                                                                                                                                Professor Emeritus

                                                                                                                                                Lincoln Christian Seminary

                                                                                                                                                Lincoln, IL 62656

[1] Magee, Karl Popper, p. 2.

[2] Ibid., p. 3.

[3] Ibid.

[4] Passmore, “Logical Positivism, “ p. 56.

[5] This term is presently rejected by many who were once closely associated with the Vienna Circle. Feigl states that he abandoned the label in 1935 because of the implied meaning often attached to “positivism.” He prefers the term “logical empiricism” which is a term suggested by Ernest Nagel. C. Feigl, p. 231-32.

[6] Kraft, p. 4.

[7] See particularly The Logic of Scientific Discovery, pp. 21-22.

[8] Bergman, p. 1.

[9] Hunnex, part II.

[10] Hempel, p. 109.

[11] Ayer, “Introduction,” p. 11.

[12] Kant, p. 3.

[13] Ibid., p. 4.

[14] Diamond, pp. 11-12.

[15] Passmore, A Hundred Years of Philosophy, p. 360.

[16] Note Popper’s statement that “Carnap and the Vienna Circle attributed it [the concept of a ‘meaningless’ category] to Wittgenstein, but it is much older. The theory goes back to Hobbes at least; “. . .asserting that words purporting to denote unobservable entities cannot have any meaning--it was clearly and forcefully used by Berkeley (and other nominalist).” Conjecture and Refutations, p. 258 n. 12.

[17] Kraft, p. 23.

[18] Note Hempel’s remark: “... the term ‘verifiability’ is to indicate...the conceivability, or better, the logical possibility of evidence of an observational kind which, if actually encountered, would constitute conclusive evidence for the given sentence; it is not intended to mean the technical possibility of performing the tests needed to obtain such evidence, and even less does it mean the possibility of actually finding directly observable phenomena which constitute conclusive evidence for that sentence--which would be tantamount to the actual existence of such evidence and would thus imply the truth of the given sentence.” See Hempel, p. 111 n. 6.

[19] Kraft, pp. 118-19.

[20] Ayer, “Introduction” p. 13.

[21] Ayer, “The Vienna Circle,” p. 82.

[22] Kraft, p. 112. It should be noted that the postulate that all concepts should be reducible to inner and external perception--to inner experience--was at the heart of Locke’s and Hume’s epistemology. Such a reduction was not actually attempted until Carnap. See Kraft, p. 102. This constitution theory was set forth in Carnap’s The Logical Construction of the World (1928).

[23] Kraft, p. 114-15.

[24] Ibid., p. 115.

[25] Ibid., p. 118.

[26] Erkenntnis, Vol. 3 pp. 209ff.

[27] Ayer, “Introduction,” p. 20. This is the general thesis of “physicalism.” This thesis was that the language of physics, i.e., statements which are used to refer to physical events, is a universal language-universal in the sense that every empirical statement can be expressed in it. Ayer says, “It was his acceptance of this doctrine of physicalism that led Neurath to insist so much upon the unity of science.” Ayer, “The Vienna Circle,” p. 82.

[28] Joergensen, Joergern. The Development of Logical Empiricism (Chicago, IL: University of Chicago Press, 1951), pp. 71-72.

[29] C. Kraft, p. 119. This is quoted from Carnap’s “ueber Protokollsaetze,” Erkenntnis Vol. 3, p. 224.

[30] Ibid., p. 119.

[31] Ashby, p. 244. Note also other responses to this problem discussed by Ashby.

[32] Hempel admits, “...this conception would attribute to the criterion a measure of arbitrariness which cannot be reconciled with the heated controversies it has engendered...” Hempel, p. 124.

[33] Passmore, “Logical Positivism,” p. 54.

[34] Charlesworth, p. 147. He continues and says, “All one, [eg. Ayer] has a right to say is that he has defined ‘meaning’ in such a way that it does not apply to metaphysical propositions.”

[35] Passmore, “Logical Positivism,” p. 55.

[36] Passmore, A Hundred Years of Philosophy, pp. 375-76.

[37] Ayer, “Introduction,” p. 14.

[38] Blanshard, p. 208.

[39] Ibid., p. 209.

[40] Ayer, Language, Truth, and Logic, p. 36. Note also Hempel’s remark, supra, pg. 6, note 1.

[41] Diamond, p. 31.

[42] Kraft, pp. 132-33.

[43] Hempel, p. 112.

[44] Ibid., p. 113.

[45] Ibid., p. 113, no. 8.

[46] Blanshard, pp. 228-29; Diamond, pp. 31-34.

[47] Hempel, p. 114, n. 8.

[48] Magee, Karl Popper,p. 43.

[49] Popper, The Logic of Scientific Discovery, p. 40 n. 3.

[50] For a brief listing of these complications see Ashby, p. 242. These problems themselves, however, should not be allowed to be a criticism of Popper’s intended thesis.

[51] It should be noted that a basic sequence was involved. When verifiability was abandoned as inadequate in its stated form, falsifiability was tried. When falsifiability proved inadequate as a criterion of meaning, the two were combined together as a “criteria of meaning.” But his combination was also inadequate. Diamond, pp. 33-34.

[52] See Ayer’s 2nd edition (1946), pp. 9, 36.

[53] Ibid., p. 11.

[54] Ibid.

[55] Refer to Hempel’s discussion on this point, pp. 114-15.

[56] Ayer, Language, Truth, and Logic, p. 11.

[57] “...a statement is directly verifiable if it is either itself an observation statement, or is such that in conjunction with one or more observation statements it entails at least one observation statement which is not deducible from these other premises alone;...” Ibid., p. 13.

[58] Ibid., p. 13.

[59] Ashby, p. 242.

[60] Diamond, p. 35.

[61] Ibid. See Blanshard’s discussion on this same issue as well, pp. 238-39.

[62] Blanshard, p. 232.

[63] Blanshard remarks that “Perhaps the Teutonic professional genius is less emendable to this kind of correction.” P. 233.

[64] Note that these articles were almost simultaneous with Ayer’s first edition of Language, Truth, and Logic (1936).

[65] Quoted from Joergensen, p. 74.

[66] Ashby, p. 243.

[67] Hempel, p. 116.

[68] Ibid.

[69] Ibid., pp. 116-17.

[70] Blanshard, p. 233.

[71] Joergensen, p. 77.

[72] Hempel, p. 118.

[73] Ashby, p. 243.

[74] Hempel, p. 117.

[75] Ibid., pp. 118-22.

[76] Supra, p. 7.

[77] Joergensen, p. 80.

[78] Kraft, pp. 112-13. Norman Martin says, “We must therefore distinguish two forms of positivism and physicalism, one that claims that all scientific concepts are definable in terms of protocols in the phenomenalistic or physicalistic language and one that makes the weaker claim of reducibility to such protocols.” See Martin, p. 30.

[79] Ayer, p. 120.

[80] Ibid., p. 122.

[81] See Joergensen’s discussion, pp. 78-79.

[82] Ibid., p. 81.

[83] Ibid., p. 82.

[84] Blanshard, p. 235.