Totalitära staters undermåliga information om resultaten av dess beslut

Igår var jag som sagt på Kristian Gerners föredrag ”Karl Poppers filosofi och kommunistsystemets kollaps” på Filosoficirkeln. Måste säga att jag blev lite besviken: det var för lite filosofi. Gerner pratade mest om hur Poppers idéer om reformism och öppet samhälle influerade politiska aktörer i Öst, inte minst genom George Soros arbete.

Jag hade hoppats på något mer i stil av en popperiansk analys av faktorerna bakom kommunismens kollaps. Exempelvis som ett resultat av den informationsbrist som uppkommer i en planekonomi (där all den information som ges av marknadens prismekanismen saknas) och där fri debatt inte tillåts. Enligt Popper får alla politiska åtgärder oavsiktliga och oförutsedda konsekvenser, och ett samhälle som saknar mekanismer att uppmärksamma och handskas med dessa kommer ofrånkomligen att stagnera.

Hur som helst, Gerner nämnde att medan han skrev texten till föredraget så hade han råkat höra en uppläsning av Orwells 1984 på radion, vilket han anmärkte var en passande inramning. Detta fick mig att tänka på en recension som Orwell gjorde av Bertrand Russells bok Power: A New Social Analysis. Tidpunkten är januari 1939, och Orwell ser en mycket mörk framtid. Därför blir han frustrerad av Russell: ”like all liberals he is better at pointing out what is desirable than explaining how to achieve it.”

[. . .] [He] merely utters what amounts to a hope that the present state of things will not endure. He is inclined to point to the past; all tyrannies have collapsed sooner or later, and ”there is no reason to suppose (Hitler) more permanent than his predecessors”.

Underlying this is the idea that common sense always wins in the end. And yet the peculiar horror of the present moment is that we cannot be sure that this is so. It is quite possible that we are descending into an age in which two and two will make five when the Leader says so. Mr Russell points out that the huge system of organized lying upon which the dictators depends keeps their followers out of contact with reality and therefore tends to put them at a disadvantage as against those who know the facts. This is true so far as it goes, but it does not prove that the slave-society at which the dictators are aiming will be unstable. It is quite easy to imagine a state in which the ruling caste deceive their followers without deceiving themselves. Dare anyone be sure that something of the kind is not coming into existence already? One has only to think of the sinister possibilities of the radio, state-controlled education and so forth, to realise that ”the truth is great and will prevail” is a prayer rather than an axiom.

Kring detta stycke kan det vara värt att reflektera med utgångspunkt i Popper och Hayek. Jag har tyvärr inte tid att göra det ordentligt just nu. Men notera exempelvis att idén om ett informationsgap mellan de styrande och de styrda är förenlig med att den mängd information de styrande besitter hade varit mycket större om de inte samtidigt hade haft behov av att upprätthålla detta gap. Kristian Gerner pratade exempelvis om de sovjetiska försöken under Krusjtjovs tid att få till stånd en genuin samhällsvetenskap. De steg som skedde i denna riktning togs sedan tillbaka av ett mer räddhågset sovjetiskt ledarskap.

George Orwell, An Age Like This, (Collected Essays, Journalism & Letters, vol. I), s. 375–6.

Om den förtjusande Professor Russell (Isaiah Berlin)

Brev till Felix Frankfurter, 23 augusti 1937. Isaiah Berlin beskriver oenigheterna inom hans college angående förslaget att bjuda in Bertrand Russell till Oxford under en termin. Det fanns många som var ytterst fientliga till detta förslag, framförallt med tanke på Russells radikalism och okristliga moral. . .

My conversation with the Warden [of All Souls college] I cannot reproduce, it was absolutely wonderful: he plainly hates Russell & all connected with him, but finally said ‘I suppose, at bottom, he is a very lovable kind of man?’ really I didn’t know what to say. I might have said, I suppose, ‘Many women appear to have found him so, sir’ but didn’t, only just though.

Isaiah Berlin, Flourishing. Letters 1928—1946, Chatto & Windus, London, 2004, s. 250.

Dewey v. Russell on Truth

Ett paper från 2006.

John Dewey, just like William James, his predecessor as the main figure of American Pragmatism, received critique from across the Atlantic by the eminent Bertrand Russell. The present paper will concern Dewey’s reply to Russell as it is put forward in his article ‘Propositions, Warranted Assertibility, and Truth’ (1). Dewey and Russell represent different schools of thought, and their disagreements are likely to involve the fundamental points where American pragmatism break with the tradition of European analytic philosophy. This paper thus focus on the fundamental issues where Russell and Dewey disagree; it will clarify the different standpoints, and show how the issues are related to the wider views in question; at the end Dewey’s argument against Russell’s classical correspondence theory of truth will be criticized. The conclusion of the paper becomes that both philosophers talk past each other and that none of them succeed in showing the inadequacy of the opponent’s general conception of philosophy. Dewey’s view is that Russell has not understood his theory, nor his wider perspective. In the course of the paper it should become clear that Dewey is right in thinking that Russell has misread him, and that Russell’s criticism does not work against Dewey. However, it will also be argued that Dewey’s criticism of Russell’s theory does not necessarily work either; he fails to convince that the classical correspondence theory must be given up, because his arguments against it only work from assumptions that can well be rejected.

Both Russell and Dewey proposed what they called correspondence theories of truth. This, however, does not mean that the differences are minor. Rather, the differences are vast. But as both philosophers state that their theories are correspondence theories, there is a natural framework for trying to understand the differences between them. If the theories both are looked upon as correspondence theories in a general sense, then the differences might be construed as differences about what is it that corresponds, and about what is being corresponded to. To put it in another way, we can regard both theories as different kinds of correspondence theory: what they have in common is that there is a correspondence relation — what they disagree on is what is thus related.

The correspondence relation for Dewey holds between a hypothesis, idea, theory or assertion on the one hand, and a problematic situation on the other. This correspondence, Dewey writes, is in the sense of a key answering to the conditions imposed by a lock [p.107]. It should be obvious that this cannot be taken as an independent truth definition in the ordinary analytical sense, but that it must be part of a wider view. This is Dewey’s view of human inquiry. The start of an inquiry is an indeterminate situation, usually coming from some practical matter. The first step is to getting a clearer and more specific picture of the problem; the indeterminate situation becomes a “problematic situation”. The inquiry then proceeds by coming up with possible solutions to the problem That is, hypothesis that might help to act upon in the problematic situation. Evaluation then takes place; possible consequences of different solutions are considered, and by doing this the relative value of the solutions are estimated. The final test is when a solution actually is tried as a guide to action. The question of truth and falsehood come into play at this last stage of inquiry, and it is dependent on whether the consequences of acting upon the hypothesis under test are such that they resolve the problem and settle the indeterminate situation. Here, it is important to keep in mind the fundamental that role that Dewey ascribes to the indeterminate situation. It is our conceptions of this situation that guide the rest of the inquiry, and that determines how to evaluate and praise ideas and hypotheses. The indeterminate situation is the beginning of an inquiry, but also controls it throughout [p.207]. Coming back to the correspondence relation, it could be said to hold between the between the first and last stage of inquiry.

Bertrand Russell, in his Inquiry into Meaning and Truth, published 1940, characterized John Dewey’s theory of truth as a theory that simply substitute “truth” with “warranted assertibility”. Russell says that he himself, on the other hand, endorses the correspondence theory of truth. His version of the correspondence theory is the classical one: the relation is between statements, propositions or beliefs to an independent reality, to empirical facts. A statement is true if what is referred to also posses the properties that the statement asserts. Dewey’s theory is obviously radically different from this, and in Russell’s view it does not even deserve to be called a correspondence theory. Russell’s description of Dewey’s whole theory of inquiry and truth is as follows (2): individuals engage in inquiry with the purpose of better interacting with their environment. In an inquiry, “assertions” are tools, and these assertions can be “warranted” to different degrees. The degree to which an assertion is warranted is determined by their ability to produce the desired results. During inquiry assertions can come to be replaced by better assertions, and sometimes they are the very means that lead to better assertions. The term “better” simply means that it produces more of the desirable results, lets us cope better with our environment, and hence “better” could basically be substituted with “more warranted”. An inquiry does not end; no assertion is the best for all times. The important point in this summary is that an assertion is warranted if it produces the desired results, and that the idea of truth thus loses its static and privileged nature.

These two typical pragmatist points are also what Russell find troublesome. He regards as untenable what he sees as the core of Dewey’s theory: that usefulness of an assertion is the only thing that counts in favour of it being true. Dewey, he says, reject the traditional view that an inquiry should lead the inquirer to come to know something about the world and that it is in virtue of this knowledge that inquirer’s actions become more successful (3). Instead Dewey holds that the success of an inquiry is simply successful action at the end of it. Assertions made in the course of this inquiry is thus valued only in terms of success; valued to the extent they lead to desired results.

The difference between Dewey and Russell seems very deep. The essential difference, Russell concludes, regards the question how it is decided which assertions falls under “true” and which falls under “false”. For Dewey, it is decided by the effects of the assertions; while Russell it is decided by the causes of the assertions (4).  In Russell’s theory there are antecedent facts that determine the truth of the propositions that are asserted. In his correspondence theory there are independent facts that have the role of ‘truth-makers’. Russell, in comparison to Dewey, maintains the objectivity and independence of truth and “the world.”

In the course of ‘Propositions, Warranted Assertibility, and Truth’, Dewey both defends his own position from Russell’s criticism, and levels an attack on Russell’s version of correspondence theory. Often, of course, these separate issues are intertwined.

Dewey does not object to the view that his theory could be said to substitute “truth” with “warranted assertibility”, but does object that Russell does not look at the context and conditions for how this is done. Dewey denies that his analysis of “warranted assertibility” simply means that assertions are warranted by their usefulness in reaching “desired results” [p.201]. The reason Dewey’s theory look objectionable as it is described by Russell is that we intuitively would not like to regard any spurious but beneficial consequence of an assertion as something that make it more warranted. This is the deep intuition that works against pragmatism: that it is useful to believe something does not show that it is true.

However, when Russell talks about “desired results” and “successful action” he is not giving the conditions and requirements that Dewey’s theory actually contain, and hence it is a somewhat false picture that Russell paints. What Russell has failed to see when he counts any satisfied desire and every type of consequence as relevant, is the extent to which the indeterminate and problematic situations are influencing the inquiry: they are not only the staring point and source of the inquiry, but controls it throughout [p. 208]. They set conditions for which consequences can be counted as giving an assertion any validity, and it is they that are to be satisfied at the end of enquiry.

Dewey thinks that the root of the misunderstanding must be that Russell not understood the impersonal character of the indeterminate situation. For had he done that, Russell would never got into thinking that just about any personal desire were relevant in evaluating theories. What it means that the indeterminate situation is impersonal or “objective” is that it is not dependent on a doubting subject. Rather it is, in one sense, the other way around: a doubt is only valid if it is evoked by such an indeterminate situation [p.209]. If it is not, the doubt is “pathological” on behalf of the doubting subject. Russell however, cannot imagine a doubtful situation without a personal doubter (5).  What Russell has failed to grasp, presumably, is that indeterminate situations are existential situations that the human, as an organism, finds itself in when it interacts in its environment.

That Russell misread Dewey shows that he did not understand the impersonal character of the indeterminate situation, and that he did not understand the impersonal character shows, in turn, that he never really understood one of the core parts of Dewey’s whole pragmatic project in epistemology. That is the naturalization of epistemology. Dewey writes that the question of the objectivity of indeterminate situations is intimately linked to his wider view of “experience” as behavioral. The objectivity of the indeterminate situations is given in this biological sense, as an imbalance in the interactions between organisms and environment. As inquiry starts with this kind of situations it seems that we get a whole other picture of epistemology than the traditional one; we get a picture of a human practice that is always carried out as a matter of course, the questions are practical rather than intellectual, and even have a Darwinian sense in that they concern how we cope with the world rather than how the world is.

Now, this project of Dewey is far from Russell’s conception of philosophy, and it is perhaps not surprising that he failed to internalize it sufficiently. Russell is working with a traditional conception of knowledge, world and truth. Russell’s account of correspondence relation is of epistemological character as opposed to Dewey’s experimental-behavioral [p.211]. Knowledge in Russell’s traditional conception is a question of individuals doubting and believing things about an independent reality. This conception of Russell’s is, however, on Dewey’s view untenable, or at least fruitless. Dewey levels an attack on the positions that Russell has gotten into. On Dewey’s view, the subjectivistic psychology of a doubt and of knowledge, and traditional ideas of certain belief and independent truth, have lead Russell to obvious contradictions. This is especially so with Russell’s position that there is a causal link between an assertion and an antecedent fact.

Dewey’s first point of criticism of Russell concerns this relation between experience and propositions. On Russell’s theory there is a possibility that propositions can be known directly by some immediate experience [p.202]. Atomic propositions such as “This is red” or “redness-here”, in virtue of its simplicity and the directness of the sense experience, can be known directly and with certainty because of a causal relationship between the percept and the proposition. Dewey challenges Russell on this point. First of all is there no self-assured meaning of the term “here”, rather the alleged simplicity of “redness-here” disappears on reflection, because it seems that any such assertion presuppose a theory of space-time [p.202]. In addition, a causal link between a sense experience and a judgment seem to presuppose a physiological theory to guarantee that the assertions are correct. The causal link is thus not as direct as Russell assumes. This is of great importance. Russell has distinguished a class of judgments such that they are caused by what it asserts, and hence cannot fail to be true. These “basic propositions” are fundamental, because the justification of more complex propositions depends on their relation to them. Dewey thus mounts an attack on the fundamental points of Russell’s epistemology. An attack on the “purity” of these basic propositions must shake the whole mass of knowledge as it is construed in Russell’s account. This critique of Russell is very much in line with Dewey’s pragmatism and conception of philosophy. By trying to find secure foundations, Russell is engaged in a quest for certainty just as Descartes was, even though it is now about experience rather than rationalistic principles. This is a project that Dewey thinks it is time to give up on. Dewey remarks that Russell’s theory in this way give good reasons for complete skepticism, because the assumption is that a condition for knowledge is certainty, and this certainty is plainly impossible, which means that we end up in skepticism.

The argument that Russell must end up in skepticism also comes up in another form, namely as a fundamental criticism of the correspondence relation on Russell’s view. If the proposition “This is red” is true when caused by a simple “atomic” event, then the question arises how we know whether it actually is caused by such an event. That is, can we test correspondence? The atomic event both causes the proposition and at the same time is what makes the proposition true (it is its verifier). Dewey’s charge is that this becomes absurd, because it is the proposition itself that is our belief and our only way of speaking about the event [p.206]. It cannot be possible to look at both the proposition and the event to judge if they correspond, because it is through the proposition that we can talk about the event in the first place. And if such a comparison could be done it seems that propositions are surplus, without function [p.209]. So either comparison is possible, which make propositions surplus, or the comparison is impossible, which would mean that there is no possible way of telling when the correspondence relation is at hand. This shows that the classical correspondence theory leads to skepticism, because it can never be tested whether the truth conditions of a proposition obtains or not, which means that we can never justify a proposition and regard it as knowledge.

Dewey wants us to give up on the classical correspondence theory of truth and the traditional view of knowledge. From where Dewey stands, the theories of Russell seem detached from practical reality and hence rather pointless. However, Dewey’s arguments against the correspondence theory is not necessarily successful in its aim to make way with the conception of truth as an independent and objective property of propositions or beliefs about the world.

Dewey’s argument seem to say, that as we can never have direct access to the event independent of the proposition we can never really have any knowledge. The correspondence theory in this way leads to skepticism. This points to the fine balance that one who holds a correspondence theory must take into account. One of the motivations for holding a correspondence theory, presumably, is to save our intuition that truth is not dependent on our beliefs, and that no matter how justified we are in a belief there is always the possibility that we may turn out to be wrong. Truth just cannot be reduced to justification or “warranted assertibility,” because if so that possibility seems, on the realist account, to disappear. On the other hand, when the truth is given this transcendent status the effect necessarily is that the justifiability claims become more insecure. To cure the insecurity the philosopher wants a safe link somewhere between the two realms; in Russell this is the causal relationship between the atomic event and the basic proposition. In this there seem to be conflicting ideas: the transcendence of reality but at the same time the existence of some propositions that one can know with certainty. The transcendence is combined with knowing with certainty. It seems like it is this contradiction, if we may call it that, that Dewey’s argument concern.

Dewey might very well succeed in the argument against Russell’s basic propositions. But if taken to be the wider argument that correspondence theory makes knowledge impossible because it is impossible to compare the proposition with the event itself, then there seem to be ways to counter the argument. The point is that skepticism does not necessarily follow, because what is shown is that we cannot determine when the correspondence relation actually occurs, but not that it in fact cannot occur. The more correct way to put it then, would be to say that we could never know if we have a true belief, not that we never can have knowledge. If knowledge is defined simply as true belief, the objection that it is impossible to decide when this is the case does not threaten the possibility of knowledge. The skepticism charge against classical correspondence theory seem to rely on the assumption that to know something it is required that we have criteria for true belief, and so that the belief can be possible to justify. The correspondence theory can be defended against this critique, by settling with true belief as the only conditions for knowledge: knowledge without justification.

This move however, will probably not be found satisfactory to Dewey and pragmatism in general. The argument took the impossibility of testing and justifying claims of correspondence to the conclusion that knowledge with this conception of correspondence gets impossible. To drop the demand for justification does not solve the problem, because the very problem was that correspondence transcended justification. To talk about propositions being true without any possible way of justifying it, is the core problem. The underlying pragmatic assumption behind the argument is the view that talking about properties that transcends our knowledge about them is meaningless.

The move to drop the demand for justification as condition for knowledge shows, however, that the pragmatist argumentation is not very conclusive. Rather it seems to depend on assumptions that may very well be rejected. The argument relies on the theory that a concept must be supplied with criteria for its use if it should be regarded as useful or meaningful at all. But it is not obviously faulty to think that the concept of truth can have a meaning even without the possibility to supply it with criteria that can determine whether a particular proposition is true or not. Nor that knowledge is impossible because it is impossible to determine when we have it.

These remarks are intended to show that Dewey’s argument against the objectivist or ‘trancendent’ view of truth only works from verificationist assumptions about the need for criteria. Moreover, it also seems to be the case that it is only from these assumptions that the transcendent view of truth will be found to be problematic in the first place. Hence, the argument does not carry far in the eyes of someone who does not share these verificationist assumptions of pragmatism.


(1) John Dewey, ‘Propositions, Warranted Assertibility, and Truth’, in The Essential Dewey, vol 2, Indiana University Press, 1998. The article was first published in Journal of Philosophy, 38, 1941. In the present paper, page numbers in brackets refer to The Essential Dewey.

(2) Bertrand Russell, Inquiry into Meaning and Truth, Allen & Unwin, London, 1961, p.319.

(3) Ibid., p.321.

(4) Ibid., p.325

(5) Ibid., p.323.