home page           'A Theory of Moral Objectivity' contents page

A Theory of Moral Objectivity (Section 4d - Analytic philosophy)

By Robert M. Ellis, originally written as a Ph.D. thesis 'A Buddhist theory of moral objectivity' in 2001. This html version copyright 2008.

A hard-copy paperback book of this thesis is now available from and also on Amazon, price UK£25 (or equivalent). This relatively high cost is necessary because it is A4 size and has 487 pages (296,000 words). This print version includes an index.

Support independent publishing: Buy this book on Lulu.

A downloadable pdf version of this thesis is available from the British Library at (you will need to search the original title 'A Buddhist theory of moral objectivity', and register with the ethos site, but registration is open and the download pdf is free for researchers). Alternatively you can download a pdf for a small cost from

Join discussion or ask questions on any aspect of the thesis on the new phpbb discussion board


4d)      Analytic Philosophy


i)                    Scientism from Hume to Frege


Whilst Hume’s scientism was representative of the forces that were gradually emerging to challenge eternalism in the eighteenth century, his thinking generally represented an extreme of thought in a society in which eternalism still ruled. That Hume’s Dialogues concerning Natural Religion[1] is such a masterpiece of indirectness and irony is one instance of the pressures against him still exerted by the Church of his time. So whilst Hume represents perhaps the first reasonably consistent scientistic philosopher, the forces he represented were in general still scattered and unsystematic during the eighteenth century. They could be seen more in the growth of political individualism than in other areas of philosophy: the arising of rightsism in Thomas Paine’s Rights of Man and its triumph in the democratic and egalitarian movements of the American and French Revolutions. However, as I have already commented[2], these movements, like that of the Enlightenment as a whole, were as much indications of an eternalist retrenchment as of a nihilist victory. The bourgeois concept of autonomy could be identified with either a Romantic or a Kantian understanding of transcendental humanism as well as a scientistic one.


Perhaps this is the reason why, although Hume had followers, he had no obvious successors in the eighteenth and early nineteenth centuries. Perhaps the descriptivist utilitarianism of Adam Smith, who was a personal friend of Hume, offers the nearest figure, but Smith in general avoided the more controversial areas of Hume’s doctrines, refusing even to arrange for the publication of Hume’s Dialogues after his death despite Hume’s request that he do this. The absolute utilitarians such as Mill and Sidgwick inherited many of Hume’s empiricist assumptions, but not the grounds of his ethics, and Mill in particular strove to reconcile scientistic arguments with eternalism[3].


Perhaps Hume’s most important influence was over Kant. Although Kant’s account of the basis of scientific knowledge was quite different from Hume’s, he was stimulated by Hume into realising the strengths of empiricist argument and into accepting that scientific knowledge could not be gained through reason alone[4]. Although Kant had great temperamental and philosophical differences from Hume, they shared a scientistic concern with the grounds of  knowledge. The fact that for Kant this concern was ultimately subordinated to an eternalistic desire to find absolute grounds for ethics did not prevent Kant from making a large contribution to the scientistic tradition. This contribution consisted particularly in his account of absolute space and time based on a priori intuition, and his account of the synthetic a priori, which attempted to give the absolute status of the a priori both the mathematics and to the primary laws of science.


Those who immediately followed Kant or were influenced by him, particularly the idealists, likewise maintained some of these rationalist tenets of thought, but appropriated them to a generally eternalist understanding of value. From about the time of the death of Hegel in 1831, however, German idealism began to lose its influence (only to re-emerge in both Britain and Germany later in the century) and was eclipsed by a Humean empiricism which became dominant in Germany until a further Kantian reaction set in about 1870[5]. The turn to empiricism was marked by the increasing influence of the natural sciences (particularly after Darwin), the dwindling in importance of philosophy as a distinct discipline, and a reaction against Christian dogmatism, which idealist philosophy was seen as supporting. This coincided with the rise of thinkers such as Mill in England and Comte in France who used scientistic arguments to support a transcendental humanism, as well as more fully scientistic thinkers such as Herbert Spencer (1820-1903), whose popularity in Victorian England showed to what extent capitalism and science together were gradually increasing the audience to which scientism could appeal.


In some ways Spencer’s evolutionism marked an important new stage in the development of scientism, as in its popularised form it could provide an ethic which could not only claim the authority of scientific knowledge, but could support the conventions of capitalistic competition through an appeal to nature. In some ways, however, this ethic was still strongly influenced by eternalism and merely marked a transitional stage in the movement from eternalism to scientism. It did not involve a mere appeal to conventional assumptions based on an argument that no higher standards could be known, like that of Hume, but an explanatory theory which appealed to a new form of cosmic justice whereby strength and adaptability were rewarded with success. An appeal to rules existing in the universe itself, rather than just the explanatory power of science, was still required, and the reward resulted not just from following scientific advice but from working with a power believed to operate in the universe. Popular evolutionism also gained little support from scientists themselves because it crossed the fact/value boundary so crucial to the scientistic identity.


The developments of philosophy between Hume and Frege, then, illustrate much more of the new ways in which eternalism responded to Hume's challenge, and of the ways in which Hume’s strict scientism could be compromised, than they do developments in the justification of scientism itself. During this period public receptivity to scientism increased immensely because of the practical results which science was achieving (in partnership with capitalism) rather than because any great advance had been made in their theoretical justification. During this period, too, the power of Christianity was gradually waning under the influence of individualism fostered by social and economic change, together with the questioning of biblical certainties produced by historical criticism of the Bible and new understandings of geology and biology. The result was confusion.


Into this confusion stepped a new breed of philosophers who saw philosophy quite clearly as a handmaiden of science. For them scientific knowledge was the main ground of objective value and the task of philosophy was to overcome the theoretical difficulties which might still stand in the way of the full acceptance of that scientific authority which had already been demonstrated practically. These were the analytic philosophers, so called because of their belief that the scope of philosophy should be strictly confined to analysis alone.


Although analytic philosophy as it later developed in the form of logical positivism was strongly empiricist and owed much to Hume, its initial development emerged as much from a scientistic type of Kantianism. At the centre of this initial development was the figure of Frege, whom I shall consider first.



ii)                  Frege


Frege himself appears to have written nothing about ethics, yet as perhaps the most significant  founder of analytic philosophy the new strategy he adopts to try to provide a firm grounding for science is crucial in providing the foundations on which later analytic philosophers would build in order to support a descriptivist approach to ethics. Throughout his career his chief motivation was the search for an analytic account of the truth of mathematics, but in the process of creating the conditions on which he believed he could do this Frege created a much more systematic account of representationalism than had ever previously existed. He tried to create an ideal logical language which would be capable of a perfect representation of the form of human knowledge, and in doing so circumscribed representationally meaningful statements by restricting them to only those which were capable of reduction to this logical form.


Frege’s approach is Kantian in that his argument is pitched at the level of the transcendental a priori, by which he believed he could identify the conditions for truth[6]. He does not, however, share Kant’s belief in the existence of a category of synthetic a priori distinct from the analytic, and his attempt to prove the analyticity of arithmetic, which Kant believed to be synthetic a priori, illustrates this. His method consisted primarily in the analysis of language, but later he was also obliged to introduce an element of epistemology[7]. Whilst Hume had based his epistemology on an empirical foundationalism, Frege applied a Kantian coherentism based on transcendental argument, but in doing so provided a new and much more sophisticated justification for the truth of scientific knowledge: that it lay within a sphere of meaning defined by logic. The implication of this was that the relativity of the truth of scientific knowledge could now be acknowledged because its privileged status was now supported instead from a different quarter.


The new criterion of meaning seems to be by far the most important development in Frege’s work, but to him it evidently seemed merely incidental, since he took scientism for granted and concentrated his fire on the Humean empiricism which he was reacting against. The passage in which Frege almost carelessly introduces his best-known criterion of meaning in the Foundations of Arithmetic illustrates this:

That we can form no idea of its content is therefore no reason for denying all meaning to a word, or for excluding it from our vocabulary. We are indeed only imposed on by the opposite view because we will, when asked for the meaning of a word, consider it in isolation, which leads us to accept an idea as the meaning. Accordingly, any word for which we can find no corresponding mental picture appears to have no content. But we ought always to keep before our eyes a complete proposition. Only in a proposition have the words really a meaning. It may be that mental pictures float before us all the while, but these need not correspond to the logical elements in the judgement. It is enough if the proposition taken as a whole has a sense; it is this that confers on its parts also their content.[8]


Frege does not stress here the way in which he believes meaning is limited to truth-conditions as it is limited to the logical content of propositions (though he does believe this), because he is concerned to refute the more naïve Humean belief that it is limited to the direct representational content of words when they are analysed into simple impressions. Nevertheless, the limitation of meaning to truth-conditions was to later form the basis of the distinction made by the logical positivists between scientific propositions with acceptable truth-conditions and other propositions without them.


In his own context, however, Frege is concerned to assert the power of his account, which is indeed an advance on a Humean form of representationalism because it takes the context of words into account. Hume took little account of the ambiguity of language, but Frege is acutely aware of it and thus wants to construct an artificial logical language which will purge natural language of all its ambiguities and enable us to see what is precisely represented. In order to do this he realises that he has to not only work with the meaning of sentences rather than of words, but take into account the context of sentences, which may alter the representational meaning of a sentence[9]. Like all coherentists, however, at some point Frege has to limit the extent of context which he will take into account: he will consider the context of, say, the whole scientific theory in which a sentence is embedded, but not the social or psychological context of that theory, its relationship to desire and action, or the role of the interpreter when she reads a sentence embedded in a theory. It was essential to Frege’s enterprise that he should not recognise contextuality to this degree because this would also involve the recognition of ways in which the ambiguities of natural language have a meaning related to their contextual functionality and not merely to their representationality.


Frege does indeed give careful consideration to terms in language which appear to be merely functional and not representational, but only within a framework of representationalism. His designation of functions as “incomplete” enables him to draw a contrast with the completeness of objects, which can be linked by functions so as to form propositions which have truth-value. Neither objects nor functions are said to have truth-value alone, but a proposition brings objects into a relationship, such as an identity or a set-relationship, by sticking them together using functions. Objects, however, are distinguished from functions because they are “complete”, meaning that they can be combined into identity relations with other objects[10]. Both objects and functions are logical, not verbal, constructions and must be distinguished from the terms in natural language that indicate their presence. Frege’s purpose in abstracting from natural language, however, is to provide a perfect representation, and this representationalism forces him to give a status to objects which he does not give to functions, for representationalism sees the world as consisting in objects and their relationships rather than either the subject (the focus of expressivism) or functions (the focus of pragmatism). In the logical world Frege wishes to construct, subjects must conform to the status of objects and functions only attain completeness when they are turned into value-ranges, which consist of the set of truth-values of all possible relations between objects that could be created by that function[11]. Frege’s distinction between complete object and incomplete function thus only serves a purpose within the logic he wishes to create, which presupposes a restriction of the meaning of language to a representationalist one.


Sluga’s comment on this weakness in Frege’s approach is indicative of the kind of weak argument which can be used to attempt to cover up the reliance of the representationalism in analytic philosophy on pragmatic distinctions (which merely serve the purposes of representationalism rather than being representational themselves):

At first, the distinction between the complete and the incomplete can be considered as merely conventional or pragmatic, but the need for the distinction justifies us in concluding that it stands for something real or objective.[12]


What is “the need for the distinction” but itself something conventional or pragmatic? And apart from this distinction and its role in the construction of  Frege’s logic, what reason is there for assuming that the relationships between objects in natural language are any more meaningful than the relationships between functions? Since Frege restricts the scope of the contextual influence on his logic to the representational context and does not extend it to the pragmatic context, no purpose other than a representational one is encountered and the way in which functions can be understood to relate to each other through a subject of experience is ignored.


Frege begins to incorporate epistemological assumptions as well as linguistic ones in the change which comes over his view about 1891 (though these assumptions could also be read back into his earlier theory), when his theory of sense and reference was in formation. Frege uses the distinction between sense and reference in order to try to make it clearer how an analytic statement of identity could be informative, since an identity of sense (i.e. of the meaning of two terms according to our understanding) is trivial, whereas one of reference (i.e. of two terms of which the sense differs, revealed to refer to the same object) is claimed to be informative[13]. In order to determine that the reference of two senses is identical (such as Frege’s Morning Star and Evening Star, which both refer to Venus) epistemological criteria are required, but Frege claims that the procedure, though cognitive rather than merely linguistic, is still analytic. The purpose of this is to change the Kantian synthetic a priori, which included arithmetic, into a purely analytic a priori with the same status as logical truth. A simple sum such as 7+5=12 is not just an analysis of the meanings of the numbers in the way they are designated, but nevertheless when analysed contextually according to Frege’s new logic its propositional meaning will prove to be universally true without any recourse to empirical investigation. In his scheme the numbers should be understood as the value-ranges of functions, and for that reason can be considered complete logical objects[14]. Identity relationships between numbers, like those between other identities of logic which are also informative, are thus claimed to be identities of reference.


Frege’s failure to prove the analytic status of numbers is thus an example of a failure to prove the independence of reference from sense. His project was proved to have failed in logical terms by Russell’s discovery of a contradiction in his logical account of sets (a contradiction which also applied to propositions)[15]. According to Frege a set is a value-range of a function which must be defined, not according to its contents, but according to the logical extension of the function alone, making empty sets and sets of one member entirely explicable. Russell pointed out, however, that “there is no class (as a whole) of those classes as wholes that do not contain themselves. From this I conclude that under certain circumstances a definable set does not form a whole”[16]. The effects of this are far-reaching, as both Frege and Russell recognised, since it revealed a respect in which ambiguity could not be removed from a purified logical language without creating contradiction, and a respect in which reference was not independent of sense and thus could not be used to consistently represent truth-value.


Curiously neither Frege nor his successors took this as an indication that their representational assumptions might not be adequate. Instead, the combination of unquestionable representationalism and logical paradox led Frege merely to despair of a way forward, though he never ceased in his devotion to the hopeless project. One of Frege’s letters to Russell makes his unquestionable allegiance to scientism clear together with the way in which he felt completely trapped by the implications of that allegiance:

I myself fought against the recognition of value-ranges and thereby of classes for a long-time; but I did not see any other possibility of giving logical foundations to arithmetic. The question is, how do we grasp logical objects? And I have not found any answer but the following: we grasp them as extensions of concepts or, more generally, as value-ranges of functions. I have never been unaware that this leads to difficulties and these have been multiplied by your discovery of the contradiction. But what other way is there?[17]


For him there simply was no other way but to continue to seek this authentication of the nature of the truth of science, just as for many medieval thinkers the function of philosophy was merely to give an account of the truths of God. In this he reflected both the attitudes of his context and the psychology of dualism.


Even if Frege had not encountered purely logical contradictions which prevented him from creating a perfectly unambiguous logical language, his project would still appear hopeless as a way of providing support for the status of scientific knowledge. Whilst the terms of a logical language may be unambiguous, this purity is purchased at the expense of any possible relationship to ambiguous empirical objects. This difficulty is recognised by some of Frege’s more empirical successors such as A.J.Ayer, who reinstate Hume’s clear-cut view that all a priori propositions are analytic, and analytic propositions are tautologous and uninformative[18]. At the same time, Ayer, like the other logical positivists, adopts the connection between truth-value and meaning first proposed by Frege despite the fact that according to his view of analytic propositions this truth-value must be entirely abstract and uninformative, inapplicable to empirical objects.


As the most significant founder of analytic philosophy, Frege is quite a touching figure, devoted to his doomed project despite his awareness of its limitations. Although, like many dualists before him, he wanted desperately to reconcile the absolute with the relative, he stuck firmly to the a priori methods of Kant and worked with slow persistence to try to create a sort of launching pad amongst the absolutes of the a priori from which the relative could be reached. Those aspects of his theories which were adopted by the analytic philosophers who followed him were largely incidental details for him, revealing him as a dignified figure in contrast to the scramble of dogmatic attempts to give an absolute (or at least privileged) status to science which followed him.


iii)                Logical positivism 


In the movement known as logical positivism, which grew up through the combination of Humean empiricism and Fregean analysis, we find the defining core of the whole project of analytic philosophy. In it we find the conviction that the nature of philosophy could be revolutionised through the application of linguistic analysis, turning philosophy, castrated of all pretensions to metaphysics, into a eunuch in the service of victorious science. Whilst this movement is mainly concentrated into the first half of the twentieth century, gaining its classic expressions in the Vienna Circle and the Cambridge Apostles, I shall argue later that in many respects it refuses to die and has had great influence over later twentieth-century analytic philosophy.


Oswald Hanfling identifies three main features in logical positivism: the verification principle, the elimination of metaphysics, and the unity of science[19]. Of these the unity of science is perhaps the most central: “The idea of the unity of science was based on the belief in the uniformity of language, and this in its turn on the view that there must be a single type of act of verification and a single type of verification-statement corresponding to it”[20]. This uniformity in science could then become the basis for the rejection of any kind of investigation (labelled “metaphysics”) which did not conform to the verification criteria.


The scientistic values which informed logical positivism were clear not only from the explicit concern with converting all scientific statements into the same currency, but from the zeal with which the logical positivists set about imposing a binary distinction on all statements as either verifiable (and therefore “meaningful”) or unverifiable (and therefore “meaningless” or “nonsensical”). The fact that of course they would not have translated these categorisations into such metaphysical terms as “good” and “bad” only served to obscure to them the extent to which they were merely setting up a new moral vocabulary. Occasionally it would occur to the logical positivists that their vocabulary was supposed to be a neutral and descriptive one in harmony with the basis of the claim of objectivity in science, in which case they would claim that their language had no connotations of value, as in this ‘Advice to the Reader’ placed at the beginning of Carnap’s The Unity of Science:

Nonsense (or pseudo-expression) is intended to carry none of its usual abusive connotation. Technical use = whatever cannot be verified in experience.[21]


This serves to illustrate the contradiction at work in the whole movement of logical positivism, whereby a prescriptive position claims its authority covertly by assuming a mantle of “neutral” descriptivity. It is an approach already used particularly by Marx and more broadly by the whole naturalistic tradition of eternalism. The implicit moral prescription of logical positivism rested on the appeal to the authority of science and the possibility of discovering natural laws which was attributed to a verifiable scientific approach. If we follow these verifiable procedures, the logical positivists assumed, we must gain better results than if we do not: cosmic justice in epistemological form.


That these values effectively provided a rationalisation for established conventions is clear not only from the logical positivist treatments of ethics (which I shall discuss later) but from their whole conception of philosophy. This can be seen in Frank Ramsey’s definition of philosophy:

In philosophy we take the propositions we make in science and everyday life, and try to exhibit them in a logical system with primitive terms and definitions etc. Essentially a philosophy is a system of definitions or, only too often, a system of descriptions of how definitions might be given.[22]


Just like many dualist philosophers before them, the logical positivists saw themselves as supporting and rationalising assumptions and practices that already existed. The difference between their case and that of earlier philosophers (such as Kant) was that they saw this process of rationalising not in moral but purely in cognitive terms. The effect of this denial of the moral element of their goals meant that these goals remained unexamined, conforming to the conventions of their society and/or to their individual inclinations.


The logical positivist claim to separate “science” from “metaphysics” rests on the verification principle. In the form offered by the early Wittgenstein, this was that “The meaning of a proposition is the method of its verification”. A proposition here is distinguished from a sentence as being a logical construct (as identified by Frege) capable of truth or falsity because shorn of its contextual ambiguities. As Hanfling points out, such a proposition is already by definition meaningful on logical positivist criteria: the logical positivist is therefore concerned with distinguishing real propositions from sentences which merely appear to contain such[23].


This distinction is made on the basis of an appeal to possible experiences which would be produced by a method of verification. Whether a statement implied possible experiences which could be used to verify it could be found by a process of analysis. This analysis led either to a logical atomism of “elementary propositions” (in the case of the early Wittgenstein[24]) or to a phenomenal atomism (in the case of Ayer or Waismann[25]) in which either the logical components through which we interpret sense-experience, or the description of sense-experience itself in the form of “sense-data”, is broken down into elementary phenomenal propositions which permit the possibility of verification.


It is conceded by most logical positivists, however, that such verification is incomplete. For example, the number of observations required to completely verify any given object’s existence would be infinite. But Frege had already provided the logical positivists with the basis of a way of avoiding this sceptical problem in his entirely formal account of meaning. It is the coherence of the verifiable account, demonstrated by its reducibility to logical form, that counts in setting it apart from unverifiable accounts, rather than its foundational basis in experience. Carnap and Neurath took precisely this road, believing that formal linguistic analysis was all that was necessary to establish the verifiability of statements, with no need for an appeal to experience at all[26]. Ayer, on the other hand, made a distinction between verification (which he believed to be impossible in a “strong” sense[27]) and verifiability. In order to distinguish between verifiable and unverifiable statements he argued that it was only necessary to define a verifiable statement as capable of verification in the weaker sense, that is “that some experiential propositions can be deduced from it in conjunction with certain other premises without being deducible from those other premises alone”[28]. Ayer later specified that these “other premises” must be either analytic or themselves verifiable[29].


Here Ayer came rather closer than his fellow logical positivists to identifying a genuine difference between metaphysics and hypothesis. In its basic structure, in fact, this account of verification foreshadows Lakatos’s “negative heuristic” (the core statement), “auxiliary hypotheses” (the “other premises”), and “positive heuristic” (the deduced experiential hypotheses), which I have already discussed above[30]. In adopting this structure Ayer is recognising that the core statement of a scientific theory is often only verifiable in a very indirect and partial way, but that nevertheless some criteria need to be applied to identify helpful theories and differentiate them from unhelpful ones. However, Ayer is still burdened by logical positivist assumptions which Lakatos has shaken free of. His criteria of verification are still understood as criteria of meaning and hence as ways of providing a clear-cut reason for the rejection of metaphysical hypotheses; and so he fails to recognise the ways in which falsification can provide relatively more definite information about the acceptability of a theory than verification, despite the fact that Popper had already suggested this at the time he was writing[31].


As I have already argued, philosophy cannot provide a complete rational account of objective criteria for the acceptability of theories or of other beliefs. This argument is confirmed by the widely accepted fact that the logical positivists failed to do so. In the absence of such positive criteria for verification, then, there are also no completely rational grounds for verifiability: for there are no positive grounds for assertion of a complete distinction between statements that can be verified and statements that cannot. Ayer argues that it is a sufficient criterion of verifiability that there should be some evidence making a statement  derived from the one under examination probable, yet we are still left with a need for judgement as to what constitutes relevant evidence contributing to probability. Given Ayer’s specification that auxiliary hypotheses should also be either analytic or themselves verifiable, it seems that a further judgement as to the acceptability of auxiliary hypotheses will be required in order to curtail an infinite regression of verifiability-requirements for the auxiliary hypotheses that may be required to show the auxiliary hypotheses to be verifiable. Ayer’s attempt to find rational criteria, like that of Lakatos, only points to its own limitations and indicates that the only criteria of objectivity in such cases can be psychological.


Ayer makes generally less progress than Lakatos in obtaining even relatively valid rational criteria partly because of his adherence to verifiability rather than falsifiability. Ayer rightly argues that falsification is just as relative as verification because it depends on the acceptance of an observation-statement which might falsify a theory as true. Yet despite this, criteria  which specify a way in which a theory may be falsified have a value in providing a clear indication of a way in which a universal theory may be related to experience which criteria of verification do not have. An apparent verification, as Lakatos points out, may be encouraging, but only provides one example of the apparent truth of a theory with universal application, and thus falls foul of the problem of induction. An apparent falsification, however, could conclusively show a problem either with a theory or its auxiliary hypotheses provided that the observation-statement that indicates it is acceptable. Even if no actual falsification is conclusive, then, the ability to specify a possible falsification for a theory provides a way in which it can be tested for a relationship to experience which is not available through verification.


To illustrate this it is only necessary to consider the central assumptions of the eternalist tradition, which largely produced the “metaphysics” that Ayer was attacking. The hypothesis of absolute value was supported by the auxiliary hypothesis of cosmic justice, implying experiences of  requital at some point in the future. Such a hypothesis can be repeatedly verified through experiences which are interpreted as just requital for previous actions, yet no possible falsification can be specified for it, since pleasant or painful experiences that seem to be undeserved can always be given ad hoc explanations such as the just requital of secret or unremembered deeds, a type of causality outside the operation of cosmic justice, or simply God’s inscrutable will.


Ayer’s adherence to verifiability rather than falsifiability is perhaps connected to the ways in which his own scientistic allegiances still depended on the same dynamic. His arguments, like those of all the logical positivists, ultimately rest on Frege’s appeal to the coherence demonstrated through the reducibility of statements to his logical system, yet like all coherentisms this is unfalsifiable because it will accept no terms of defeat outside itself. Ayer states the principle of verification itself to be a definition rather than an empirical claim, and concedes that the term “meaning” may be defined differently from the way it is understood in the principle. Nevertheless, he goes on to support the special status that he believes should be given to the principle through an empirical claim:

It is indeed open to anyone to adopt a different criterion of meaning and so to produce an alternative definition….Nevertheless, I think that, unless it satisfied the principle of verification, it would not be capable of being understood in the sense in which either scientific hypotheses or common-sense statements are habitually understood.[32]


Ayer seems to feel that an appeal to common-sense is sufficient here, given the way in which, like the other logical positivists, he sees himself as merely defining and rationalising accepted conventions. For him the field of science and common-sense delimits “meaning”, and the strength of his case rests on the respects in which this is true for the world as represented according to the dominant beliefs within these spheres. Yet this approach is essentially inward-looking, satisfied with a coherence, not expecting further challenges from beyond. Verificationism is the best way of supporting such an approach because it constantly reaffirms it, and appeals to the a priori provide it with a spurious universality.


In contrast to the verificationists, the falsificationism of Popper and Lakatos is much more representative of the open psychological disposition underlying the leading edge of science. Though they share some of the weaknesses of the logical positivists, they are much more inclined to understand science undogmatically as an open and incomplete set of hypotheses which are merely modified by each falsification. The criterion of falsifiability helps to differentiate theories which are helpful because open to experience from those which are closed because they are not (i.e. because they are Dogmatic), whereas verifiability only offers a criterion for separating two types of closed theory from each other, created by imposing the expectations of a given coherent set of principles on all other theories. As Lakatos writes: “Intellectual honesty does not consist in trying to entrench, or establish ones position by proving it (or ‘probabilifying’) it – intellectual honesty consists rather in specifying precisely the conditions under which one is willing to give up ones position”[33].


The idea that the logical positivists were replaying scenes of the eternalist drama in scientistic terms can also be applied to their attitude to language. Effectively Frege opened a new vein of absolutism through the claim that science could be justified through recourse to an ideal logical language, and verificationism merely develops this approach. That this approach proved so appealing can perhaps be accounted for in terms of the failure of previous versions of absolutism, whether based on Plato’s Forms, Stoic empiricist foundationalism, Christian revelation, or Kant’s transcendental a priori. The pattern of dualism according to which an absolute verbal justification of conventional doctrines needed to be found continued, but the rational failure of previous versions led it to emerge in a new linguistic form. What distinguished this new form particularly from its predecessors was its unquestioned acceptance of Hume’s fact-value distinction, which I shall consider in the next subsection, yet the distinction between ordinary relative language and verifiable absolute language persisted not in spite of but on the strength of this distinction.


In time, however, this appeal to absolute language predictably created a reaction within analytic philosophy, and a swing to ordinary language particularly led by the later Wittgenstein, whose work will be considered in the next section. This reaction led analytic philosophy closer to an existentialist position in which not only values were relative, but the position from which that view was asserted was also understood in contingent terms. Nevertheless this more recent tradition maintains some of the basic scientistic principles of logical positivism, as I shall argue later[34].


The duality between ideal and ordinary language, then, is just another version of the absolute-relative distinction. This duality appeared in the work of the logical positivists as the distinction between analytic and empirical statements and through the attempt to give some (scientific) empirical statements an analogous status to that of analytic certainty. The grounds of the dichotomy between analytic and synthetic were later strongly attacked by Quine, who showed, even on an entirely analytic basis, that the standard ways in which logical positivists had defined analyticity had presupposed their own conclusions[35]. As with previous versions, this dualism between types of language can be avoided by resort to incrementality. Sceptical arguments about the most basic empirical expressions make it clear that even these are to some extent hypothetical and have an ambiguous relationship to experience, whilst even Ayer was prepared to admit the hypothetical nature of most scientific theories. Frege’s failure to prove the analyticity of mathematics also suggests that even mathematical statements may be regarded as to some extent hypothetical. Apart from strictly analytic logical statements with no reference to the world, then, the wasted labours of the earlier analytic philosophers to impose their linguistic duality on the world seem to have revealed, not that any such duality can be satisfactorily believed in, but rather that there exists an incrementality in types of statement within a broad hypotheticality. At the one extreme exist statements with a relatively precise representational correlation to experience within a certain framework of belief, whether this experience is understood in more concrete specific terms or general and abstract theoretical terms. At the other are statements with an extremely vague and ambiguous relationship to experience rendering them easily manipulable in the service of dogmatic ideologies.


This accounts for both the power of logical positivist attacks on metaphysics in their context, and for their failure to disprove the meaningfulness of metaphysics. For the logical positivists (and their successors, the falsificationists) were able to point to the rational correlates of dogmatism as a psychological disposition, and thus appeal to those of a more sceptical disposition who already suspected the harmfulness and ethical ineffectiveness of the dogmatism in the eternalist tradition. Yet because dogmatism (as a psychological disposition) has no necessary relationship to Dogmatism (the philosophical appeal to dogmatically-derived premises) their argument remained rationally incomplete, and thus would have remained unconvincing even if the logical positivists had had a convincing positive alternative to offer in place of the eternalism on which many still relied in the early twentieth century. Eternalists could always offer an alternative framework of belief in which their “nonsensical” statements were both verifiable and meaningful. In a well-known broadcast debate with the Christian F.C. Copleston, Ayer invented a “drogulus”, which was an imperceptible thing with no physical effects, and argued that it must be meaningless. But Copleston was quite prepared to admit its meaningfulness on the grounds that he could form the idea of such a thing, even if he had no particular reason for believing in it (although he did have reason to believe in other metaphysical entities)[36]. Perhaps the most interesting point about this episode was that (in pragmatic terms) the drogulus was probably more meaningful to Ayer than to Copleston.


A further reason for the relative ineffectiveness of logical positivist attacks on metaphysics was their association with materialism and behaviourism. However, the issues raised here are probably best discussed in relation to the ‘Private Language Argument’ of the later Wittgenstein, so I will postpone consideration of them until they can be discussed in relation to other aspects of Wittgenstein’s thought[37].  More immediately I shall turn to logical positivist treatments of ethics.


iv)                Positivist ethics


The understanding of ethics put forward by the logical positivists and their associates took various forms, but all of these were motivated by the application of the verification principle to ethical statements. According to any such application, ethical statements could be ruled either verifiable (and hence meaningful) or unverifiable (and hence at least prima facie meaningless). The logical positivists were united in accepting purely descriptive ethics, in the form of sociological accounts of ethical belief, as verifiable, and in accepting purely analytic accounts of ethical terms (which is what they understood their own “ethics” to be) as self-evident[38], but were divided over exactly how normative ethics was to be regarded. The chief division here lay between those who interpreted normative ethics in relative and descriptive terms which they believed gave meaning to absolute normative statements, and those, subsequently labelled “non-cognitivists”, who took ethical statements as referring to absolute ethical claims which could not be verified.


The first of these groups largely followed the assumptions of Hume and can by represented by Moritz Schlick. Schlick appeals to the “material” rather than the “formal” properties of good: “One of the worst errors of ethical thought lies in the belief that the concept of the moral good is completely exhausted by the statement of its purely formal property”[39], he writes. The material properties of good consist in their manifestations in human belief and behaviour. To the problem of relativism which this raises Schlick responds with only a slight modification of the kind of naturalist conviction one could find in Aquinas or Hume:

The modes of behaviour which we group together under the names reliability, helpfulness, sociability, are everywhere judged to be ‘good’, while, for example, thievery, murder, quarrelsomeness pass for ‘evil’ so unanimously that here the question of the common property can be answered with practically universal validity. If such characters are found for a large group of actions, then one can apply himself [sic] to the ‘exceptions’ and irregularities, that is, to those cases in which the same behaviour evokes divergent moral judgements in different times, among different peoples. Here one finds either that there is no different ground for the judgement from that in all ordinary cases, but that it is merely more remote, hidden, or applied under altered circumstances; or one must simply note the fact as indicating a new or ambiguous meaning of the word ‘good’.[40]


The appeal to hidden factors here is identical to that commonly used in the eternalist tradition to support the compatibility of universal claims with moral conflict in relative contexts; but it is the further appeal back to an ad hoc analysis which gives this a distinctively logical positivist style, adding a new element to its unfalsifiability in case the previous, more traditional, arguments should fail. Since any possible moral conflict could be subsumed under this description without giving us any idea of how to resolve it, it is clear that Schlick is here merely trying to give universal support to conventional ethics by offering a pseudo-scientific account of them which is actually almost entirely Dogmatic.


The more common interpretation of logical positivism in the realm of ethics, however, was non-cognitivist. The non-cognitivists shared the view that normative ethics were unverifiable, but diverged enormously in their interpretation of how that unverifiability should be regarded. At one extreme is the view of G.E. Moore and Bertrand Russell that goodness was an indefinable non-natural quality and that the ascription of the quality of “good” to a state of affairs can be objectively true or false, but that this can only be known through intuition[41]. Even if Moore and Russell are not regarded as logical positivists in the same strong sense as Ayer and the Vienna Circle, they shared many of the same assumptions, and their view of ethics, generally known as “intuitionism”, is quite compatible with logical positivism. At the other extreme was the view of Carnap, Ayer and C. L. Stevenson that became known as emotivism. According to this view, the unverifiable nature of prescriptive ethical statements makes them not meaningful statements at all, not even of an individual desire, since some kind of pseudo-absolute status is being claimed which is entirely incompatible with the verification principle. Instead, ethical statements should be understood as (not statements of, but) expressions of emotion equivalent to “Hurrah!” or “Boo!”[42].


These two views themselves represent a microcosm of the eternalist-nihilist debate within the sphere of logical positivism. Intuitionism represents a return to ideas of an intuitive conscience found in the eternalist tradition, but now asserted in the pseudo-scientific context of early analytic philosophy[43]. Emotivism, as MacIntyre observes, probably arose in reaction against the pretensions of intuitionism[44]. Yet the similarities between the two movements are perhaps more striking than the differences. Both depend on a strong assertion of the fact-value distinction (which Moore restated as a condemnation of “the naturalistic fallacy”), and hence are basically scientistic in their attempt to authoritatively define the facts about value. Both are strongly individualistic in their account of the meaning of ethical beliefs, both contrast that account of ethical beliefs with the socially-shared and public criteria used to determine cognitive knowledge, and both thus take an irrationalist view of ethics. The main difference seems to lie in the value ascribed to the ethical in each case: for intuitionism this is absolute, for emotivism entirely relative. This complete divergence of approaches illustrates the effect of completely separating ethics from the naturalist moorings to which Schlick firmly tied it: ethics has become anything at all, from the absolute to the trivially personal, because it has been excluded from the field of rational consideration and control.


MacIntyre’s account of the relationship between intuitionism and emotivism (in the Cambridge setting where they most famously developed) clearly shows the confusion which prevailed and the ways in which that confusion could be manipulated. In his account of how Moore and his supporters taught intuitionism he draws on the recollections of Keynes:

Keynes gives examples of central topics of discussion among Moore’s followers….:’If A was in love with B under a misapprehension as to B’s qualities, was this better or worse than A’s not being in love at all?’ How were such questions to be answered? By following Moore’s prescriptions in a precise fashion. Do you or do you not discern the presence or absence of a non-natural good in greater or lesser degree? And what if two observers disagree? Then, so the answer went, according to Keynes, either the two were focussing on different subject matters, without recognising this, or one had perceptions superior to the other. But, of course, as Keynes tells us, what was really happening was something quite other: ‘In practice, victory was with those who could speak with the greatest appearance of  clear, undoubting conviction and could best use the accents of infallibility’ and Keynes goes on to describe the effectiveness of Moore’s gasps of incredulity and head-shaking, of Strachey’s grim silences and of Lowes Dickinson’s shrugs.[45]


MacIntyre goes on to comment here on the “gap between the meaning and purport of what was being said”: in other words, the gap between philosophical statements and the psychological states they reflected. The appeal to intuition, as frequently in the eternalist tradition, seems to operate largely as a rationalisation for the assertion of social power. MacIntyre also brings out the respects in which it appealed to a particular group in educated English society represented by the ‘Bloomsbury Group’: a group whose values were individualistic and aesthetic, being part of an insulated class of beneficiaries of capitalism. For them the scientistic rather than eternalistic roots of intuitionism made it an ideal rationalisation of refined hedonism, now that they were in a position to enjoy the fruits of the constant alienation produced by Christianity and capitalism during the nineteenth century. This hedonism was supported by two other, logically independent, aspects of Moore’s doctrine: his utilitarianism (which aimed to maximise the intuitively-perceived good) and his view that the greatest goods we can imagine are ‘personal affections and aesthetic enjoyments’[46].


MacIntyre notes that some of the founders of emotivism, including C. L. Stevenson, were pupils of Moore. It is easy to imagine the ways in which Moore’s approach could also provoke negative reaction and how Stevenson and his contemporaries could maintain many of Moore’s other assumptions but reject the appeal to intuition from which he derived his authority. MacIntyre’s account of emotivism is linked to his brilliant portrayal of  “the nature of moral disagreement today”, in which he describes the widespread unconscious acceptance of a paradox in modern moral disagreement, whereby “we simultaneously and inconsistently treat moral argument as an exercise of our rational powers and as mere expressive assertion”[47]. This indicates the widespread acceptance of emotivist attitudes, according to which moral discussion naturally appeals to an absolute, but it is a nonsensical absolute which can never be agreed upon by those who disagree. Moral disagreement is thus interminable.


Emotivism thus represents a far more popular and important attitude to ethics than does intuitionism, but wherever there is an emotivist there may well be an intuitionist lurking in the background, claiming that her judgements are objective and that the disagreement can be ended by the recognition that her view is correct. The disagreement between intuitionist and emotivist admits no more resolution in its own terms than the larger one between eternalist and nihilist on which it depends. But it is the scientism of the fact-value distinction that makes the disagreement interminable in the modern context, and the discontinuity between absolute and relative in the whole history of dualism in the West which stands behind that application of the fact-value distinction and continues to support it.


In such an atmosphere of interminable disagreement individualism and hedonism can be easily justified under either intuitionist or emotivist values, and it makes little difference which are adopted in terms of their long-term effects. In this confused situation analytic philosophers have found two other ways forward, neither of which ultimately succeed in escaping nihilism but which deserve further consideration. One is the adaptation of scientism to include approaches from utilitarianism and Kantianism, and the other the development of theories of rationality independent of ethics. Both of these take place within the overall context of the development of liberal democracy, which will also require further attention in the remainder of this section. I shall tackle these in a largely thematic order which takes leave of strict chronology but is mainly concerned with developments in the second half of the twentieth century.


v)                  Hare


The work of R.M. Hare deserves consideration here as probably the foremost among the twentieth century analytic philosophers who have tried to reconcile the scientistic assumptions of the logical positivists with a more positive approach to ethics which adopts ideas from the eternalist tradition. As with Kant, Marx and Mill on the eternalist side, Hare’s approach makes it clear how thin the line between eternalism and scientism can be and how, with certain compromises on each side, they can appear compatible. The chief of these compromises is the agreement of a boundary between a domain of “facts” over which science and its description rules, and a domain of prescriptive “values”. If the agreement is adhered to and neither side attempts to poach on the wrong side of the divide, the basis of an unholy alliance is formed whereby dualism is perpetuated.


Hare puts forward an ethical theory which allows such a demarcation agreement to be maintained, but does so on scientistic grounds. These grounds consist in a strong distinction between descriptivity and prescriptivity, which is similar to that of the emotivists and intuitionists except that Hare thinks that a systematic logical account of the nature of  ethical thought can be given, making ethical prescriptions both meaningful and rationally accountable. But Hare also attempts to steer clear of the reef of naturalism by insisting on the uniquely prescriptive nature of moral language, explaining this in terms of linguistic analysis which provides a basis for a universalisability which (at least in some formal respects) resembles Kant’s, and which he also believes to yield conclusions entirely compatible with those of utilitarianism. Hare thus draws on a neo-Kantian account of why moral statements are prescriptive, but tries to improve upon it by showing how universalisable prescriptions can be applied using the utilitarian hypothesis that it is preferences which should be universalised.


So what prevents us from seeing Hare as a sophisticated Kantian in the eternalist tradition? His use of a “deontic logic” has its origins in Kant’s account of practical reason, and his claim that universalisability provides a basis for rational objectivity is largely Kantian. It is only Hare’s context which is different. Whilst Kant lived in an eternalist world and needed to justify his dual scheme for phenomenal scientific description and noumenal ethical prescription to an audience still largely eternalist, he did so mainly in epistemological and metaphysical terms which required the hypothesis of the synthetic a priori; whereas Hare, living in a world which is increasingly nihilist, does so primarily in the linguistic terms of the analytic a priori. Both want an accommodation between eternalism and nihilism, but Kant, true to his time, does this in terms of an appropriation of the scientistic case by eternalism, whilst Hare attempts to appropriate the eternalist case to the scientistic framework set up by the logical positivists.


Hare’s account makes use of three features that he claims justifiable moral language to have: prescriptivity, universalisability and “overridingness”. I shall examine each of these features to bring out the scientistic and dualistic aspects of Hare’s approach.


Hare’s understanding of prescriptivity is based, as he admits, on a conceptual analysis of moral terms as they are generally used, thus confining their prescriptive power to the conceptual scheme within which they are used. “I am not suggesting that we are tied to using the words in the way that we do, or to having the conceptual scheme that we have” he writes, “But if we were to alter the meanings of our words, we should be altering the questions we were asking, and perhaps answering, in terms of them”[48]. He goes on to suggest that his criterion for the justifiability for asking different questions should be that “I should require to be satisfied, not merely that the new ones are important, but that the old ones are unimportant”[49]. This goes far enough towards the recognition that moral logics are culturally restricted to show that Hare shares the general scientistic perception of ethics as essentially relative. It incidentally sets a challenge which I hope to meet in this book as a whole. But Hare goes on later to show the restricted way in which he actually understands any possible advances on his own analysis:

If anybody either does not believe that I have given a true account of the moral concepts as we have them, or thinks that investigations of the logic of words in ordinary language are unhelpful, either because they yield no determinate results, or because, even if they did, these would have no force in moral arguments, it would be open to him to proceed in a different way. He could set out the logic of an artificial language, identical in its properties with the logic which I claim our ordinary words have; then he could show that anyone reasoning in such a language would have to think in the same way in which I say our existing concepts constrain us to think; and then he could, using the same arguments that I shall use, show that there would be an advantage in our adopting such a language.[50]


Hare here shows himself to be caught in the analytic duality between ordinary and artificial language, since the only alternative to his ordinary-language based analysis that he can imagine is an artificial-language based one. The constraints of language show for him the bounds of the meaning of ethics, and any language which depicted ethics better would merely represent the same (essentially conventional) ethics but do so more faithfully. The idea that ethical language might change because it might describe a different function, or that existing ethics might be justified (or condemned) in functional terms does not seem to occur to him here[51].


Hare contrasts the prescriptivity based on people’s use of moral terms with the descriptivity used in purely factual and supposedly neutral statements. Hare rejects the arguments of some “descriptivist” philosophers that the very way in which we understand and describe a situation contains an element of evaluation inseparable from the description, maintaining that even what he calls “secondarily evaluative words” such as “cruel”, “lazy” or “rude” can be used in a purely descriptive way which makes the prescriptive element of their meaning distinguishable[52]. However, the examples Hare offers, such as the allegedly purely descriptive use of “cruel” to describe a practical joke by someone who didn’t disapprove of it, merely show that the evaluative role played by the use of the word by different people is a different one, not that it is not evaluative. The word “cruel” could be used in a way which showed approval, or for that matter an attempt neutrality, but even neutrality (which may be impossible in a pure form anyway) is an evaluative stance understood only in relation to the possibility of approval or disapproval. Hare’s attempt to bolster his claims of the existence of pure “prescriptivity” by contrasting it with “descriptivity” fails because he is clearly imposing a dichotomy on an incremental scale of degrees of affect which accompany language, to which the choice of language itself bears only a contingent relationship.


Hare’s treatment of descriptive language here is related to his account of objectivity. Here he claims that factual objectivity, such as “being in principle publicly observable” should be clearly distinguished from the sort of objectivity which he believes ethical prescriptions can achieve, which is achieved purely by virtue of their universalisability[53]. Hare sees universalisability as a purely conceptual property of value judgements, namely that they “entail identical judgements about all cases identical in their universal properties”[54], it following from this that

…if I now say that I ought to do a certain thing to a certain person, I am committed to the view that the very same thing ought to be done to me, were I in exactly his situation, including having the same personal characteristics and in particular the same motivational states.[55]


Hare makes it clear that this applies only to a conceptually identical situation and not to the unlikely scenario of an actually identical situation, but its coherence even in conceptual terms depends on assumptions about personal identity. What would it mean for me to be in exactly someone else’s situation, other than being them? Logically it is contradictory for me to be someone else, since I can either be me or someone else, but not both. Even if “I” am defined purely in terms of my actual identifications this is the case, since there will always be “someone else” that I don’t identify with. If it is thus, as Hare claims, a purely conceptual procedure it works on the basis of an incoherent hypothesis. It would be more coherent to suggest that this was a useful imaginative procedure: to imagine myself in someone else’s shoes to develop empathy for them (in which case I would be me imagining I was someone else) but Hare’s case adheres to a purely rational basis[56].


In fact it appears that the convincingness of Hare’s theory of universalisability depends on what he claims to be purely rational features (in order to claim rational objectivity) unconsciously accreting psychological elements for the reader. It would be possible for me to reason about what it would be like to be someone else in two different ways: I could reason in a way which is alienated from any actual identification with the other person, or in a way which is integrated with such an identification to some degree. If we understand Hare as referring to an alienated reasoning process, then it obviously functions as pure rationalisation: a person could go through this process and know exactly what was the right course of action and yet have no motive to perform that course of action. If, on the other hand, we understand the reasoning to be in necessary combination with any degree of imaginative empathy, it is this empathy which provides the necessary element without which the combination of reason and empathy would have no claims to objectivity. Hare’s argument, however, depends on the reverse assumption that it is the reasoning process which is necessary.


This becomes even more obvious when we consider Hare’s account of how universalisability is applied through utilitarian calculation. Universalisability is only given substantial content when it is related to the facts of the case under consideration[57]. The relevant facts are selected through a process of hypothesis until we find those which fit the pattern of universalisable reasoning about the case. Hare offers preference-utilitarianism as such a hypothesis: the facts that are relevant to a particular moral judgement are those of the preferences of those involved (known indirectly through hypothesis) and of the effects of possible actions on those preferences[58]. I have already discussed the dogmatic empiricism involved in the consequentialist approach offered here[59], and Hare relies just as much on this acceptance of descriptive “facts”, without allowance for our degree of doubt and ignorance, as his utilitarian predecessors. His proposed calculations involve facts about preferences, not beliefs understood as influenced by preferences, attempting to appropriate all such considerations to an entirely rational framework.


Hare’s account of the relationship between reason and emotion, however, cannot be fully explained without reference to his third criterion of justifiable moral statements, that is their “overridingness”. This means that they must override all other prescriptions with regard to the same case. Many of these other prescriptions will be what Hare calls “intuitive” moral prescriptions, which Hare stresses have a prima facie validity as moral principles, but which are not necessarily adequate to every case and so must be overridden by purely rational, or “critical” moral prescriptions. Intuitive moral prescriptions will also be “moral” in the sense that they have been selected by critical morality in the light of their general validity as moral rules[60]. For Hare the distinction between intuitive and critical morality provides the key to resolving many conceptual problems, but this is through the employment of a very old eternalist strategy: when inconsistencies occur at the relative level, this strategy is to appeal to the absolute and to appropriate that appeal to your relative position.


Hare does this, perhaps, more disingenuously than usual, because he introduces two characters which represent absolute and relative positions: the archangel, who has knowledge of all facts and employs only critical thinking, and the prole, who is incapable of critical thinking and thus has to rely entirely on intuitive versions of morality. He then goes on to ask

‘When ought we to think like archangels and when like proles?’ ….the answer…depends on how much each one of us, on some particular occasion or in general, resembles one or other of these two characters. There is no philosophical answer to the question; it depends on what powers of thought and character each of us, for the time being, thinks he possesses. We have to know ourselves in order to tell how much we can trust ourselves to play the archangel without ending up in the wrong Miltonic camp as fallen archangels.[61]


There is insight here in the sense that Hare recognises the ways in which the usefulness of reason depends on our nature in other respects, which will determine its premises. But there is also cause for suspicion that we could easily make the wrong judgement. Does he really think that proles will always believe themselves to be proles and not archangels? In one sense, indeed, we are all proles because we have no absolute grounds for belief in the facts on the basis of which we judge, or indeed for believing that we have the correct method of judgement. Philosophical grounds then do exist for believing that there are only proles and no archangels, in the form of ordinary Sceptical arguments. The fact that Hare gives no psychological, only rational, criteria for being an archangel also provide a terrible vision of archangelhood, for a supposed archangel is not just really a deluded prole, but possibly alienated at that. An archangel may believe he has absolute grounds for moral judgement when he has none, and only apply those moral judgements in any case when it fits his desires to do so. Such a figure might well be a monster who combines absolute certainty with the power of knowledge, but without psychological integration: something like the despotic Yahweh of the Old Testament, perhaps, or a cleverer Hitler less prone to strategic errors.


The curious feature of Hare’s view is that he recognises the fact that human beings are not archangels, and yet continues to appeal to an archangelic perspective of rationality. He claims that the solution for our general lack of rationality lies in intuitive ethics, yet we also need to be able to appeal to critical ethics despite our inability to practise it. Given our lack of absolute ethical foundations, the appeal to rationality can produce only an limited judgement which takes into account conditions within the limited sphere of knowledge we are able to deal with, but Hare appears to regard this ethical coherentism as adequate grounds for objectivity in ethics.


Perhaps this limitation seems more acceptable to him because of  a general trivialisation in his habitual conception of ethical problems, to judge from the examples he gives. For example, he illustrates his model of universalisation by reference to the example of one person needing to move another’s bicycle because they want to park their car in that place, although the bicycle-owner has a mild aversion to their bicycle being moved. He argues that universalisation requires that the person wanting to park their car should treat the aversion of the other person to their bicycle being moved just as he would treat a similar aversion of his own, that is, as overruled by the stronger desire to avoid inconvenience by parking his car in that place[62]. This is trivialised not because the situation doesn't require the exercise of an ethical sense (it plainly does), but because Hare seems to think that a critical judgement of such a situation (not a quick judgement based on prima facie rules) can be exercised only by considering these factors. He ignores, for example, the difficulties in knowing what the bicycle-owner actually feels, the unconscious psychological effects on the bicycle-owner (whose temperament is not specified), the effect on other observers who might see the bicycle being moved, the contribution that might be made to power relationships whereby car drivers maintain undeserved hegemony over cyclists, the possibility of the judgement being influenced by the psychological states or dispositions of the judge, and the psychological effects that making the judgement might have on him subsequently. The spreading wave of possible effects even in the case of  such an apparently trivial judgement is infinite, and to call a judgement “critical” which takes into account so few of them amounts to a trivialisation, a narrowed appreciation of our degree of responsibility for events in the world[63]. In effect, the way in which Hare limits the judgement (without any allowance for the unknown) puts it into his category of “intuitive” judgements.


Hare’s ethical theory thus in general has a curious status. Whilst insisting on “prescriptivism” it is “descriptivist” in the sense I have applied so far, since his appeal to “deontic logic is justified by the empirical appeal to the way people use language[64]. The way in which he attempts to give authority to his theory of rational ethical objectivity by appeal to analysis is at the same time reminiscent of logical positivism, and in many ways involves the mere extension of logical positivism to the ethical sphere. Hare’s theory is ineffective as a challenge to relativism because, as he states himself, it has no purchase on the “amoralist” who does not accept his premise that there are affirmative universal prescriptions[65]. This is hardly  surprising given that Hare begins from relative and descriptive premises: his “prescriptions” thus lack normative power. But his account of ethics is also flawed even as a description of what people believe by his adherence to rationalism and ignoring of the psychological.


Hare also exhibits some of the other features of scientism that I have previously identified. He deals with the problem of freewill by recourse to a form of compatibilism similar to Hume’s[66]. His individualism in the broad sense of conventionalism I have already commented on, though it also appears in the form of an appeal to the interests of the individual person in relation to his account of individual rationality, which also involves an appeal to cosmic justice. I shall be considering this account, in relation to the accounts of some other modern analytic philosophers, in the next subsection.


vi)                Theories of rationality


The development of theories of practical rationality is not a feature new in the emergence of scientism: rather it is an adaptation of an approach existing in the eternalist tradition to serve a slightly different function. In the eternalist tradition, theories of rationality were the mechanism whereby the absolute nature of ethics could be shown. Although it was assumed that individual rationality in the form of prudent self-interest could be appealed to to support ethics through the structures of cosmic justice, that rationality could not have existed without the wider ethical purpose which it served. There was thus no question of prudent self-interest alone being the theoretical foundation of ethics in eternalism, even if the way in which the alleged foundations were actually a rationalisation of such self-interest could be revealed through notions of cosmic justice. Thrasymachus in Plato’s Republic can thus put forward a relativist view, but he can offer only individualism, not a consistent or universal rationality, to support it[67]. Similarly for Kant, practical rationality and ethics are indistinguishable, but whichever label is applied the appeal is to universalisability and the avoidance of a response based on relative desires. In scientism, however, the adoption of scientific knowledge as the prime value meant that the rationality of science itself became normative and ethics had to be justified in terms of that rationality. Whilst scientistic rationality, like that of eternalism, had pretensions to universality, its practical nature as a way of guiding action could only be expressed in relation to the individual because the Humean distinction between facts and values was taken to rule out any universal status for ethics.


In this way analytic philosophy gave birth to the idea of an independent rationality of the individual as a kind of substitute for universal ethics. In the absence of a universal prescription it was reasoned that criteria of consistency could still be applied to the actions of the individual. Such criteria could then often provide the basis for an understanding of ethics: either because ethics amounted to nothing more and could be explained in entirely self-interested terms, or because self-interest was claimed to provide a motive for adhering to an ethics which was defined in other terms. Such an approach could only be based on a strong assumption of individualism, though fortunately the analytic approach when trained on this individualism has also begun to reveal some of its inadequacies.


I shall briefly discuss three modern accounts of practical rationality by analytic philosophers. The first, of R.M. Hare, shows the kind of approach to rationality which is consistent with the account of ethics discussed in the last subsection. The other two, by Derek Parfit and Simon Blackburn, are both sophisticated enough to reveal many of the weaknesses of individualism: however, due to the way in which they are influenced by the general scientistic assumptions of analytic philosophy neither of them succeeds in entirely disentangling himself from that individualism[68]. In general my aim will be to show that there cannot be a convincing account of practical rationality distinct from that of ethics, and that the grounds of ethics, which are in any case not purely rational in nature, must be prior.


I have already explained in the previous subsection how Hare’s understanding of moral objectivity depends on an account of logical rationality. At this level Hare’s account of ethics is conventionalist rather than individualist in a sense relating to the individual person. However, Hare evidently also feels that his account of ethics leaves a gap of motivation. In part this consists in the recognition of some of the difficulties I raised in the last subsection. As Hare writes

…it is possible for somebody even to use the moral words in the senses, and with the logical properties, that we have claimed they have, and still not make, using them, any moral judgements except ones of indifference…. Even if someone does make moral judgements that are not judgements of indifference, he may still not act on them; so we have, by providing a system of reasoning, done little to improve people’s morals. And if, as prescriptivists are entitled to do, we reply that on our view there can be no gap between a prescriptive moral judgement and the disposition to act on it (for if a person were not disposed to act on his moral judgements, he would not be treating them as prescriptive), still it may be objected that there has to be a gap somewhere. Logic cannot take us all the way from beliefs in non-moral facts to dispositions to action.[69]


He still has to explain why anyone ought to follow universalisable moral prescriptions if they do not want to, and thus he offers what he calls a “non-moral” explanation of how he believes prudent self-interest should support prescriptive ethics. In this explanation he eschews the use of the word “rational” to represent consistent self-interest, since he has already appropriated that term for ethics, and tries to avoid appealing to cosmic justice in the traditional eternalist style, but nevertheless argues that it would be in a child's best interests to bring it up to be moral rather than to be egoistic[70]. Hare’s arguments in support of this are empirical ones. He claims that “to be a successful immoral egoist…requires capacities far beyond the reach of most human beings"” that "crime does not in general pay” and that the easiest way to appear to be virtuous so far as is needful for self-interest is to actually be so[71].


All of these claims may be relatively true in restricted empirical circumstances (ones which perhaps require a stronger degree of social cohesion than is sometimes present in the world), but what is interesting about them is the rather restricted range of examples of immorality they focus on, all of which fit the model of possibly self-interested immorality as a confrontation of the individual with society, where egoism is individualism and moral truth is social convention. There is no mention of collective immoralities such as social exclusion, economic oppression or environmental degradation, even where these are obviously motivated by self-interest exercised collectively. It seems obvious that Hare would not be able to sustain his claims through plausible empirical argument in relation to such examples, for whilst it may be true that a person who wishes to be successful in social and economic terms should have a conscience which leads them to avoid crime or deceit, too much concern about matters of  broader justice is likely to impede them because they will be constantly questioning and undermining the terms of social and economic distinction which provide such prosperity.


This selectivity of examples reveals weaknesses not just in Hare’s understanding of ethics, the conventionalism of which I have already discussed, but also in his understanding of what he calls prudence (a term compatible with what some other analytic philosophers call rationality or rational self-interest). Hare relies on a conventional understanding of self-interest which assumes that the desires of the individual have both a certain consistency, over time and at a given time, and a distinctness from those of other people, which enable us to identify self-interest as having an actual normative power over behaviour. Hare then attempts to appropriate this supposed normative power for socially defined ethics, which is not really such a large step as it appears if we strip it of its pretensions to be anything other than an association between two conventional constructions where value is believed to focus. It takes relatively little analysis of these assumptions about self-interest to show that far too much has been built on them.


Hare runs into similar problems in his arguments about supererogation. In brief, Hare claims that it is not morally desirable for us all to be saints, because we do not all have the capacity to be so, and those with differing capacities should follow differing prima facie moral principles[72]. In the idea of “capacity” however, lurks another conventional assumption about the self, giving it features which are fixed but not necessarily easily identifiable. If morality consists only in conventional expectations, then it will be possible (bizarrely) to over-fulfil them and thus become excessively moral simply through being unconventional. Supererogation will be then be possible because morality will consist in a set of fixed and discontinuous expectations without any incrementality, with excessive morality becoming effectively a sort of immorality. Hare tries to reconcile this with the moral defensibility of having some saints by effectively making it a conventional expectation that some will be saints and others not. However, since anyone can claim not to have the capacity to be a saint, this effectively maintains the conventional expectation that nobody need be challenged beyond the point of conformity, by shaping theory to fit that expectation. Thus supererogation, an invention of the medieval Catholic church, maintains a curious afterlife in the rationalisations of modern analytic philosophy.


The basic problem here is the assumption that the ego is identical with the psyche, which leads to the confusion of the identifications of the ego with the interests of the individual[73]. The idea of the interests of the individual can then be manipulated in support of conventional ethics in a way in which the notion of the ego cannot. Parfit and Blackburn go on to make more sophisticated analyses of the self and its rationality which deal with some of the symptoms of this problematic assumption, but do not clearly identify it[74].


Parfit’s account of the rationality of the self appears in his discussion of the relationship between three theoretical positions, all of which represent possible understandings of rational self-interest: P (present-aim theory), S (a theory of self-interest) and CP (a critical adaptation of P). Parfit criticises the assumption that a deliberative version of P is always the same as S, a view he describes as “psychological egoism”: whatever understanding of self-interest is applied, Parfit argues convincingly that this is unlikely to be coincident with our present desires even after deliberation, and that even if the two are coincident in content the justification will be different[75].


Parfit also offers a strategy for showing the incoherence of self-interest theories by showing that an advocate of S must use inconsistent types of argument against P on the one hand, and against morality on the other. Against P, S must argue for rational neutrality between contradictory desires which appear at different times, whilst against morality S must argue for bias in favour of oneself over others. Since inconsistent types of rational consistency are appealed to in each case, Parfit argues that we should reject S as irrational.


However, Parfit attacks one kind of theory of rational self-interest only to support another, which he calls CP, a critical version of present-aim theory. He defines CP as follows:


CP: Some desires are intrinsically irrational. And a set of desires may be irrational even if the desires in this set are not irrational….Suppose that I know the facts and am thinking clearly. If my set of desires is not irrational, what I have most reason to do is what would best fulfil those of my present desires that are not irrational. This claim applies to anyone at any time.[76]


But what makes a desire irrational? Parfit gives a number of examples[77]: for example, he claims that a person who was altruistic towards people who lived less than exactly a mile away, but not towards people who lived more than a mile away, would be irrational. He then summarises the criteria.


In these cases the concern is not less because of some intrinsic difference in the object of concern. The concern is less because of some property which is purely positional, and which draws an arbitrary line. These are the patterns of concern which are, in the clearest way, irrational.[78]


But the distinction here between properties that are purely positional and those that are “intrinsic” simply repeats the scientistic duality between descriptive and non-descriptive statements. We can only understand properties as intrinsic if we can describe them accurately and know what objects they refer to with certainty: Parfit’s distinction thus only makes sense if one assumes realism, representationalism and essentialism. Certainly the “within-a-mile-altruist’s” criteria for the bestowal of altruism are arbitrary, but so are any criteria we use to define objects and their properties. The "within-a-mile-altruist” is thus no more nor less rational than any other person who uses non-universal criteria to limit their altruism: to only those of their own group, for example.


Parfit’s appeal to CP as a definition of rationality depends entirely on this scientistic dualism. He claims that moral theorists should accept CP, because moral desires could be included in the rational ones which it claims should be prioritised[79]. But CP offers no solution to the problem of relativism, as it offers no new reason to see a relative view as irrational. Parfit has made some progress in demolishing the long-standing attachment to S in Western philosophy, but is unable to offer anything positive in its place due to his attachment to the idea of there being a theory of rationality distinct from ethics. His criticism of S also proves of limited value because of this attachment: because S is only an abstract idea of self-interest and not an actual egoistic identification, to show it to be inconsistent does not provide any reason to overcome egoistic identification. Rather, in contrasting S with CP, Parfit merely advocates the replacement of one egoistic theory with another, maintaining the beliefs which support the division between the ego and other parts of the psyche through the conviction that there is a wholly rational answer to questions of value which allows the ego to re-appropriate what has been cast into doubt.


Parfit’s reliance on essentialism about the objects of rational distinction is all the more surprising given his nominalist treatment of the question of personal identity. Parfit applies a sophisticated Humean argument which attempts to show that we ought to understand ourselves in reductionist terms[80], but repeats Hume’s mistake of seeking only the “facts” about the self, rather than understanding it in dynamic terms as a process of identification. Parfit’s claim that the Buddha would have agreed with his reductionism is thus just as misleading as that of those who make similar claims about Hume’s[81]. Parfit’s reductionism, like Hume’s, appears to depend on a contrast between the presence of “facts” about other objects and the absence of “facts” about the self, not, as in a non-dualist approach, on a recognition of uncertainty about facts extended consistently to our self-experience.


Blackburn’s account of the self and its rationality is also broadly Humean, but offers a quite different emphasis. He offers a very useful analogy illustrating the difference between the deliberative operation of reason on a Kantian and on a Humean account: that of a ship.

For Hume…the ship is worked by a crew, each representing a passion or inclination or sentiment, and where the ship goes is determined by the resolution of conflicting pressures among the crew. After one voice has prevailed, various things may happen to the losers: they may be thrown overboard and lost altogether, or more likely they may remain silenced just for the occasion, or they may remain sullen and mutinous, or they may continue to have at least some effect on the ship’s course….

For Kant, so the contrast goes, there is indeed the Humean crew. But standing above them, on the quarter-deck, there is another voice – a voice with ultimate authority and ultimate power. This is the Captain, the will, yourself as an embodiment of pure practical reason, detached from all desire. The Captain himself is free. But he always stands ready to stop things going wrong with the crew’s handling of the boat.[82]


This analogy throws an important aspect of the distinction between eternalism and nihilism into stark contrast, showing how Kant’s understanding of the will, whatever its relationship to disavowals of Cartesian metaphysics[83], bears a strong structural similarity not just to Descartes’ understanding of the relationship between soul and body, but to the Christian understanding of the relationship between God and the universe and Plato’s understanding of the relationship between the rational soul and its other parts.


Blackburn offers an impressive attack on this Kantian assumption about the will, arguing cogently that deliberation is not the activity of an independent will making a choice between desires, but rather involves a process  motivated by a particular desire in which the attention moves between external objects which it sees as the possible means to its fulfilment. If desires are understood as the objects rather than the subjects of concern, Blackburn argues, then they will appear arbitrary - a threatening mutinous crew before whom the subject is passive, and it is this sort of fear which has motivated the Kantian mistake[84].


Blackburn suggests that there are two major mistakes that have been made about deliberation, mistakes which have not only been made by Kantians but even by Humeans like Parfit. The first mistake is the objectifying of our desires, made by Parfit in his theory that there can be a critical perspective on present aims independent of those aims[85]. The second is “confusing the absence of an experience of causation with the experience of the absence of causation”, which underlies the assumption of freewill.

Since we do not and cannot be aware of all the forces that mould our agency, we think we have knowledge of the absence of such forces. But….in reality we need to recognise the inevitable existence, not of a perspective of free or rational agency, but of an absence that can easily be mistaken for it. We have here the necessary existence of a ‘blindspot’, or aspect of ourselves that cannot be seen, by ourselves, at the time we deliberate.[86]


This ‘blindspot’, with its admission of our ignorance about freewill without a corresponding assertion of determinism, comes close to a non-dualist aporesis on the topic.


For all the strength and insight of his arguments against Kantianism, however, Blackburn fails to transcend the scientism of his analytic background. Blackburn puts forward a theory of rationality which also constitutes ethics, and follows Hume’s understanding of rationality of judgement as clarifying the relationship of means to ends through applying knowledge in the service of our desire. Blackburn stresses, though, that Hume’s view is not the whole story: knowledge of matters of fact is relevant in other ways, such as appreciating the consequences of our ends being achieved, and is not always sufficient, as we can have too many facts and not be able to apply them due to lack of imagination, maturity, sympathy, or culture[87]. Nevertheless, for Blackburn there is no moral method beyond the rational one, and rationality consists largely in coherence. Where it goes beyond coherence, he has no systematic account to offer of why or how it does so[88].


Blackburn thus has no convincing response to relativism. His argument against it is based partly on pointing out the false dichotomy between absolute truth and the complete abandonment of moral truth made by postmodern relativists such as Rorty[89], but it also asserts that objectivity must emerge from our own desires without specifying any ways in which those desires themselves need to be shaped. The context of meaning and belief for ethical statements is a “quasi-realist” one, in which our moral concerns themselves are projected in coherent propositional structures, but Blackburn still wants to distinguish between this kind of functionality and that used in describing facts[90], leaving scientism intact with its relativistic implications. It seems as though Blackburn is content to define ethics as it is conventionally understood by a person in his social setting, noting the limitations of doing so, without providing any prescriptive challenge. In this respect he is typical of the nihilist tradition in that he simply does not understand the eternalist (in his context of discussion, Kantian) motivation for wanting independent norms. As frequently in nihilist argument, the demolition of the eternalist argument also involves an incidental loss of the scaffolding of universal moral meaning within which something more convincing could be erected.


Blackburn thus illustrates the tendency in analytic philosophy for scientism to live on in more and more sophisticated expositions which all nevertheless begin from the same basic premises. His degree of sophistication and mature humanity tends to attract isolated psychological insights as a magnet attracts iron filings, but the basic scientistic assumption that no system of objective ethics can be constructed without recourse to metaphysical claims restrains him from moulding those insights together into a stronger account from which any positive ethical guidance could be derived. Blackburn has gone a long way beyond Hare’s assumptions about the anti-social rationality of the self and even Parfit’s realisation that “self-interest” is itself a theoretical construction, to a position where desires are recognised as variously objective and subjective, but they are still scattered and unsystematisable, there being no account available of where the ship should be going even when there seems to be consistency in its navigation and fair knowledge about the seas through which it is passing.


I shall try to show in Part 2 that such a much more consistent account is still possible, and that scientistic pessimism about its possibility is merely the product of scientistic assumptions. Once the core assumption that science is fundamentally different from ethics is removed, and due allowance is made for our degree of ignorance, there is no reason for ethics not to be as precise and systematic, within the limitations imposed by its sphere of investigation, as science. But in order to reach such a systematic account, the psychological nature of the investigation, and the limitations of rationality as a model, need to be fully appreciated.


vii)              Liberal neutrality


Before leaving the topic of the effects of the core scientistic assumptions of analytic philosophy, and going on to the changes in those core assumptions wrought by Wittgenstein, I want to consider the ways in which those core assumptions can be correlated to broader social and political attitudes in the twentieth century. That they can be so correlated is central to my broader thesis about nihilism and its failure as an ethical approach.


I have already discussed MacIntyre’s important account of the emotivist assumptions which characterise modern moral debate, and linked this general prevalence of emotivism with the scientism represented philosophically by logical positivism[91]. The inconsistent assumptions of moral absoluteness and moral relativity accompanying these emotivist assumptions can also be found in Hare’s approach and its unholy alliance of prescriptive universality and descriptive scientism, as well as in the idea of individual rationality. All these features reflect general implicit assumptions about ethics in Western society in the twentieth century: that ethical questions admit of no final resolution, but are a matter of individual taste, interest or conscience. These assumptions about ethics have inevitably had political implications.


The natural political correlate for relativism and individualism may at first sight appear to be anarchism, but that anarchism is relatively rare in Western society is an indication both of the prevalence of scientistic justifications for individualism (the pursuit of knowledge requires an organised society within which individualism can operate) and its dependence on broader group values (as actual individual desires take the form of identifications either with a group or common to a group). Instead the main political correlate of individualism appears to be liberalism. Liberalism provides a set of political values that are compatible with widespread assumptions about the nature of ethics, and supports a philosophy of government which allows maximal freedom to the individual. But there are a variety of reasons for being liberal: I have already discussed the eternalist liberalism of Kant and of Mill[92] and the way in which a generally utilitarian justification and method are typical of modern Western democracy. However, there are also scientistic liberal attitudes which see the governmental support of individual autonomy as justified, not by an absolute ethics, but by the supreme value of knowledge. From such a standpoint individualism is instrumental rather than foundational in the way it is for existentialists.


The strongest characteristic of this kind of scientistic liberalism is its appeal to an ideal of neutrality which enables impartial and objective judgement. This neutrality is to be exercised by agents of the state in dealing impartially with citizens, leading to impartial judgements based on facts. It requires the fact-value distinction to be applied consistently and thus relies on scientistic accounts of ethics, but at the same time requires a higher level political justification for the values exercised by the state. How can individualists justify the values of the state which protects their rights and freedoms? Either through a recourse to eternalism at the political level only, resulting in a strong split between the values applied to public and private morality, or through a disingenuous separation of “politics” from “ethics” (or “the right” from “the good”) which somehow takes politics (like science) not to be a matter of ethical judgement at all, but the application of a principle of neutral justice in which differing values are treated equally by being reduced to facts[93]. The myth of political neutrality is thus interdependent with that of scientific neutrality.


Aspects of this approach are found in Mill’s argument that liberty of thought and discussion serves the utilitarian goal of advancing human knowledge and thus should be encouraged by governments, though here they still exist under an eternalist banner[94]. A more modern philosophical account of such a stance of neutrality is that of Ronald Dworkin, who argues from grounds of moral relativism not just to the importance of individual rights, but to the ultimate reducibility of all plausible political theories to a concern with equality which is based on this fact. Disagreements in political theory are thus for Dworkin merely disputes about the facts as to what measures will bring about equality, which is the only default value remaining given the fact of moral relativism[95]. I am not here concerned with arguing with the philosophical expressions of this political philosophy, however, since I believe them to be mistaken due to their prior assumptions about ethics in ways I have already specified. Not all scientistic liberals follow the same line as Dworkin, with others, such as Popper[96] or more recently Blackburn[97], maintaining that the fact-value distinction implies that facts are not enough for the basis of government and that liberal values should be adopted together with a recognition that they are values: perhaps this approach provides some resources for improving liberalism along the lines suggested by Popper, in which governments take an experimental and provisional approach (“piecemeal technology”) to formulating policy[98]. But while the fact-value distinction helps to define the role of government, its application will continue to impose an unnecessary dualism, as I shall now attempt to argue by charting some of the wider effects of scientistic assumptions.


In much of the Western world, the fact-value distinction now stands in the place of the now widely discredited idea of moral objectivity as the injunction under which agents of the state operate. I distinguish this aspect of government (at which policies are applied) from that at which policies are formulated, developed and justified, where different forms of eternalist justification (particularly utilitarian) are often invoked.  It appears in the operation of the police and the judiciary, the civil service, the armed forces, social work, education, public service broadcasting, the subsidising of cultural activities and provision of museums and libraries, the provision of health services and the regulation and support of economic activity. In all of these areas of public employment a tension is evident which is at its worst seriously problematic: this is that the employee of the state who has personal values which contrast to any extent with an institutional framework of values constantly has to present the appearance of not having them. In psychological terms, this means that each employee has to alienate a large section of their own desires under the compulsion enforced by the ego to conform to the expectations of their role as agent of the state and maintain the economic position this provides them with.


I need only give a few examples of the kind of result that can issue from this alienation, which will be familiar to some extent to every public sector employee (and some private sector employees too). The public neutrality of the police can conflict with an actual racism or class-prejudice on the part of individual officers. Civil servants fail to reveal corruption despite its immediate conflict with their personal values, or do so and get prosecuted. The armed forces and the police in Northern Ireland fail to gain the confidence of the Catholic sector of the population and are seen by them as an occupying force, despite their supposedly neutral role in keeping the peace. Teachers communicate mixed messages between their actual religious or political convictions and their official neutrality, creating confusion amongst their pupils. The solution usually advanced for this problem is legislation, bureaucratisation, or economic pressure which attempts to keep public employees under closer and closer control, but this only increases the degree of alienation.


It may appear that such tension and alienation are an inevitable side-effect of the operation of the state in these areas of public life, and are justified in the same way as liberal government in general is justified, as a lesser evil than the effects of unchecked conflicts of interest and unmitigated inequalities which would result without it. But this argument can only work on the assumption that inescapable subjectivity is an essential feature of individuals and that a conflict between such subjectivity and the “neutrality” of government is inevitable. If the argument I put forward in this book as a whole is correct and moral objectivity is a property of individual psychological dispositions, then the state should be able to improve the effectiveness of its employees by both recognising and encouraging their moral objectivity. If moral objectivity is understood as an incremental matter, then the degree of responsibility and autonomy given to an employee should reflect, not only practical competence and experience, but primarily that degree of objectivity, every advance being rewarded by greater trust and thus reducing the division between the personal desires of an employee and the requirements of her public role. The training of an employee should also reflect the importance of developing the moral objectivity that is required to perform the task with integrity, as much as the other practical demands of the task.


This kind of recognition of the moral objectivity of the employee, however, depends in turn upon a re-assessment of the nature and justification of government in relation to the employee. If government maintains a myth of its own neutrality, then the subjectivity of employees is bound to contrast unfavourably with that neutrality, but if government acknowledges its own relative partiality and ignorance in relation to employees, there no longer need be a duality between the impersonality of one and the personality of the other. Since the policies of government, and the actual partiality and ignorance of government, are those of politicians, this requires that politicians should also cultivate moral objectivity as a genuine replacement for a false idea of neutrality. I shall be saying more in chapter eight about exactly how this could be achieved (without recourse to the kind of radical eternalism found in Plato’s Republic or Maoist China)[99], with appeal to further arguments in the intervening chapters about the nature, justification and cultivation of moral objectivity.


Similar arguments apply not just to the state and its agents but to the recipients of the services offered by the “neutral” state. To receive a service of the state, whether this be the support of social workers, welfare benefits, or education, is also often an alienating experience as the state imposes its own impersonally-justified crude prescriptions on the desires of individuals with very specific backgrounds and requirements, but with the recipients forced to accept the offered service because of their fear of legal sanctions or economic consequences. This is particularly evident in school education, where scientistic expectations of the priority of knowledge dominate a curriculum geared towards the gaining of knowledge and practical skills, together with a “neutral” and descriptive understanding of values, rather than the cultivation of moral objectivity. Where religious education takes place in state schools, as it does in the UK[100], it reflects the unholy alliance with eternalism and offers a syllabus of incompatible revelations: an approach which is only likely to perpetuate relativism, as I have argued in more detail elsewhere[101].  A much greater stress on the cultivation of moral objectivity in school (and also in further, higher and adult) education is again the non-dualist solution that I would propose to the alienation that occurs between the “neutral” state and the individual in society at large.   


I thus contend that the neutralist understanding of the role of the state is no more inevitable than the scientism that underlies it. For the moment my delineation of a non-dualist alternative must remain undeveloped, but by now the correlation between liberal neutrality and the assumptions of analytic philosophy about ethics should be clear, and that these assumptions create problems which are psychological in nature. That there also exists a psychological solution to such problems remains to be more fully argued.


[1] Hume (1966)

[2] See 3.h.i

[3] See 3.k.ii

[4] Cassirer (1981) p.90-92

[5] Sluga (1980) p.12-19

[6] ibid. p.58-64

[7] ibid. p.156

[8] Frege (1950) p.71

[9] Sluga (1980) p.135. Sluga also argues (p.131-4) against Dummett’s view that Frege abandoned this contextual view about 1890. Since Sluga generally shows much more awareness of historical context I assume his view to be correct here.

[10] ibid.137-144

[11] ibid. p.147-8

[12] ibid. p.141

[13] ibid. p.154

[14] ibid. p.123-8

[15] Also subsequently and more completely by Gödel’s Incompleteness Theory (1931).

[16] Russell’s letter to Frege of June 16th,1902. Quoted by Sluga (1980) p.163

[17] Frege’s letter to Russell. Quoted by Sluga (1980) p. 164

[18] Ayer (1946) ch.4

[19] Hanfling (1981) p.13

[20] ibid. p.13

[21] Quoted by Hanfling (1981) p.36

[22] Ramsey (1931) p.321

[23] Hanfling (1981) p.15-18

[24] Wittgenstein (1961) § 4.21-4.26

[25] Ayer (1946) p.55-57; Hanfling (1981) p.45-6 (on Waismann)

[26] Hanfling (1981) p. 77-89

[27] Ayer (1946) p.18-19. This was his view in the first edition of Language, Truth and Logic; in the second he added an introduction in which he claimed that some elementary statements about sense-data (of the “this is white” variety) are indubitable (ibid. p.177-8).

[28] ibid. p.20

[29] ibid. p.183

[30] See 2.b.iii

[31] Ayer (1946) p.19-20

[32] ibid. p.185

[33] Lakatos (1974) p.92

[34] See 4.e.i

[35] Quine (1953) p.20-46. Quine’s influential Scepticism about the analytic-synthetic distinction led him to an epistemological holism and even an incrementalism in relation to this issue, but this did not prevent him from holding to strongly scientistic premises in other areas of his writing. He perhaps serves as an example of the limitations of internal reform in analytic philosophy.

[36] Hanfling (1981) p.35

[37] See 4.e.iii. Also see 6.b.iii on the mind-body problem.

[38] E.g. Ayer (1946) p.105

[39] Schlick (1962) p. 213

[40] ibid. p.214-5

[41] See Moore (1959)

[42] See Ayer (1946) p.104-116; Stevenson (1937). Stevenson makes a distinction between two components of the ethical statement, one being descriptive of approval as a psychological state, but the other being a prescription which was entirely expressive. Ayer also modified his position later to one which admitted that ethical statements are descriptive, but of some “queer” state of affairs (Ayer 1965 p.231-3).

[43] The claim of moral objectivity involved in intuitionism might lead one to think that it is actually eternalistic, despite its generally scientistic context and justification. I am happy to leave this as an open question, though I am generally more inclined to see it as a type of scientism because of its relationship to logical positivism and to the psychological features of nihilism. As in the case of Marx, a reclassification would pose no threat to my overall argument. The more important point to stress is the way that intuitionism fractally reflects the overall pattern of eternalism in relation to nihilism on a smaller scale.

[44] MacIntyre (1985) p.17-18

[45] ibid. p.16-17

[46] ibid. p.14-15

[47] ibid. p.11

[48] Hare (1981) p.18

[49] ibid. p.19

[50] ibid. p.20

[51] For more details of what I mean by “functional” here, in the context of linguistic pragmatism, see 5.c.i.

[52] Hare (1981) pp.17, 73-5

[53] ibid. p.206-213

[54] ibid. p.108

[55] ibid. p.108

[56] This is at least the case in Hare (1981); though in Hare (1963) a more imaginative approach is suggested, without the psychological implications of this being recognised.

[57] Hare (1981) p.88

[58] ibid. p.91

[59] See 3.k.iv

[60] Hare (1981) p.60

[61] ibid. p.45

[62] ibid. p.109-111

[63] Hare makes several of the “five mistakes in moral mathematics” identified by Derek Parfit (Parfit 1984 p.67-86), but obviously on my account the mistakes are not ones which can be rectified purely by further analysis: they are symptomatic of the kind of mistake made in purely rational accounts of the objectivity of ethics.

[64] This follows from Hare’s appeal to ordinary language: e.g. (1981) p.15

[65] Hare (1981) p.187

[66] Hare (1963) p.61-3

[67] Plato (1941) p. 14-21: Republic I. 336b et seq.

[68] This general point is also relevant to MacIntyre’s account of rational ethics, discussed in 4.b.iv.

[69] Hare (1981) p.188-9

[70] ibid. p.190-1

[71] ibid. p.195-8

[72] ibid. p.198-205

[73] In fairness to Hare it must be said that he does hint at a recognition at this: “Our intuitions are not even utilitarian, let alone egoistic” (ibid. p.205). However, this does not prevent him from offering the arguments I have been discussing, in which he evidently takes self-interest to be a coherent idea.

[74] For a more detailed account of the argument about Parfit which follows, see Ellis (2000).

[75] Parfit (1984) p.127-8

[76] ibid. p.119

[77] ibid. p.120-126

[78] ibid. p.126

[79] ibid. p.194

[80] ibid. part 3

[81] ibid. p.273. See also 4.c.iv.

[82] Blackburn (1998) p.245-6

[83] Blackburn urges cautious respect for the limitations of the metaphor in this respect, in the process making it clear how the Kantian understanding of the will has wriggled through to the modern era by being interpretable as not involving an ontological claim (ibid. p.249).

[84] ibid. p.250-256

[85] ibid. p.259

[86] ibid. p.260

[87] ibid. p.261-263

[88] ibid. p.309-310

[89] ibid. p.291

[90] ibid. p.319. Blackburn’s account of the fact-value distinction is sophisticated: I shall not discuss it further here, but in 4.f.iv I shall be considering some of the issues raised by Dewey’s similarly subtle and residual type of scientism.

[91] See 4.d.iv

[92] See 3.g.viii & 3.k.v

[93] Jones (1989) discusses the way in which this neutrality is applied in much more detail than I can here. In particular he points out the distinction between negative forms of neutrality, where the state attempts not to promote any value, and positive ones where it attempts to do so even-handedly. Both of these types fall under the idea of neutrality discussed here.

[94] Mill On Liberty ch.2 from Mill (1962)

[95] Dworkin (1978)

[96] Popper (1962b) p. 392-3

[97] Blackburn (1998) p.269-278

[98] See Popper (1957) passim., but esp. p.64-70

[99] See 8.c.ii, with supporting material also in 6.d.iv

[100] In the US and France there is no religious education in state schools, but in Germany a “confessionalism” in which religious education is understood as an instruction in Christianity still prevails: a fact which can perhaps be related to the greater prevalence in Germany of a Kantian style of liberalism rather than a scientistic one.

[101] Ellis (1997)


Return to thesis contents page

Return to home page



A Theory of Moral Objectivity: quick links to other sections


1. introduction

2a. Psychology of belief

2b. Heuristic process

2c. Psychology & philosophy

3ab. Eternalism

3cd. Plato

3e. Stoicism

3f. Christianity

3g. Kant

3h. Hegel

3i. Marx

3j. Schopenhauer

3kl. Utilitarianism

4a. Nihilism

4b. Scepticism & Aristotle

4c. Hume

4d. Analytic Philosophy

4e. Wittgenstein

4f. Pragmatism

4g. Nietzsche

4hi. Existentialists

5. Integration

6. Philosophical Problems

7. Normativity

8. Middle Way Ethics

9. Conclusion

10. Appendix



Other books:

A New Buddhist Ethics

The Trouble with Buddhism


List of all books and papers


Objections & responses


Comment on the blog


Home page