Any statement of fact is true or false in virtue of some existing state of affairs in the world. In many cases the truth-value of a statement is determined by appealing to the truth-values of certain other statements, but this process must terminate somewhere if the truth-value of any statement of fact is to be assessed at all. An epistemological view according to which the process of verification or falsification terminates with statements of a logically distinct kind is a view to the effect that there is a distinct class of basic statements. The principal questions that have been considered are (1) Is there such a class of statements? (2) If there is, what is the relation between these statements and certain nonverbal occurrences called experiences? (3) Are basic statements descriptions of the private experiences of the speaker or of publicly observable events? (4) Are these statements either incorrigible (that is, of such a character that they cannot be false, or cannot be shown to be false) or indubitable (that is, such that they cannot rationally be doubted)? These questions have been much discussed by modern empiricists, especially in connection with the verifiability criterion of meaning. The problems concerning basic statements are not, however, essentially confined to empiricist theories of meaning and truth; they are fundamental in any theory of knowledge.
The thesis that there is a class of basic or elementary propositions is powerfully presented in Ludwig Wittgenstein's Tractatus Logico-Philosophicus (1921; first English translation, 1922). Wittgenstein argues that if a proposition contains expressions standing for complexes, the sense of the proposition will depend upon the truth of other propositions describing those complexes. This will again be the case if any one of those other propositions contains expressions standing for complexes. Thus, the determinateness of the sense of the original proposition requires that its analysis should terminate in elementary propositions consisting only of names of simple things (see 2.0211–2.0212, 3.23). An elementary proposition is an arrangement of names that represents a possible arrangement of simple things; it is a logical picture of an elementary state of affairs. Wittgenstein gave no explicit interpretation of "simple things," "names," or "elementary propositions." He is reported as saying that at the time he wrote the Tractatus he thought it was not his business, as a logician, to give examples of simple things, this being a purely empirical matter; the Tractatus view is that the application of logic decides what elementary propositions there are (5.557).
Moritz Schlick and some other members of the Vienna circle gave an empiricist interpretation to Wittgenstein's theory. In "Über das Fundament der Erkenntnis" (1934) and other articles, Schlick inquired whether there is a class of statements which provide an "unshakeable, indubitable foundation" of all knowledge. This kind of incorrigibility, he argued (against Otto Neurath and Rudolf Carnap), cannot depend simply upon the coherence of a statement with the existing system of science, nor simply upon someone's decision to accept a statement as true. It is possessed only by the statements a person makes about his own experiences. Schlick called such statements Konstatierungen "confirmations" and contrasted these with the "protocol sentences" described by Neurath and Carnap.
Konstatierungen have the following characteristics: (1) They have the form "here, now, so and so"; examples are "here two black points coincide," "Here yellow borders on blue," "Here now pain." (2) In the case of other synthetic statements, understanding their meaning is quite distinct from the actual process of verifying them, and their meaning does not determine their truth-value; but in the case of a Konstatierung (since "'this here' has meaning only in connection with a gesture … one must somehow point to reality"), the occasion of understanding it is the same as that of verifying it. Therefore a (significant) Konstatierung cannot be false. (3) Unlike "protocol sentences," these statements cannot be written down or recorded at all because of the fleeting reference of the demonstratives that occur in them; but they provide the occasions for the formation of protocol sentences. (4) They are the only empirical statements that are not hypotheses. (5) They are not the starting points of science in either a temporal or a logical sense, but simply the momentary consummations of the scientific process; they are the means by which all scientific hypotheses are confirmed.
The first and most obvious objection to the view that there are Konstatierungen (in Schlick's sense) is that it results immediately in a radical form of solipsism. It may also be objected that Konstatierungen are either genuine contingent statements, in which case they cannot be of such a nature that they cannot be false, or they are purely demonstrative, in which case they are not statements. Following Wittgenstein's later work, many philosophers would deny the possibility of the essentially private use of demonstratives and descriptions that are supposed to occur in Konstatierungen. Further, no adequate account is given of the relation between these private statements and the public protocol sentences to which they give rise. Moreover, if the Konstatierungen are meaningful only at the moment at which they are verified, they cannot occur in predictions, and hence it cannot be through them that scientific hypotheses are confirmed.
Rudolf Carnap, in "Die physikalische Sprache als Universalsprache der Wissenschaft" (1931; translated as The Unity of Science, 1934) and elsewhere, had at first held that science is a system of statements based upon sentences describing the experiences of scientific observers. These "primitive protocol sentences," Carnap supposed, contain no inferential or theoretical additions; they describe only what is directly given, and hence they stand in no need of any further justification. At this time Carnap left it an open question whether protocol sentences describe the simplest sensations and feelings of the observer (for example, "here now red," "joy now"), or partial or complete gestalts of single sensory fields (for example, "red circle now"), or the total experience of the observer during an instant, or macroscopic material things (for example, "A red cube is on the table"). Later, however, in Logische Syntax der Sprache (1934) and other publications, due mainly to the criticisms of Neurath, Carnap held that the question of what protocol sentences describe is not a factual but a linguistic question and that we are free to choose whatever form of language is most convenient for reporting observations in science.
Otto Neurath, in "Soziologie im Physikalismus" (1931/1932; English translation, 1959) and other articles, had argued that sentences cannot be compared with the private experiences of the observer, nor with public material things, but only with other sentences. Some sentences are reports of acts of observation, in the sense of being behavioral responses to those acts, and such protocol sentences may have whatever form we find most convenient. In "Protokollsätze" (1932/1933), Neurath maintained that for the purposes of science it must be possible to incorporate the protocol sentences expressed at one time in those expressed at another time, and that comparison of protocols, even with one's own past protocols, requires an intersubjective language. Neurath remarks, "every language as such is inter-subjective." Carnap later agreed that if protocol sentences were regarded as describing the observer's private experiences, they could be understood, if at all, only solipsistically. Neurath suggested that a convenient form for protocol sentences would be one which contained a name or description of an observer and some words recording an act of observation; he gives as an example "Otto's protocol at 3:17 o'clock [Otto's word-thought at 3:16: (In the room at 3:15 was a table perceived by Otto)]." In this example, it is supposed that the entire sentence is written down by Otto at 3:17, simply as an overt verbal response; the sentence in brackets is Otto's response at 3:16, and the sentence in parentheses is his response at 3:15. The word "Otto" is repeated, instead of using "my" and "me," in order that the components of the protocol may be independently tested, for example, by being found in the protocols of other observers. The protocols of different observers or of the same observer may conflict, and when this happens, one or more of them is to be rejected.
According to Neurath, Carnap, and also Carl Gustav Hempel in "On The Logical Positivists' Theory of Truth" (1934/1935) it is a matter of convenience and decision which of the conflicting protocols should be rejected; hence, no protocol is incorrigible. The aim of science is to build up a coherent system of sentences, but no sentence at any level is sacrosanct; every sentence in science is in the end accepted or rejected by a decision made in the interests of coherence and utility. This view was strenuously opposed—by Schlick, Bertrand Russell, and A. J. Ayer, among others, who argued that (1) on this account protocol sentences are distinguished from others only in respect of their syntactical form; (2) a purely syntactical criterion of truth cannot do the work required of it; and (3) the Neurath–Carnap doctrine is a complete abandonment of empiricism.
According to Bertrand Russell's early doctrine of knowledge by acquaintance and knowledge by description, "every proposition we can understand must be composed wholly of constituents with which we are acquainted." A person is acquainted with those objects that are directly presented to his mind, and Russell held that sense data and universals are so presented. Later, in The Analysis of Mind (1921), Russell maintained that it is not possible to make a distinction between sensation and sense datum and that a sensation is not itself a cognition, although it is a cause of cognitions. This view led to the account of basic propositions that Russell gives in An Inquiry into Meaning and Truth (1940). In epistemology, he says, we can arrange our propositions about matters of fact in a certain order such that those that come later are known, if they are known, because of those that come earlier. At the beginning of such an ordering there will be "basic propositions"—those which "on reflection appear credible independently of any argument in their favour."
A basic proposition is one whose utterance is caused as immediately as possible by a perceptual experience. It is known independently of inference but not independently of evidence, since the perceptual experience that causes it to be expressed also gives the reason for believing it. The perceptual experience in question provides the strongest possible evidence for the basic proposition; no previous or subsequent occurrence and no experiences of others can prove that the proposition is false. Nevertheless, according to Russell, a basic proposition is not incorrigible; it cannot be disproved, but it may be false. Since one of the aims of epistemology is to show that all empirical knowledge is based upon these propositions, it is desirable that they should be given a logical form which makes contradiction between them impossible. Russell therefore defines a basic proposition as one "which arises on the occasion of a perception, which is the evidence for its truth, and … has a form such that no two propositions having this form can be mutually inconsistent if derived from different percepts" (Inquiry into Meaning and Truth, p. 139). Examples are "there is a canoid (shaped) patch of color," "I am hot," "that is red." Alternatively, "we can consider the whole body of empirical knowledge and define 'basic propositions' as those of its logically indemonstrable propositions which are themselves empirical" (ibid.). Russell believes that this logical definition is extensionally equivalent to his epistemological definition.
Whether basic propositions are incorrigible or indubitable, and if so in what sense, has been considered at length by A. J. Ayer. In "Basic Propositions" (1950) he defends the view that if a sentence is a direct description of a private experience, it may be verbally incorrect, but it cannot express a proposition about which the speaker can be factually mistaken. He explains this in the following way. Many descriptive sentences, for example, "That is a table," may be used correctly (that is, in accordance with the rules of the language and on occasions generally agreed to be appropriate for their use), and yet the propositions they express may turn out to be false. But in the case of a sentence which directly describes a present experience, if the sentence is used correctly (that is, in accordance with the speaker's rules), the proposition it expresses cannot turn out to be false. Thus, "the sense in which statements like 'This is green,' 'I feel a headache,' 'I seem to remember——' can be said to be indubitable is that, when they are understood to refer only to some immediate experience, their truth or falsehood is conclusively determined by a meaning rule of the language in which they are expressed" ("Basic Propositions," p. 72).
Later, in The Problem of Knowledge (1956) and elsewhere, Ayer argues that language rules may be essentially private and that basic statements may be expressed in a sense-datum terminology, provided that this terminology is translatable into a terminology of seeming. Incorrigibility is not a property belonging to statements as such; "the sentences 'He has a headache,' when used by someone else to refer to me, 'I shall have a headache,' used by me in the past with reference to this moment, and 'I have a headache' all express the same statement; but the third of these sentences alone is used in such conditions as make it reasonable for me to claim that the statement is incorrigibly known" (The Problem of Knowledge, p. 58). But Ayer here allows that if he were asked, regarding two lines in his visual field, which looked to him to be the longer, he might very well be uncertain how to answer; and this uncertainty would not be about the meaning of the expression "looks longer than" but about a matter of fact. If anyone can have doubt about such matters of fact, he can presumably come to the wrong decision, that is, he can judge that one of the lines looks to him longer than the other when in fact it does not. No direct test of such a mistake is possible, but there may be various kinds of indirect evidence to show that it has occurred; hence, Ayer concludes, there is no class of descriptive statements which are incorrigible.
The requirements made upon basic statements are very often governed by the general nature of the theory of knowledge held by a philosopher. Thus, according to Karl Popper, our experiences cannot justify or establish the truth of any statement; the question for epistemology is not "on what does our knowledge rest? … or more exactly, how can I, having had the experience S, justify my description of it and defend it against doubt," but rather "how do we test scientific statements by their deductive consequences … what kind of consequences can we select for this purpose if they in their turn are to be intersubjectively testable?" (The Logic of Scientific Discovery, p. 98). Popper requires a class of basic statements by reference to which it can be decided whether a theory or hypothesis in science is falsifiable. Evidently a theory can be falsified by a basic statement only if the negation of the latter is derivable from the theory. Popper finds that his requirements are met by taking singular existential statements of the form "There is a so-and-so in space-time region k " as basic. It follows that the negation of a basic statement is not itself a basic statement (Popper allows some simple exceptions to this in Conjectures and Refutations, Addenda, p. 386); it also follows that any conjunction of basic statements which is not a logical contradiction is a basic statement and that the conjunction of a nonbasic and a basic statement may be a basic statement (for example, the conjunction of "There is no pointer in motion at k " with "There is a pointer at k," which is equivalent to "There is a pointer at rest at k ". Given a theory t conjoined with a statement of initial conditions r, from which a prediction p can be derived, it follows that r·∼p will be a falsifier of t and a basic statement—since if (t·r )→p, then t →(r →p ), that is, t →∼ (r·∼p ).
Popper also stipulates that the event referred to in a basic statement should be observable, that is, a basic statement must be intersubjectively testable by observation. He claims that the concept of an observable event can be elucidated either in terms of the experiences of an observer or in terms of macroscopic physical bodies, and hence that his account is neutral regarding the issue between psychologism and physicalism. In Popper's theory, the expression "observable event" is introduced "as an undefined term which becomes sufficiently precise in use: as a primitive concept whose use the epistemologist has to learn." According to Popper, "a science needs points of view and theoretical problems"; hence, in the practice of science we should not accept stray basic statements but only those which occur in the course of testing theories. Every test of a theory must terminate with some basic statement, but every basic statement can itself be subjected to further tests. There are no logical grounds for stopping at any particular basic statement. It is a matter for agreement and decision among those engaged in testing a theory; the process of corroboration or falsification terminates at the point at which they are satisfied for the time being.
From the preceding selection of views, held by recent and contemporary philosophers, it will be seen that there is no consensus concerning basic statements. The questions listed at the beginning of this article can be answered only in relation to a more general semantic and epistemological theory. Many such theories allow that there is a distinct class of basic statements. It seems that the relation between these statements and certain "experiences" of the speakers who express them must be partly semantic, and perhaps also partly causal, but the correct analysis of this relation is a matter of great difficulty. Many philosophers at the present time deny that there can be a class of statements that describe the private experiences of the speaker, on the grounds that there cannot be a language that is essentially private; but this latter view is also strongly contested. Finally, although on some views basic statements are indubitable, it seems that these statements cannot be incorrigible, at least in any sense that implies that they cannot be false. For if basic statements are to play the role assigned to them—namely, of being the terminating points of empirical verification—they must be genuine contingent statements; and a contingent statement is one whose negation is significant and could, as far as logic is concerned, be true.
See also Ayer, Alfred Jules; Carnap, Rudolf; Empiricism; Hempel, Carl Gustav; Knowledge and Belief; Neurath, Otto; Popper, Karl Raimund; Propositions; Russell, Bertrand Arthur William; Schlick, Moritz; Verifiability Principle; Wittgenstein, Ludwig Josef Johann.
Wittgenstein's view in the Tractatus Logico-Philosophicus (London: K. Paul, Trench, Trubner, 1922) is the object of his own criticism in Philosophical Investigations (Oxford: Blackwell, 1953), especially in Part I, Secs. 1–64.
Schlick's "Über das Fundament der Erkenntnis," in Erkenntnis 4 (1934), has been translated by David Rynin in Logical Positivism, edited by A. J. Ayer (Glencoe, IL: Free Press, 1959), pp. 209–227. The same volume also contains English translations by Morton Magnus and Ralph Raico of Otto Neurath's "Soziologie im Physikalismus" (which originally appeared in Erkenntnis 2 [1931/1932]) on pp. 282–317, and by Frederic Schlick of Neurath's "Protokollsätze" (which originally appeared in Erkenntnis 3 [1932/1933]) on pp. 199–208.
For Carnap's views, see "Die physikalische Sprache als Universalsprache der Wissenschaft," in Erkenntnis 2 [1931/1932]), translated by Max Black as The Unity of Science (London: K. Paul, Trench, Trubner, 1934), and Logische Syntax der Sprache (Vienna: Springer, 1934), translated by Amethe Smeaton as The Logical Syntax of Language (London: K. Paul, Trench, Trubner, 1937).
The views of Hempel may be found in "On the Logical Positivists' Theory of Truth," in Analysis 2 (4) (1934/1935).
Russell's views on basic statements can be found in The Analysis of Mind (London: Allen and Unwin, 1921) and An Inquiry into Meaning and Truth (London, 1940).
Ayer's contributions to this topic include the following: Language, Truth and Logic (London: Gollancz, 1936; 2nd ed., 1946), Ch. 5 and Sec. 1 of introduction to 2nd ed.; "Verification and Experience," in PAS 37; Foundations of Empirical Knowledge (London: Macmillan, 1940), Ch. 2; "Basic Propositions," in Philosophical Analysis, edited by Max Black (Ithaca, NY: Cornell University Press, 1950), reprinted in Philosophical Essays (London: Macmillan, 1954); and The Problem of Knowledge (London: Macmillan, 1956).
Further discussion of Quine's views may be found in "Two Dogmas of Empiricism," in Philosophical Review (1951), reprinted in From a Logical Point of View (Cambridge, MA: Harvard University Press, 1953); Methods of Logic (London: Routledge and Paul, 1952), introduction; and Word and Object (New York: Wiley, 1960), Secs. 8–10.
other recommended titles
Alston, William. The Reliability of Sense Perception. Ithaca, NY: Cornell University Press, 1993.
Audi, Robert. The Structure of Justification. Cambridge, U.K.: Cambridge University Press, 1993.
BonJour, Laurence. The Structure of Empirical Knowledge. Cambridge, MA: Harvard University Press, 1985.
Chisholm, Roderick. The Foundations of Knowing. Minneapolis: University of Minnesota Press, 1982.
Chisholm, Roderick. Theory of Knowledge. Englewood Cliffs, NJ: Prentice-Hall, 1966; 2nd ed., 1977; 3rd ed., 1989.
DePaul, Michael, ed. Resurrecting Old-Fashioned Foundationalism. Lanham, MD: Rowman & Littlefield, 2001.
Foley, Richard. The Theory of Epistemic Rationality. Cambridge, MA: Harvard University Press, 1987.
Fumerton, Richard. Metaphysical and Epistemological Problems of Perception. Lincoln: University of Nebraska Press, 1985.
Haack, Susan. Evidence and Inquiry: Towards Reconstruction in Epistemology. Oxford: Blackwell, 1993.
Moser, Paul K. Knowledge and Evidence. New York: Cambridge Unversity Press, 1989.
Putnam, Hilary. "'Two Dogmas' Revisited." In Realism and Reason, Philosophical Papers. Vol. 3. Cambridge, U.K.: Cambridge University Press, 1983.
Sellars, Wilfrid. "Givenness and Explanatory Coherence." Journal of Philosophy 70 (1973): 612–624.
Sosa, Ernest. Knowledge in Perspective: Selected Essays in Epistemology. Cambridge, U.K.: Cambridge University Press, 1991.
Van Cleve, James. "Foundationalism, Epistemic Principles, and the Cartesian Circle." Philosophical Review 88 (1979): 55–91.
Wittgenstein, Ludwig. On Certainty, edited by G. E. M. Anscombe and G. H. von Wright. Oxford: Blackwell, 1969.
R. W. Ashby (1967)
Bibliography updated by Benjamin Fiedor (2005)