Related Links in the Library:
Source: Editor's introduction to The Collected Works
of John Stuart Mill, Volume IX - An Examination of William Hamilton’s Philosophy
and of The Principal Philosophical Questions Discussed in his Writings, ed. John M. Robson, Introduction
by Alan Ryan (Toronto: University of Toronto Press, London: Routledge and
Kegan Paul, 1979).
Copyright Statement: The online edition of the Collected
Works is published under licence from the copyright holder, The University of Toronto Press. ©2006 The University of Toronto Press. All rights reserved. No part of this material may be reproduced in any form or medium without the permission of The University of Toronto Press.
Introduction by ALAN RYAN
OF THE WORK
an examination of sir william hamilton’s philosophy is not a widely read work; nor is it very highly regarded, even by those who are most attracted to Mill’s writings on philosophy. It contains some instructive set-pieces, which have preserved a sort of exemplary interest: Mill’s analysis of Matter in terms of “permanent possibilities of sensation,” his confessedly abortive analysis of personal identity in similarly phenomenalist terms, his analysis of free-will and responsibility, and his ringing declaration that he would not bow his knee to worship a God whose moral worth he was required to take on trust—all these still find their place in contemporary discussions of empiricism. Mill’s analysis of the nature of judgment and belief perhaps engages the interest of those who hope to explore the problems raised by A System of Logic in a secondary source. But it is doubtful whether many readers who leave the Logic wondering quite what Mill really thought about the epistemological status of arithmetic and geometry find themselves helped by reading the Examination; nor does it add much to Mill’s earlier account of causation, beyond the effective demonstration that whatever rivals there were to Mill’s account, Hamilton’s was not one.
In part, the fallen position of the Examination is the result of the obscurity into which its target has fallen. If the Examination is not much read, then Hamilton’s edition of Reid’s Works is certainly not read now, as it was in Mill’s day, for Hamilton’s elaborate “Dissertations on Reid.” The most recent discussion of Reid’s philosophy, for example, treats Hamilton as a late and somewhat eccentric contributor to the philosophy of common sense. Hamilton’s Lectures on Metaphysics and Logic, of whose repetitive and elementary character Mill was severely critical, were something of an embarrassment to their editors when they appeared after Hamilton’s death. Now they are simply unreadable. The one accessible source for Hamilton’s opinions is the volume of collected essays, Discussions on Philosophy and Literature, Education and University Reform, in which he reprinted his contributions to the Edinburgh Review. Even those essays now attract the educational historian rather more than the philosopher; Hamilton’s attack on the corruption and incompetence of early nineteenth-century Oxford excites more interest than his critique of Cousin’s views on the Absolute.
To the destruction of Hamilton’s philosophical reputation, Mill’s Examination contributed a good deal. Mark Pattison, reviewing the Examination in The Reader, exclaimed:
The effect of Mr Mill’s review is the absolute annihilation of all Sir W. Hamilton’s doctrines, opinions, of all he has written or taught. Nor of himself only, but all his followers, pupils, copyists, are involved in the common ruin. The whole fabric of the Hamiltonian philosophy is not only demolished, but its very stones are ground to powder. Where once stood Sebastopol bidding proud defiance to rival systems is now
- a coast barren and blue
- Sandheaps behind and sandhills before.
The enthusiasm with which Pattison contemplated the ruin of Sir William’s followers may have had rather more to do with the academic politics of Oxford, in which Pattison and Hamilton’s disciple H. L. Mansel were fiercely opposed to one another, than to any very exact appreciation of just which of Hamilton’s doctrines had suffered just what damage. But, although Hamilton’s friends and followers ignored Pattison’s advice that they “had better erect a monument to him, and say nothing about Mr Mill’s book,” they could not restore Hamilton’s status. Mill might not have shown that the intuitive school of metaphysics was inevitably doomed to obscurity and muddle, but it was generally held that he had shown Hamilton himself to be at best obscure, at worst simply incompetent.
Whether Hamilton was worth the expenditure of Mill’s powder and shot is another question. W. G. Ward, writing some years after in the Dublin Review, thought that Mill had done well to take on one representative figure of the anti-empiricist school and pursue him steadily through all the cruces of the argument between associationism and its opponents. But Mark Pattison thought that the cracking of dead nuts just to make sure they were empty was a task which wearied both those who undertook it and those who watched them do it. It is, at the very least, doubtful whether Mill was wise to devote quite so much attention to Hamilton, for the Examination falls awkwardly between the twin tasks of providing a complete critical exposition of Hamilton’s philosophy on the one hand and of providing an equally comprehensive defence of associationism on the other. In effect, Mill’s defence of associationism is spread over the notes he supplied to James Mill’s Analysis of the Phenomena of the Human Mind, and over his reviews, as well as through the Examination. Whatever else may be said for this defence, its organization impedes the reader of the Examination, who is likely to resent having to recover Mill’s views on perception, say, from an argument conducted at several removes from the issues, in which Mill complains of the injustice of Hamilton’s attacking Thomas Brown for supposed misrepresentation of the views of Thomas Reid. It also does something to account for the fact that the criticisms of Mill were criticisms of his positive claims on behalf of associationism more frequently than they were positive defences of Hamilton. Perhaps Mill should have ignored Hamilton altogether, and stuck to the positive task; he certainly left a great many openings for his critics, and might have been better advised to stop them up rather than triumph over Hamilton.
There are more serious problems than these in the way of the reader of the Examination. Mill’s critique of Hamilton and Mansel was one engagement in the battle between empiricism and rationalism. But it was an engagement in which the combatants employed intellectual weapons which we find difficult to use. The argument between Mill and Hamilton is, in their terms, an argument about the nature and contents of “consciousness”; it is in some sense an argument about psychological issues. But whereas we now tend to draw a sharp distinction between the empirical inquiry into the mind and its powers which we call psychology, and the non-empirical inquiry into the possibility of knowledge or into the intelligibility of knowledge-claims which we now call philosophy, no such distinction appears in the Examination. Where we are tolerably sure that philosophical claims about the nature of space and time, or about the nature of perception, ought to be immune from empirical confirmation and disconfirmation, Mill and Hamilton were not. This difference does not make for difficulties with Mill alone; it means that the views of all other philosophers are “read” rather differently by Mill and Hamilton from the way it is natural to us to read them. Thus, Kant’s contribution to philosophy is treated as a contribution to psychology. Where, for instance, we might interpret Kant’s account of the synthetic a priori as entailing that it is a sort of nonsense, though not strictly a grammatical or syntactical sort of nonsense, to suggest that there might be regions of space and time in which the laws of geometry or arithmetic do not apply, Hamilton plainly took the claim to be one about the incapacity of the mind to conceive non-Euclidean space or things which were not countable; and Mill was equally ready to understand Kant in this way, differing over the issue of whether our incapacity to conceive such a space or such objects was part of the original constitution of the mind or the result of experience. To some extent, therefore, readers of the Examination have to engage in a process of translation in order to feel at home with Mill’s argument. Sometimes there are cases which seem to defy the process. Mill’s discussion of how we might come to have the concept of space, for instance, is, as we shall see, very awkward if it is read as an empirical hypothesis about how the furniture of the mind might have been built; and it is more awkward still if it is read as what we now call philosophy.
Against such a background, the proper task of a critic is a matter for debate. Even if we can decently evade any obligation to show that the Examination is a neglected masterpiece, there is a good deal left to do. The task is partly historical and partly philosophical, and it is perhaps an instance of those cases where the history is unintelligible without the philosophy, as well as the other way about. Firstly, something has to be said about why Mill should have decided to write the Examination at all, and about the reasons for its immediate succès both d’estime and de scandale. Then, something must be said about the life and career of Sir William Hamilton, and at least a little about the role of Mill’s other main antagonist, H. L. Mansel. Once the appropriate background in Mill’s career has been filled in, and the main characters have been identified, I shall go on to provide a substitute for the extended analytical table of contents which was once (though it was not part of the Examination) such a useful feature of scholarly works. My account will be both expository and critical, and some at least of the distinctive philosophical views of Hamilton and Mansel will be there explored.
why should mill in particular have devoted himself to writing such a book as the Examination? From his reading of the Discussions shortly after its appearance, Mill had inferred that Hamilton occupied a sort of halfway house, subscribing neither to his own enthusiasm for the principle of the association of ideas nor to the excesses of post-Kantian Continental philosophy, in which, as Mill saw it, we were supposed to know intuitively all sorts of implausible things. Mill explains in his Autobiography, however, that his reading of Hamilton’s posthumously published Lectures during 1861 alerted him to the fact (a fact confirmed by his subsequent study of the “Dissertations on Reid”) that Hamilton was a much more committed and unrestrained intuitionist than he had previously supposed.
As readers of the Autobiography will recall, Mill was very insistent that the struggle between the intuitionists and the school of “Experience and Association” was much more than an academic argument over the first principles of the moral sciences. In explaining why he had written the System of Logic, Mill had said that “it is hardly possible to exaggerate the mischiefs” caused by a false philosophy of mind. The doctrine that we have intuitive and infallible knowledge of the principles governing either our own selves or the outside world seemed to him
the great intellectual support of false doctrines and bad institutions. By the aid of this theory, every inveterate belief and every intense feeling, of which the origin is not remembered, is enabled to dispense with the obligation of justifying itself by reason, and is erected into its own all-sufficient voucher and justification. There never was such an instrument devised for consecrating all deep seated prejudices.
The System of Logic was in quite large part directed at William Whewell, and, up to a point, Mill was right to see Whewell as the defender of conservative and Anglican institutions—he was Master of Trinity, and Mill had refused to attend Trinity as a youth for obvious anti-clerical reasons. The Examination is described in terms which suggest that Mill thought it necessary to return to the attack on the same front. The difference between the intuitionists and the associationists, he says,
is not a mere matter of abstract speculation; it is full of practical consequences, and lies at the foundation of all the greatest differences of practical opinion in an age of progress. The practical reformer has continually to demand that changes be made in things which are supported by powerful and widely spread feelings, or to question the apparent necessity and indefeasibleness of established facts; and it is often an indispensable part of his argument to shew, how those powerful feelings had their origin, and how those facts came to seem necessary and indefeasible.
One might doubt whether there was any very close practical connection between, say, a Kantian view of knowledge and conservatism on the one hand, and a Humean view and liberalism on the other. Certainly it is hard to imagine Hume welcoming the French Revolution, had he lived to see it, and it is not very difficult to construct radical political philosophies of a broadly intuitionist kind. Kant at least welcomed the French Revolution, even if he trembled before the execution of Louis XVI.
But Mill had no doubt that some such connection did hold.
I have long felt that the prevailing tendency to regard all the marked distinctions of human character as innate, and in the main indelible, and to ignore the irresistible proofs that by far the greater part of those differences, whether between individuals, races, or sexes, are such as not only might but naturally would be produced by differences in circumstances, is one of the chief hindrances to the rational treatment of great social questions, and one of the greatest stumbling blocks to human improvement.
He therefore decided that it was right to produce something more combative and controversial than a treatise on the associationist philosophy of mind. It was necessary to attack the chief exponent of the opposite view—hence what some readers will surely think of as the grindingly negative tone of a good deal of the Examination. Mill, in many ways, was ill-fitted to assault Hamilton in this fashion; he was too fair-minded to let Hamilton’s case take its chances, and therefore encumbered his attack with enormous and tedious quantities of quotation from Hamilton. Yet at the same time he was so entirely unsympathetic to Hamilton that he rarely paused to wonder if some rational and useful case might be extracted from the confused jumble, which was all that Hamilton’s writings eventually seemed to him to amount to. In a way, he could neither do his worst to Hamilton, nor could he do his best for him.
Yet the attack was a sort of duty, especially in view of the use made of Hamilton’s philosophy of the conditioned by his pupil Mansel. H. L. Mansel’s Bampton Lectures had aroused a good deal of indignation from the time of their delivery in 1858, and they went into several editions, with replies to critics appended to new editions. Mansel’s aim had been something like Kant’s—to limit the pretensions of reason to make room for faith. Accordingly, he had argued that we were obliged as a matter of faith to believe that God was everything that was good, although “good,” as applied to the Almighty, was a term which was at best related only by analogy to “good” applied to a human being. Mill thought that this conclusion amounted to using Hamilton’s doctrine to justify a “view of religion which I hold to be profoundly immoral—that it is our duty to bow down in worship before a Being whose moral attributes are affirmed to be unknowable by us, and to be perhaps extremely different from those which, when we are speaking of our fellow-creatures, we call by the same names.”
The implausibility of Mill’s attempt to line up the progressives behind the doctrine of association and the reactionaries behind the doctrine of intuitive knowledge is neatly illustrated by his conjoining Hamilton and Mansel in this fashion. Their political allegiances were practically as far apart as it was possible to get. Mansel was politically a Tory, and was conservative in educational matters too. He was one of the most powerful defenders of the old tutorial arrangements that characterized teaching at Oxford and distinguished it from the Scottish and German universities. Hamilton, on the other hand, was a liberal in politics, thought the tutorial system beneath contempt, thought Oxford colleges entirely corrupt, and, had he been able, would have swept away the whole system in favour of something modelled on the Scottish system.
Mill’s intention of provoking a combat à outrance was wholly successful. The Examination attracted much more attention than the System of Logic had done. Mansel’s long review of it, The Philosophy of the Conditioned—which only covered the first few chapters on the principle of the relativity of knowledge and the attack on his Bampton Lectures—came out within months. James McCosh produced a volume, In Defence of Fundamental Truth, intended to defend those parts of Hamilton’s philosophy which were most characteristic of the Scottish philosophy of common sense. Within two years Mill was preparing a third edition of the Examination in which these and several other extended attacks were answered; the furore continued in the years before Mill’s death, with the appearance in 1869 of John Veitch’s Memoir of Sir William Hamilton Bart., a pious defence of the opinions as well as the life of his old teacher, and W. G. Ward’s further assault on associationism in the Dublin Review in 1871. The balance of the comments was undoubtedly hostile to Mill, less because of a widespread enthusiasm for the doctrines of Sir William Hamilton than because of a widespread fear that their rejection must lead to what McCosh almost invariably conjoined as “Humeanism and Comtism”—a mixture of atheism and dubious French politics. In this sense Mill’s belief that he was fighting the pious and the conservative was absolutely right, for it was they—with the exception of some support from Herbert Spencer on the one topic of self-evidence—who were his hostile reviewers. Even then, some of the supposedly pious and the conservative were more in sympathy with Mill than with Hamilton. Two notable adherents were William Whewell, who, for all that he was Mill’s victim on many occasions, had no doubt that Hamilton was an intellectual disaster who had set the course of speculation back by twenty years, and F. D. Maurice, who had been a harsh and persistent critic of Mansel for years.
It is difficult to know when this interest in the argument between Mill and Hamilton died. From what evidence there is, it looks as though an interest in the Examination lasted so long as the System of Logic was still doing its good work in changing the philosophical syllabus in Oxford and Cambridge. But during the 1870s a new and in many ways more professional generation of philosophers became prominent, who had in one sense absorbed as much as they needed of Mill’s work and, in another, were determined to clear away his intellectual influence. In Oxford at any rate, it was T. H. Green and F. H. Bradley who set the pace; and they were not inclined to defend Hamilton for the sake of refuting Mill, especially when their epistemological allegiances were Hegelian rather than patchily Kantian. So Bradley’s Ethical Studies contains an extremely effective analysis of Mill’s account of personal identity, but does not bother with the rest of the contest between the transcendental and empiricist analysis of the relations between mind and matter. And Green, though he applies to Mill the criticisms he develops against Hume, does not treat the Examination as the locus classicus of Mill’s views. Thereafter, it seems that anyone much interested in Mill’s philosophy would look into the Examination only for the range of topics mentioned at the beginning of this Introduction.
HAMILTON AND MANSEL
ALTHOUGH THE NAME OF HAMILTON is scarcely mentioned now, except in connection with his doctrine of the quantification of the predicate, it seems a proper estimate of his eminence in the first half of the nineteenth century to say that he and Mill were the two people in Britain whose names might occur to a philosophically educated foreigner who was asked to name a British thinker of any distinction. Sorley’s History of English Philosophy, for instance, links the two names together in precisely this sense. And it seems that if one had asked teachers in American universities during the middle years of the century what contemporary influences they felt from Britain, they would have talked of Hamilton and Mill—though a little later the influence of Spencer would no doubt have been, if anything, stronger.
Hamilton was born in Glasgow on 8 March, 1788, in one of the houses in Professors’ Court, for his father was Professor of Botany and Anatomy. His father died when William was only two years old, but there is no evidence that the family suffered any financial difficulties in consequence, and Mrs. Hamilton’s character was quite strong enough to ensure that the absence of the father’s hand was not much felt.
After attending both Scottish and English schools and Glasgow and Edinburgh Universities, Hamilton began in 1807 a distinguished academic career at Balliol College, Oxford. In spite of his exceptional erudition and an epic performance in the final examination in Classics, as a Scot he received no offer of a fellowship, and returned to study law at Edinburgh, being admitted to the bar in 1813. His legal career was distinguished solely by a successful application (heard by the sheriff of Edinburgh in 1816) to be recognized as the heir to the Baronetcy of Preston and Fingalton.
If his nationality cost him the first opportunity of academic preferment, it was his Whig sympathies that scotched the second when, in 1820, he failed to succeed Thomas Brown in the Chair of Moral Philosophy in Edinburgh. The following year he obtained an underpaid and undemanding Chair in Civil History, but he made no mark in intellectual circles until 1829, when he began to contribute to the Edinburgh Review.
His first article, on Cousin, was an editor’s nightmare, being late in arrival, much too long, and completely beyond the grasp of most of the readers of the Review. But it was a great success with Cousin himself, and it served notice on the outside world that someone in the British Isles was abreast of European philosophy. It was for the Edinburgh that Hamilton wrote the most readable of his work: the two essays on “The Philosophy of the Conditioned” and on “Perception,” his essay on “Logic” which contains (at least on Hamilton’s reading of it) the first statement of the doctrine of the quantification of the predicate, and his condemnation of the intellectual and legal condition of the University of Oxford. It cannot be said that they were thought, even at the time, to be uniformly readable; Napier, the editor, was frequently reduced to complaining of the excessive length, the overabundant quotations, and the archaic forms of speech which Hamilton indulged in. But, as Mill’s account would lead one to expect, it is these essays, reprinted in his Discussions, which show Hamilton at his best and most accessible. Even then, there are longueurs attributable less to the mania for quotation that to the combative manner of the author. The essay on perception, for instance, is so grindingly critical of Thomas Brown that the reader loses patience with the argument.
In 1836, however, academic justice was at last done. The Chair of Logic and Metaphysics in Edinburgh fell vacant, and this time the City Council elected him, by eighteen votes to fourteen. The composition of lectures for the courses he was now obliged to give followed very much the same pattern as his literary exploits—everything was done too late and too elaborately; so in his first year Hamilton not infrequently worked until dawn the night before delivering his lectures, and then took what rest he could while his wife got the day’s lecture into shape for delivery. Shortly after the election, he embarked on his edition of the Works of Reid. This was a characteristically acrimonious business, in which Hamilton started work at the suggestion of Tait, the Edinburgh bookseller, then took offence at the financial arrangements proposed by Tait (who seems to have expected a volume of Reid’s writings with a short preface, rather than something with as much of Hamilton’s erudition as Reid’s thinking in it, and who was not willing to pay for labours he had no wish to see anyone undertake), and published the edition at his own expense in 1846.
Hamilton’s active career was relatively brief. In 1844 he suffered a stroke, which did not impair his general intellectual grasp, but left him lame in the right side and increasingly enfeebled. He had to have his lectures read for him much of the time, although he managed to keep up a reasonably active role in the discussion of them. He was, however, well enough to see the republication of his earlier essays and to carry on a violent controversy with Augustus De Morgan, both about their relative priority in the discovery of the principle of the quantification of the predicate, and about its merits. De Morgan was vastly entertained by the violence of Hamilton’s attacks, both because he enjoyed the resulting publicity it conferred on his own work and, so far as one can see, because he liked having an argument with someone so uninhibited in his aggression as was Hamilton. Others were less sure: Boole, thanking Hamilton for the gift of a copy of the Discussions, took the opportunity to say: “I think you are unjustifiably severe upon my friend Mr De Morgan. He is, I believe, a man as much imbued with the love of truth as can anywhere be found. When such men err, a calm and simple statement of the ground of their error answers every purpose which the interests either of learning or of justice can require.” The effort was wasted twice over, seeing that Hamilton was unlikely to become more moderate, and De Morgan was perfectly happy to be abused.
Hamilton’s health became worse after a fall during 1853, and he became less mentally active in the last two or three years of his life. Retirement, however, was impossible, since he could not live without the £500 a year that the Chair gave him. Despite these outward difficulties, and the acerbity of his writings, all was not gloom and grimness. Hamilton’s domestic life was strikingly happy; when he died on 6 May, 1856, he left behind a devoted family, loyal pupils, and a good many friends as well.
A matter of much more difficulty than establishing the outward conditions of his life is working out how Hamilton came to exercise such a considerable influence on the philosophical life of the country. He created enthusiastic students, of whom Thomas S. Baynes became the most professionally and professorially successful, but otherwise it seems to have been the weight of learning of a half-traditional kind which backed up the reception of his views. His innovations in logic, for instance, were produced in articles which were largely devoted to a minute chronicle of the fate of deductive logic in the fifteenth to seventeenth centuries. His views on perception, or on the relativity of knowledge, are always placed in the framework of an historical analysis of the sort which the higher education of the time encouraged. How much it assisted his, or anyone’s, understanding of Kant to yoke him with Plato for the purposes of comparison and contrast is debatable, but the weight it added to his arguments looked to some of his audience very much like intellectual power rather than mere weight. He was more or less an intellectual fossil thirty years after his death, however. Sir Leslie Stephen’s account of Hamilton in the Dictionary of National Biography presents him as an eccentric and pedantic leftover from the Scottish school of common sense. And Stephen’s marginal comments in his copy of the Discussions display the exasperation Hamilton is likely to induce; at the end of “Philosophy of the Conditioned,” the pencilled comment reads: “A good deal of this seems to be very paltry logomachy. His amazing way of quoting ‘authorities’ (eg Sir K. Digby, Walpole & Mme de Stael) to prove an obvious commonplace is of the genuine pedant. And yet he had a very sound argument—only rather spoilt.”
Henry Longueville Mansel was Hamilton’s chief disciple in Oxford. Born in 1820 he shone as a pupil first at Merchant Taylor’s School and then at St. John’s College, Oxford; and in 1843, with a double First in Mathematics and Classics, he settled down with great pleasure to the task of tutoring clever undergraduates; he was regarded throughout the university as its best tutor. He held the first appointment as Waynflete Professor of Metaphysical Philosophy, and therefore counts R. G. Collingwood, Gilbert Ryle, and Sir Peter Strawson among his intellectual progeny. With his interest in Kant and his German successors, and his astringent, largely destructive approach to the subject he professed, he might almost be said to have set the boundaries of the subsequent style.
Mansel was a productive writer: his Prolegomena Logica appeared in 1851; his Metaphysics, which was an expansion of a substantial essay for the Encyclopædia Britannica, in 1860. He was most widely known as the author of The Limits of Religious Thought, the Bampton Lectures for 1858. This work was reprinted several times, and aroused a great deal of controversy, in which F. D. Maurice played an especially acrimonious role. Philosophically, Mansel was greatly indebted to Kant, but he was very hostile to Kant’s theology and to Kant’s moral philosophy alike. The Limits of Religious Thought was described by Mansel himself as
an attempt to pursue, in relation to Theology, the inquiry instituted by Kant in relation to Metaphysics; namely, How are synthetical judgments à priori possible? In other words: Does there exist in the human mind any direct faculty of religious knowledge, by which, in its speculative exercise, we are enabled to decide, independently of all external Revelation, what is the true nature of God, and the manner in which He must manifest Himself to the world . . . ?
The answer he gave was that there was no such faculty of religious knowledge, and that natural theology was quite unable to set limits to the nature and attributes of God. Moreover, he shared none of Kant’s certainty that our moral faculty allowed us to judge supposed revelations by their consistency with divine goodness. What goodness is in the divinity is not a matter on which human reason is fit to pronounce.
Mansel was not only a productive writer; he wrote elegantly and lucidly. There are many reasons for wishing that it had been Mansel’s Metaphysics which Mill had examined, rather than Hamilton’s Lectures, and the clarity of Mansel’s prose is not the least. Even in the pious context of the Bampton Lectures he is witty—replying to a critic who complains that Mansel’s attack on rationalism in theology is an attempt to limit the use of reason, he says that it is only the improper use of reason he is rejecting: “All Dogmatic Theology is not Dogmatism, nor all use of Reason, Rationalism, any more than all drinking is drunkenness.” It was not surprising that progress came quickly. In 1855 he was elected to the Readership in Moral and Metaphysical Philosophy, and in 1859 to the Waynflete Professorship. Mansel’s wit and exuberance were, however, not matched by physical strength. His acceptance of the Chair of Ecclesiastical History in 1866 was a partial recognition of the need to conserve his energy, and a move to London as Dean of St. Paul’s in 1868 more explicit recognition. Besides, by the mid-1860s he was finding the moderately reformed Oxford increasingly uncongenial to his conservative tastes. In 1871 he died suddenly in his sleep.
The contrasts between Mansel and Hamilton are so complete that it is difficult to know why Mansel was so devoted a follower of “the Edinburgh metaphysician”—for his devotion did indeed extend to employing Hamilton’s logical innovations in rather unlikely contexts, and even to defending them against De Morgan. What is evident so far is that Mansel required nothing much more than an ally against the pretensions of Absolute Idealism; but that judgment plainly understates the strength of his conviction. It is obviously preposterous to think of Mansel and Hamilton as sharing any political commitment which would account for such a degree of conviction. It is more reasonable to suppose that they shared something which one can only gesture towards by calling it a matter of religious psychology. Mansel genuinely seems to have thought that an acknowledgement of the limitations of human reason was a more reverent attitude towards the unknowable God than any attempt to look further into His nature, and he seems to have been impressed by a similar outlook in Hamilton:
True, therefore, are the declarations of a pious philosophy:—“A God understood would be no God at all;”—“To think that God is, as we can think him to be, is blasphemy.”—The Divinity, in a certain sense, is revealed; in a certain sense is concealed: He is at once known and unknown. But the last and highest consecration of all true religion, must be an altar—Ἀγνώστῳ Θεῳ̑—“To the unknown and unknowable God.”
Hamilton’s insistence that his doubts about Absolute knowledge are not only compatible with, but in some sense required by, Christian revelation is practically the theme of Mansel’s Bampton Lectures. Between them and Mill there was a gulf, therefore, but one less political than Mill’s Autobiography suggests. It was the gulf between Mill’s utterly secular, this-worldly temperament and their sense of the final mysteriousness of the world. The harshness of Mansel’s attack on the Examination in The Philosophy of the Conditioned reflects his resentment of this matter-of-fact approach to the world, a resentment which cannot have been soothed by the fact that in Oxford, as elsewhere, the staples of a Christian philosophy, such as Butler’s Analogy, were losing ground to such textbooks as the System of Logic.
THE PHILOSOPHY OF THE CONDITIONED
the opening shots of Mill’s campaign against Hamilton’s philosophy are directed against “the philosophy of the conditioned.” The burden of Mill’s complaint against Hamilton is that his attachment to what he and Mill term “the relativity of knowledge” is intermittent, half-hearted, explained in incoherent and self-contradictory ways. He accuses Hamilton of both asserting and denying that we can have knowledge of Things in themselves, and of giving wholly feeble reasons for supposing that we cannot conceive of, particularly, the nature of space and time as they are intrinsically, but can nevertheless believe that they are genuinely and in themselves infinite. It is this part of Hamilton’s philosophy that Mansel’s essay on The Philosophy of the Conditioned had to endeavour to rescue; his Bampton Lectures on The Limits of Religious Thought hung on the negative claim that the human mind could not conceive of the nature of the Deity, so that He remained inaccessible to philosophical speculation, and on the positive claim that there was still room for belief in such an inconceivable Deity. Mansel’s version of the philosophy of the conditioned was intended to repel the pretensions of philosophy in the sphere of religion. “Pantheist” philosophers of the Absolute, such as Hegel and Schelling, were unable to provide knowledge of an Absolute that might replace, or be recognized as the philosophically reputable surrogate of, the God of Christianity; less ambitious philosophers were shown to be unable to restrict the attributes of a Deity by the categories of human reason. As this account suggests, the Kantian overtones in Mansel’s work are very marked, and, as we shall see, The Philosophy of the Conditioned gives a very Kantian interpretation of Hamilton.
Yet the oddity, or perhaps we should only say the distinctive feature, of Hamilton’s philosophy on its metaphysical front was the combination of the critical philosophy of Kant with Reid’s philosophy of common sense. Hamilton’s position seems at first to be exactly that of Reid. He sided with Reid and common sense in holding that “the way of ideas” is suicidal, that any theory which presents the external world as a logical construction from the immediate objects of perception (construed as “ideas”) simply fails to account for the world’s true externality. In particular, he held, with Reid, that what we perceive are things themselves, not a representation of them, or an intermediary idea. Moreover, some of the properties which we perceive things to possess really are properties of the objects themselves, and not contributions of the percipient mind. The secondary qualities he was willing to recognize as not existing in the object itself, but primary qualities were wholly objective, not observer dependent. The knowledge we have of things, however, still remains in some sense relative or conditioned. The question is, in what sense?
It is at this point that the invocation of Kant’s criticalism causes difficulties, for Hamilton could afford to take only a few details from Kant if he was not to run headlong against Reid. Above all, he wanted to side with Kant against Kant’s successors, and to deny that we can know anything of the Absolute or the Unconditioned. He wanted, that is, to deny the possibility of a positive pre- or post-critical metaphysics, in which it was supposed to be demonstrated that Space and Time were in themselves infinite—or not. But he did not want to follow Kant in his “Copernican revolution”; or, rather, he could not have intended to do anything of the sort. For Hamilton did not think that the contribution of the percipient mind to what is perceived is anything like as extensive as Kant claimed. The implication for metaphysics of the “relative” or “conditioned” nature of human knowledge he certainly took to be what Kant claimed it to be:
The result of his examination was the abolition of the metaphysical sciences,—of Rational Psychology, Ontology, Speculative Theology, &c., as founded on mere petitiones principiorum. . . . “Things in themselves,” Matter, Mind, God,—all, in short, that is not finite, relative, and phænomenal, as bearing no analogy to our faculties, is beyond the verge of our knowledge. Philosophy was thus restricted to the observation and analysis of the phænomena of consciousness; and what is not explicitly or implicitly given in a fact of consciousness, is condemned, as transcending the sphere of a legitimate speculation. A knowledge of the Unconditioned is declared impossible; either immediately, as an intuition, or mediately, as an inference.
But he refused to draw Kant’s conclusions about the subjectivity of space and time, and denied that the antinomies showed that they were only forms of intuition:
The Conditioned is the mean between two extremes,—two inconditionates, exclusive of each other, neither of which can be conceived as possible, but of which, on the principles of contradiction and excluded middle, one must be admitted as necessary. On this opinion, therefore, our faculties are shown to be weak, but not deceitful. The mind is not represented as conceiving two propositions subversive of each other, as equally possible; but only, as unable to understand as possible, either of two extremes; one of which, however, on the ground of their mutual repugnance, it is compelled to recognise as true.
In effect, Hamilton’s view seems to have been that Reid and common sense were right in holding that what we perceive are real, material objects, located in an objective space and time, objectively possessed of (some of) the properties we ascribe to them, but that Kant was right in holding that those properties which we can ascribe to them must be adapted to our faculties, “relative” in the sense of being related to our cognitive capacities.
The question of the sense in which all our knowledge is thus of the relative or the conditioned is not quite here answered, however. For there remains a considerable ambiguity about the nature of this relativism, or relatedness. The simplest reading turns the doctrine of relativity into a truism. It amounts to saying that what we can know depends in part upon our perceptive capacities, and that beings with different perceptual arrangements from our own would perceive the world differently. In that sense, it is no doubt true that what we perceive of the world is only an aspect of the whole of what is there to be perceived. More philosophically interesting is an exploration of why we seem able to agree that we might in principle perceive the world quite otherwise than we do, but find it impossible to say much about how we might do so. Mill, however, pursues that topic no further than to its familiar sources in the questions asked by Locke—whether a man born blind could conceive of space, for instance (222ff.). Mill’s chief complaint is that Hamilton confuses several senses of relativity together, when talking of the relativity of knowledge, and that the only sense he consistently adheres to is this truistic sense. In any real sense, says Mill, Hamilton was not a relativist:
Sir W. Hamilton did not hold any opinion in virtue of which it could rationally be asserted that all human knowledge is relative; but did hold, as one of the main elements of his philosophical creed, the opposite doctrine, of the cognoscibility of external Things, in certain of their aspects, as they are in themselves, absolutely
When Hamilton attempts to reconcile this objectivist account with the doctrine of the relativity of knowledge, flat contradiction is only averted by retreat into banality:
He affirms without reservation, that certain attributes (extension, figures, &c.) are known to us as they really exist out of ourselves; and also that all our knowledge of them is relative to us. And these two assertions are only reconcileable, if relativity to us is understood in the altogether trivial sense, that we know them only so far as our faculties permit.
Mill was not the severest critic of Hamilton on this score. J. H. Stirling’s critique of Hamilton’s account of perception treats Hamilton’s views with complete contempt. The contradiction between the objectivist account and the relativist account of our knowledge of the outside world is so blatant that Hamilton cannot have failed to notice it. Where Mill suspects Hamilton of mere confusion, Stirling accuses him of disingenuousness. Mill demurely declines to press any such charge (cv). He did not even suggest that Reid and Kant made awkward allies in principle. In an earlier article on “Bain’s Psychology” he had indeed yoked Reid and Kant together as members of the a priori school of psychological analysis. But he went on to point out that the question of the connection between our faculties and the nature of the external reality was an issue of ontology rather than psychology; and here Reid was “decidedly of opinion that Matter—not the set of phenomena so called, but the actual Thing, of which these are effects and manifestations—is congnizable by us as a reality in the universe.” This comment suggests that Mill thought of Hamilton as discussing metaphysics in a wide sense—both “the science of being” and psychology; Reid, Kant, and Hamilton were allies in so far as they belonged to the same camp in psychology, but they made an ill-assorted trio in matters of ontology. Here Kant and Reid belonged to different camps and no one could tell where Hamilton stood. Mansel’s reply to Mill was to insist that everything in Reid, and everything in Hamilton which expressed an allegiance to Reid, should be as it were put in Kantian brackets. We might perceive things themselves, but the “thing itself” which we perceive is not the “thing-in-itself,” but only the phenomenally objective thing. The thing known in perception was the appearance to us of a noumenon of which nothing whatever could be known.
There is something to be said for Mansel’s claims. Reid at times writes as if knowledge is doubly relative: in the knower, it is a state of an ego of which we only know the states, though convinced that it exists as a continuing substance; and, in the known, what we know is states of things external to us, though again we are irresistibly convinced of their continued substantial existence. But we cannot safely go far along this path. Reid did not like to talk of substances, and certainly did not wish to introduce them as mysterious substrates; to the extent that Mansel rescues Hamilton by claiming that external things are known “relatively” as phenomena related to imperceptible noumena, he goes against the evident thrust of Reid’s views. The further one presses Hamilton’s attachment to Kant beyond his avowed enthusiasm for the destructive attack on positive metaphysics, the harder it is to get any textual backing for the case. It is doubtless true that a sophisticated Kantian would have been untroubled by Mill’s attack, but it is quite implausible to suggest that that is what Sir William Hamilton was.
At all events, Mill’s approach to Hamilton is initially entirely negative. Mill does not put forward any view of his own on the relativity of knowledge. The reason is a good one so far as it goes. Mill’s distinction between the a priori and a posteriori schools of psychology is one which only partially overlaps his main theme. For in the Examination, just as in the Logic, Mill’s hostility is directed against those who attempt to infer the nature of the world from the contents and capacities of our minds. In principle, there is no reason why there should be any overlap between a priorism in psychology and the view that mental capacities and incapacities reflect real possibilities and impossibilities in the world. A priorism, as Mill describes it, is a psychological approach which refers our most important beliefs about the world, and our moral principles, too, to instincts or to innate capacities or dispositions. The sense in which these are a priori is not very easy to characterize, although the fact that many of the instinctive beliefs described by the a priori psychologists of Mill’s account coincide with the judgments described by Kant as synthetic a priori suggests most of the appropriate connotations. Thus the perception that objects occupy a space described by Euclidean geometry embodies the instinctive judgment that bodies must occupy space, and the necessity ascribed to the truths of geometry reflects the instinctive judgment that, for instance, two straight lines cannot enclose a space, and so on. Such judgments, says Mill, purport to be a priori in the sense that they have to be presumed true before experience is possible, or at any rate characterizable. Whether they are held to be temporally prior to experience is, he recognizes, not essential: there is no need to deny that children have to learn arithmetic in order to deny that its truths reflect the teachings of experience. Mill sees that it is quite arguable that the capacity to recognize necessities of thought is one which matures in the child, and requires experience to set it to work. Indeed, at times, he seems to suggest that the dispute between a priori and a posteriori psychologists is an empirical dispute in which there need not be only two opposing sides. For if the issue is one of how much of an adult’s understanding of the world we can account for as the result of individual learning, there will be a continuum between psychologists who stress the extent to which such an understanding is as it were preprogrammed into the human organism and those who stress how much of it can be accounted for by trial-and-error learning from the organism’s environment. In like manner, with reference to the area of moral and prudential reasoning, there would be a similar continuum between those who see us as relatively plastic and malleable organisms and those who claim to see some moral and prudential attachments more or less genetically built in.
Now, in so far as the argument proceeds in these terms, it will still follow a pattern which is visible in Mill’s own approach. That is, the environmentalist must attempt to show some way in which the capacity, whose acquisition he is trying to explain, could have been built up through experience; the innatist will respond by showing that there are features of such a capacity which are simply omitted or more subtly misrepresented by such an account. The question of how much of what we perceive of the world is to be credited to the programme by which the percipient organism organizes its physical interaction with the world, and how much is to be set down to learning, is then an empirical question, or rather a whole series of empirical questions. This was the point at which Mill and Herbert Spencer came close to agreement. Spencer’s long discussion of the nature of intuitive knowledge in the Fortnightly Review is a protest against being assigned to the rationalist camp by Mill, in which Spencer’s central point is that when we refer our sensations to external objects as their causes this is, as it were, a hypothesis proferred by the organism, a hypothesis which we cannot consciously shake, and one on which we cannot help acting. Nonetheless, it is only a hypothesis; it is, however, one which seems to have been programmed into us by evolution, and one whose reliability is most readily accounted for by the theory that the external world is, indeed, much as we perceive it is. The doctrine is not one which would perturb Mill; he ascribed something very like it to Reid.
This assertion, however, does imply that Mill’s own interest in the relativity of knowledge as a central issue in epistemology rather than psychology, would necessarily be slight. That the organic constitution of human beings sets limits to what they could hope to know about the world was an uninteresting empirical truth; interesting truths about the ways in which we were prone to illusions in some areas, or about the ways in which we estimated the size, shape, movement, or whatever of external bodies, would emerge piecemeal. Mill never quite propounded a version of the verification principle, and therefore never went to the lengths of suggesting that what one might call transcendental relativism or transcendental idealism was simply meaningless, because its truth or falsity could make no observational difference. But he came very close.
He came particularly close when he turned from Hamilton’s views on the positive relativity of knowledge to Hamilton’s negative case, as set out in his critique of Cousin. In his attack on Cousin, Hamilton had denied that we can ever attain to positive knowledge of “the Infinite” and “the Absolute”; Mill dismantles Hamilton’s various arguments to this effect, distinguishing Kantian arguments to show that we can know nothing of noumena from arguments against the possibility of an “infinite being.” They are, he points out, directed at very different targets. That our knowledge is phenomenal, not noumenal, “is true of the finite as well as of the infinite, of the imperfect as well as of the completed or absolute” (58-9). The “Unconditioned,” in so far as it is to be identified with the noumenal, is certainly not an object of knowledge for us. But “the Absolute” and “the Infinite” are in considerably worse shape than the merely noumenal. These, though Hamilton never meant to go so far, are shown up as a tissue of contradictory attributes: “he has established, more thoroughly perhaps than he intended, the futility of all speculation respecting those meaningless abstractions ‘The Infinite’ and ‘The Absolute,’ notions contradictory in themselves, and to which no corresponding realities do or can exist” (58). To Mansel’s reply that Hamilton had not tried to argue that they were meaningless abstractions, Mill had a ready retort:
I never pretended that he did; the gist of my complaint against him is, that he did not perceive them to be unmeaning. “Hamilton,” says Mr Mansel, “maintains that the terms absolute and infinite are perfectly intelligible as abstractions, as much so as relative and finite.” Quis dubitavit? It is not the terms absolute and infinite that are unmeaning; it is “The Infinite” and “The Absolute.” Infinite and Absolute are real attributes, abstracted from concrete objects of thought, if not of experience, which are at least believed to possess those attributes. “The Infinite” and “The Absolute” are illegitimate abstractions of what never were, nor could without self-contradiction be supposed to be, attributes of any concrete.
Mill’s harassment of Hamilton on the Absolute and the Infinite has few lessons of great moment. It is interesting that Mill does not adopt, as he might have done, Hobbes’s method of dealing with the question of infinity. Where Hobbes had said that “infinite” characterizes not the attribute itself, but our incapacity to set a limit to whatever attribute is in question, Mill treats it as an attribute, that of being greater than any completed attribute of the appropriate sort—a line of infinite length is thus longer than any completed line. Some attributes could be characterized as absolutely present, but not infinitely so, others as infinitely but not absolutely present. The purity of water has an absolute limit, viz., when all impurities are absent, but there is no sense to be given to the notion of infinitely pure water. Concerning this issue, Mill changed his mind on minor points from one edition to another. He began by claiming that power could be infinite, but knowledge only absolute, because absolute knowledge meant knowing everything there is to be known; but under pressure from Mansel and other critics, he agreed that a being of infinite power would know everything he could think or create, so that his knowledge would be infinite also (37-8). But he is casual about such concessions, quite rightly seeing them as having little bearing on the main question, whether there is any sense at all to be attached to such notions as “the Absolute.”
It is surprising that Mill does not press his opponents harder on the meaninglessness of propositions about beings with infinite attributes and the rest. Mansel in particular, but Hamilton also, was very vulnerable to the charge that in showing God or the Unconditioned to be beyond our conceiving, they had also shown them to be beyond our believing. Both Hamilton and Mansel were utterly committed to the principle that what was not a possible object of knowledge was nevertheless a proper object of belief. Mansel stated his position with characteristic lucidity in the Preface to his Bampton Lectures:
“the terms conceive, conception, &c., as they are employed in the following Lectures, always imply an apprehension of the manner in which certain attributes can coexist with each other, so as to form a whole or complex notion. . . . Thus when it is said that the nature of God as an absolute and infinite being is inconceivable, it is not meant that the terms absolute and infinite have no meaning—as mere terms they are as intelligible as the opposite terms relative and finite—but that we cannot apprehend how the attributes of absoluteness and infinity coexist with the personal attributes of God, though we may believe that, in some manner unknown to us, they do coexist. In like manner, we cannot conceive how a purely spiritual being sees and hears without the bodily organs of sight and hearing; yet we may believe that He does so in some manner. Belief is possible in the mere fact (τὸ ὅτι). Conception must include the manner (τὸ πω̑ς).
The obvious question invited is, what is the mere fact believed in? If we cannot form any conception of the state of affairs which is said to be the object of our belief, it is not clear that we can be said to know what we believe at all. Mill’s attack on the discussion of “the Infinite” and “the Absolute” concentrates, as we have just seen, on the claim that they cannot be talked about because they are literal self-contradictions; Mansel does not quite go to the length of saying that self-contradictory propositions might be true, though we cannot imagine how, and Mill does not press on him the obvious dilemma that he must either say that, or admit that the terms he is using no longer bear their usual meaning, and perhaps bear no clear meaning at all.
What Mill does argue against Hamilton is that no sooner has Hamilton routed those of his opponents who believe that we have direct knowledge of the unconditioned, or perhaps an indirect and implicit knowledge only, than he joins forces with them by letting what they describe as “knowledge” back into his system under the label of “belief.” If one were looking for the weak points in Mill’s account of Hamilton, this brief attack would surely be one place to seek them in. In essence, Mill’s complaint is that whatever Hamilton had maintained about the relativity of knowledge, and whatever scepticism he had evinced about the Unconditioned, everything would have been
reduced to naught, or to a mere verbal controversy, by his admission of a second kind of intellectual conviction called Belief; which is anterior to knowledge, is the foundation of it, and is not subject to its limitations; and through the medium of which we may have, and are justified in having, a full assurance of all the things which he has pronounced unknowable to us; and this not exclusively by revelation, that is, on the supposed testimony of a Being whom we have ground for trusting as veracious, but by our natural faculties
Mill’s outrage is intelligible enough. If one supposes that philosophical first principles are supposed to furnish a set of premises from which we can deduce the general reliability of our knowledge, then some such method as that of Descartes is the obvious one to pursue, and it would seem that first principles must be better known than anything that hangs upon them. At least it would seem scandalous to any Cartesian to suppose that we merely believed in our own existence and yet knew that bodies could not interpenetrate or that the sun would rise again in the morning. Yet it is doubtful whether this is how Mill ought to have understood Hamilton. Spencer, who tackled the issue more sympathetically, suggested a more plausible interpretation, and one which does more justice than Mill’s to the difference between a Cartesian and a Kantian view of first principles. Mill, who treats the difference between belief and knowledge very much as twentieth century empiricism was to do—that is, regarding knowledge as justified true belief (65n)—cannot allow for a difference in the ways of treating particular knowledge claims and claims about the whole of our knowledge. But Spencer does just that. When we claim to know something, we assume that we can set our belief against external evidence; but we cannot peel off the whole of our knowledge of the world from the hidden world of which it is knowledge and claim that we now know that it is knowledge. All we can do is believe that it really is knowledge. More than one twentieth-century philosopher of science has similarly claimed that we can only make sense of the sciences’ claim to supply us with knowledge of the world if we believe in an occult, underlying, objective order in the world, which is beyond experience but accounts for its possibility.
It is only when Mill comes to sum up the successes and failures of the philosophy of the conditioned that he supplies the reader with what is most required—an explanation of what Mill himself understands by inconceivability, and how he explains it, in opposition to the intuitionists and innatists. The explanation occupies a considerable space, but it is worth noticing two main points. The first is Mill’s claim that the majority of cases of inconceivability can be explained by our experience of inseparable associations between attributes, and the other his claim that most of the things that Hamilton claims to be inconceivable are not difficult, let alone impossible, to conceive. What is most likely to scandalize twentieth-century readers is the way Mill treats it as an empirical psychological law that we cannot conjoin contradictory attributes, and therefore cannot conceive things with contradictory attributes. The source of the scandal is obvious: we are inclined to hold that it is a matter of logic that a thing cannot have inconsistent attributes, not because of any property of things or our minds, but because a proposition is logically equivalent to the negation of its negation, and to ascribe a property and its contradictory to an object is simply to say nothing. The assertion negates and is negated by the denial of it. The law of non-contradiction, on this view, cannot be interpreted psychologically, without putting the cart before the horse: that a man cannot be both alive and not alive is not the consequence of our de facto inability to put the ideas of life and death together.
Mill, however, suggests something like a gradation, from flat contradiction through decreasingly well-attested repugnances of attributes:
We cannot represent anything to ourselves as at once being something, and not being it; as at once having, and not having, a given attribute. The following are other examples. We cannot represent to ourselves time or space as having an end. We cannot represent to ourselves two and two as making five; nor two straight lines as enclosing a space. We cannot represent to ourselves a round square; nor a body all black, and at the same time all white.
But he goes on to make something nearer a sharp break between flat contradiction and everything else:
A distinction may be made, which, I think, will be found pertinent to the question. That the same thing should at once be and not be—that identically the same statement should be both true and false—is not only inconceivable to us, but we cannot imagine that it could be made conceivable. We cannot attach sufficient meaning to the proposition, to be able to represent to ourselves the supposition of a different experience on this matter. We cannot therefore even entertain the question, whether the incompatibility is in the original structure of our minds, or is only put there by our experience. The case is otherwise in all the other examples of inconceivability.
These, Mill begins by saying, are only the result of inseparable association; but he rather confusingly qualifies this by suggesting that even there the inconceivability somehow involves the contradictoriness of what is said to be inconceivable: “all inconceivabilities may be reduced to inseparable association, combined with the original inconceivability of a direct contradiction” (70). The point he is making is, evidently, the following. We cannot conceive of a state of affairs characterized as A and not-A, because the conception corresponding to A is just the negative of the conception of not-A. In other cases, there is no direct contradiction; it is A and B we are asked to conceive jointly, and if we are unable to do so it is because in our experience B is always associated with not-A. Hence the attempt to conceive A and B turns out to be special case of trying to conceive A and not-A, and the real point at issue between Mill and the opposition is the nature of our certainty that in these proposed instances B really does imply not-A. Mill thinks it is an empirical conviction, implanted by experience, reflecting the way the world actually is, but telling us nothing about how it has to be. The opposition have no common doctrine; the Kantian members of it think that the conviction reflects how the world has to be, but only in the sense that since “the world” is a phenomenal product of our minds working upon unknown and unknowable data it must obey the laws of our own minds; Catholic transcendentalists like W. G. Ward claimed to be objectivists and realists on this issue, where the Kantians were subjectivists and phenomenalists; they held that real inconceivabilities in our minds reflect the necessity of a certain rational structure to the universe, a structure that is not a matter of choice even for Omnipotence itself. So, in attacking Mill’s attempt to explain the truths of mathematics in experiential terms, Ward says:
I have never even once experienced the equality of 2+9 to 3+8, and yet am convinced that not even Omnipotence could overthrow that equality. I have most habitually experienced the warmth-giving property of fire, and yet see no reason for doubting that Omnipotence (if it exist) can at any time suspend or remove that property.
Mill himself makes something like a concession to the Kantian mode of analysis, though it is a physiological rather than a psychological version of transcendental idealism that he perhaps offers. In the body of the text he claims that “a round square” is in principle no more inconceivable than a heavy square or a hard square; to suppose that one might exist is no more than to suppose that we might simultaneously have those sensations which we call seeing something round and those which we call seeing something square:
we should probably be as well able to conceive a round square as a hard square, or a heavy square, if it were not that, in our uniform experience, at the instant when a thing begins to be round it ceases to be square, so that the beginning of the one impression is inseparably associated with the departure or cessation of the other
But in a later footnote he drew back:
It has been remarked to me by a correspondent, that a round square differs from a hard square or a heavy square in this respect, that the two sensations or sets of sensations supposed to be joined in the first-named combination are affections of the same nerves, and therefore, being different affections, are mutually incompatible by our organic constitution, and could not be made compatible by any change in the arrangements of external nature. This is probably true, and may be the physical reason why when a thing begins to be perceived as round it ceases to be perceived as square; but it is not the less true that this mere fact suffices, under the laws of association, to account for the inconceivability of the combination. I am willing, however, to admit, as suggested by my correspondent, that “if the imagination employs the organism in its representations,” which it probably does, “what is originally unperceivable in consequence of organic laws” may also be “originally unimaginable.”
The note nicely illustrates the difficulty of seeing quite what Mill’s case was. Even here he seems determined to appeal to the laws of association, and yet the case he is partially conceding is that there are structural constraints on what things can be perceived and therefore come to be associated. Evidently the one thing he is determined not to concede is that the laws of the Macrocosm can be inferred from the laws of the Microcosm; but as he says, he is here at one with Hamilton and Mansel.
Yet it is this view which Mill mostly writes to defend, and perhaps in a form which does set him apart from Hamilton and Mansel. For Mill plainly treats the question of what we can and cannot conceive as a flatly factual one, and so, in turn, he treats the laws of number or the findings of geometry as flatly factual too. Indeed, he goes so far as to claim that even with our present mental and physical constitution we could envisage alternative geometries and different arithmetical laws. “That the reverse of the most familiar principles of arithmetic and geometry might have been made conceivable, even to our present mental faculties, if those faculties had coexisted with a totally different constitution of external nature, is,” says Mill, “ingeniously shown in the concluding paper of a recent volume, anonymous, but of known authorship, ‘Essays, by a Barrister’ [i.e., Fitzjames Stephen]” (71n), and he quotes the paper at length. The gist of it is that we can perfectly well imagine a world in which 2+2=5; for all we need imagine is a world in which “whenever two pairs of things are either placed in proximity or are contemplated together, a fifth thing is immediately created and brought within the contemplation of the mind engaged in putting two and two together” (71n). Mill does not suggest, what is surely rather plausible, that such a statement of the case is self-destructive, in that it presupposes that what we should say under such conditions is not that 2+2=5, but, as he does say, that associating pairs creates a fifth object. The supposition, of course, is much more complicated in any case than Mill allows. As Frege later argued, things are only countable under a common concept—a cow and a sheep are not a pair of cows nor a pair of sheep, but they are a pair of animals, mammals, familiar English objects, and so on. Are we to suppose that they spontaneously generate a fifth something or other when conceptualized one way but not another? Can we stop the process by thinking of four things, not as two pairs but as a trio and an individual? Are addition and subtraction supposed to cease to be isomorphic, so that 5-2=3, even though 2+2=5? Nor is it clear what the notion of contemplating pairs is going to embrace. If I read a word of six letters, do I read a word of three pairs of letters, and if so, is it not a word of at least seven letters? Or will it stay one word of only six letters so long as I read it as one word only—in which case how will anyone ever learn to read? There is, no doubt, something contingent about the fact that our system of geometry and arithmetic apply in the world, but it is hardly so flatly contingent as this account suggests.
Mill is much more persuasive when he sets out to deny Hamilton’s claims about the limitations from which our thinking necessarily suffers. Mill distinguishes three kinds of inconceivability, which, he says, Hamilton habitually confuses. The first is what we have been examining until now, the supposed impossibility of picturing the states of affairs at stake, either directly or indirectly as the result of its making contradictory demands on the imagination. The second is the apparent incredibility of what is perfectly visualizable. Mill’s example is the existence of the Antipodes; we could model a globe in clay and recognize that there need be no absolute “up” or “down,” but still fail to see how people could remain on the surface of the globe at what we were sure to think of as its underside (74-5). Finally, there is a sense in which an event or state of affairs is inconceivable if it is impossible to see what might explain it: “The inconceivable in this third sense is simply the inexplicable.” Mill says, and quite rightly, that it merely invites confusion to employ “inconceivable” to cover mere inexplicability:
This use of the word inconceivable, being a complete perversion of it from its established meanings, I decline to recognise. If all the general truths which we are most certain of are to be called inconceivable, the word no longer serves any purpose. Inconceivable is not to be confounded with unprovable, or unanalysable. A truth which is not inconceivable in either of the received meanings of the term—a truth which is completely apprehended, and without difficulty believed, I cannot consent to call inconceivable merely because we cannot account for it, or deduce it from a higher truth.
Oddly enough, it was Mansel who got into the most serious muddle here, and for no very obvious reason. He denied that Hamilton had ever used the term “inconceivable” to cover more than the unimaginable, and yet, as we have seen already, employed the term himself in Mill’s third sense. We believe that the will is free, but we cannot explain how it is, and so, on Mansel’s view, we have here a believable inconceivability. Had he stuck simply to saying that we can conceive that something is the case where we cannot conceive how it is, there would be no problem—what is imaginable and credible is the bare fact, what is unimaginable is a mechanism which might account for it. The connection, as Mill is quick to see, between the narrower, proper senses of inconceivable, and the wider, improper sense, is that the offer of a hypothetical mechanism to account for a phenomenon makes it so much the easier both to visualize it and to believe in its existence. None of this, of course, is to deny that Mansel is quite right to suggest that the mind does indeed boggle at the task of explaining how the physical interaction of brain and world results in perceptions which are themselves not in any obvious sense physical phenomena at all; all it shows is that there is no point in muddying the waters by suggesting that the facts are inconceivable when what one means is that they are in certain respects inexplicable.
Having cleared up these terminological difficulties, Mill then embarks on the question of whether, as Hamilton claims, the philosophy of the conditioned shows that there are propositions about the world which are inconceivable and yet true. The examples Mill has in mind, as we have seen, are such propositions as that space is finite, or, conversely, that space is infinite. The language of conceivability causes a few more difficulties, even after Mill’s sanitizing operations, for between Mill and Mansel there remains a difference of opinion on the question of what it is to have a conception of any state of affairs. Mansel seems to require that there should be some kind of one-to-one relationship between the elements in our conception and that of which it is the conception. Mill does not entirely repudiate this view; it will serve as a criterion for having an adequate—or perhaps one had better say, a complete—conception of the phenomenon that one should be able to enumerate the elements in one’s conception and match them to the components of the thing conceived. But, says Mill, in one of his most felicitous moves, it is impossible to have a wholly adequate conception of anything whatever, since everything and anything can be envisaged in an infinite number of ways. The obsession with the infinite and absolute in Hamilton and Mansel is ill-defended by Mansel’s arguments about adequacy, since, says Mill, there is no suggestion that a number like 695,788 is inconceivable, and yet it is pretty clear that we do not enumerate its components when we think of it (84).
What, then, is it for us to conceive of space as infinite, or conversely, as finite? On Mill’s view, we can conceive of an infinite space by simply conceiving of what we call space and believing that it is of greater extent than any bounded space.
We realize it as space. We realize it as greater than any given space. We even realize it as endless, in an intelligible manner, that is, we clearly represent to ourselves that however much of space has been already explored, and however much more of it we may imagine ourselves to traverse, we are no nearer to the end of it than we were at first. . . .
The same confidence applies to conceiving of space as finite. Mill supposes that all we need to imagine is that at some point or other an impression of a wholly novel kind would announce to us that we were indeed at the end of space. The extent to which neither Mill nor Hamilton, nor Mansel for that matter, takes the full measure of Kant is somewhat surprising. There is no suggestion that drawing the boundaries of space is conceptual nonsense because boundaries are something one draws in space, so that if space is finite it must be finite but unbounded. There is no attempt to explore further what could lead us to recognize an experience as, say, the experience of reaching the end of time or the end of space.
For, as we have seen, Mill does not do more than skirt round the suggestion that “infinite” may have something odd about it, if it is treated as an ordinary first-order predicate, or that “Space” may be the name of an object to which it is only dubiously proper to apply a predicate like “finite.” Mill does not extend the notion of “meaninglessness” beyond its most literal applications. He thinks that it is impossible to conceive what is meant by a literally meaningless utterance, or one to which we can attach no meaning, but that this is not a philosophically interesting sort of inconceivability:
If any one says to me, Humpty Dumpty is an Abracadabra, I neither knowing what is meant by an Abracadabra, nor what is meant by Humpty Dumpty, I may, if I have confidence in my informant, believe that he means something, and that the something which he means is probably true: but I do not believe the very thing which he means, since I am entirely ignorant what it is. Propositions of this kind, the unmeaningness of which lies in the subject or predicate, are not those generally described as inconceivable.
For Mill, then, in so far as the states of affairs described by Hamilton as inconceivable are picked out by intelligible propositions, it becomes a question of fact, even if one which there is no hope of deciding, which branch of the antinomies proposed by Hamiton is true. In that case, what of the philosophy of the conditioned? The answer, says Mill, is that there is in it a good deal less than meets the eye. Hamilton’s claim that “Thought is only of the conditioned,” and that the “Conditioned is the mean between two extremes—two inconditionates, exclusive of each other, neither ofwhich can be conceived as possible, but of which, on the principles of contradiction and excluded middle, one must be admitted as necessary,” turns out to be nothing better than noise. It “must be placed in that numerous class of metaphysical doctrines, which have a magnificent sound, but are empty of the smallest substance” (88).
GOD AND PROFESSOR MANSEL
with hamilton thus routed, Mill turns to meet Mansel’s application of the philosophy of the conditioned to religious thought. Neither Mill’s attack nor Mansel’s response stands out as a model of dispassionate and impersonal inquiry. Mill all but accuses the clergy of being under a professional obligation to talk nonsense (104), and Mansel replies in kind. Mill opens his assault by paying Mansel a backhanded compliment: “Clearness and explicitness of statement being in the number of Mr. Mansel’s merits, it is easier to perceive the flaws in his arguments than in those of his master, because he often leaves us less in doubt what he means by his words” (91). In fact, it is not always quite clear where Mansel does and where he does not rest on arguments borrowed from Hamilton; against Mill he tended to argue by complaining of Mill’s defective appreciation of the history of philosophy, a procedure which has the defect of turning the interesting question of where Mill and Mansel disagreed over the possible extent of a human knowledge of God’s nature into a much less interesting question, about the extent of Mill’s acquaintance with traditional natural theology. Mansel was probably right in his conjecture that in some sense Mill thought traditional metaphysics was pointless and nonsensical, but he was far too annoyed to tackle the question that he had really set for himself—namely, if traditional natural theology and traditional metaphysics were as essentially flawed as The Limits of Religious Thought maintained, was Mill not right? Why was not agnosticism the proper resting place?
Still, Mill hardly encouraged Mansel to adopt a conciliatory attitude. After a rapid summary of Mansel’s argument that we cannot form an adequate conception of God—since God as Absolute and Infinite is inconceivable by us—he comes to Mansel’s conclusion that we can only fall back on revelation. That the God thus revealed can or cannot have any particular characteristics, Mansel says it is not for reason to declare; the credibility of a revelation is a matter of historical probabilities, “and no argument grounded on the incredibility of the doctrine, as involving an intellectual absurdity, or on its moral badness as unworthy of a good or wise being, ought to have any weight, since of these things we are incompetent to judge” (90). It is not, says Mill, a new doctrine, but “it is simply the most morally pernicious doctrine now current . . . ” (90).
Readers who have begun to weary of the hunting of the Absolute will probably take it on trust that in so far as “the Absolute” means the unrelated-to-anything-in-our-experience it is no great achievement to show that we have no knowledge of the Absolute. But Mill presses Mansel rather harder than this, for he at last challenges him to make good on the claim that we are able and indeed obliged on the strength of revelation to believe in this unknowable entity. Mansel, says Mill, succeeds in showing that “the Absolute” and “the Infinite” as defined by himself are simply self-contradictory; but, on Mill’s view, this entails their being also unbelievable. “Believing God to be infinite and absolute must be believing something, and it must be possible to say what” (98). Mansel’s argument to the effect that “the Absolute” and “the Infinite” are involved in self-contradiction is altogether too devastating for his own good, for Mansel certainly does not want to say that the divine nature is really and inherently contradictory. Mansel, indeed, went out of his way to deny any such suggestion; credo quia impossibile he thought unworthy of any sane man. His reply to Mill, abusive though it is, shows how little he wished to get himself into such depths, for when Mill taunts him with not being able to say what the object of his belief is, he falls back on propositions which Mill readily admits to be intelligible, such as the proposition that God made the world, though we cannot tell how He did it. The explanation of the trouble is simple, though rather strange. Mansel thought it an aid to Christian belief to show that the sceptic could not attack its doctrines on rational grounds; but the way in which he rescued them from the sceptic was by making them too elusive to disbelieve. Inevitably the price he paid was making them too elusive to be believed either.
The single thing in the Examination that most heartened his allies and most outraged his opponents was Mill’s assault on what he took to be the immorality of Mansel’s doctrine of the unknowability of the moral attributes of God. To Mill the issue was simple enough. When the clergy talked of God’s power they generally meant what we would mean by talking of human power, for instance the divine ability to throw us into the inferno; only on God’s moral attributes did they equivocate and suggest that God’s goodness was not as mortal goodness.
Is it unfair to surmise that this is because those who speak in the name of God, have need of the human conception of his power, since an idea which can overawe and enforce obedience must address itself to real feelings; but are content that his goodness should be conceived only as something inconceivable, because they are so often required to teach doctrines respecting him which conflict irreconcilably with all goodness that we can conceive?
Whether it is or not, Mill’s case is that Mansel cannot hope to argue that God’s moral attributes are unlike their human analogues without thereby sacrificing the right to expect us to worship Him. There is, as any reader of Mansel’s Bampton Lectures can see, an awkwardness in Mansel’s case, analogous to the awkwardness of his epistemology. The case he presents is the familiar one: the Christian who believes in the infinite power and goodness of God is confronted with a world in which the just suffer and the wicked flourish. The austere Mansel does not argue in the Kantian manner that we are thereby licensed to expect a reconciliation of virtue and happiness in the life hereafter. What he does instead is suggest that the inscrutability of God extends to the inscrutable goodness He exhibits. It is not clear that Mansel intends to show that God’s goodness is not ours; mostly, he argues that how God is working out an overall plan for His universe, a plan which is good in the same sense as a human plan would be good, simply remains unknowable. The goodness of God’s agents particularly exercises Mansel: what would be cruelty or injustice if done otherwise than in obedience to God’s commands is, we must hope, not cruelty or injustice after all. But, once again, it is less a matter of the imperfect analogy between human and divine attributes (which is the object of Mill’s complaint) than of the imperfection of our knowledge of the Almighty’s programme, for the sake of which these orders were given. In this light one can understand why Mansel’s reply to Mill takes the form of a rather querulous complaint that surely Mill cannot deny that a son may recognize the goodness of his father’s actions without wholly understanding them—and Mill does not deny it.
Mill, however, surely gets the best of the dispute, with his famous outburst, for all that Mansel tries to dismiss it as “an extraordinary outburst of rhetoric.”
If, instead of the “glad tidings” that there exists a Being in whom all the excellences which the highest human mind can conceive, exist in a degree inconceivable to us, I am informed that the world is ruled by a being whose attributes are infinite, but what they are we cannot learn, nor what are the principles of his government, except that “the highest human morality which we are capable of conceiving” does not sanction them; convince me of it, and I will bear my fate as I may. But when I am told that I must believe this, and at the same time call this being by the names which express and affirm the highest human morality, I say in plain terms that I will not. Whatever power such a being may have over me, there is one thing which he shall not do: he shall not compel me to worship him. I will call no being good, who is not what I mean when I apply that epithet to my fellow-creatures; and if such a being can sentence me to hell for not so calling him, to hell I will go.
OTHER MAJOR ISSUES
as one might guess from the title of Mansel’s The Philosophy of the Conditioned, it was that doctrine which Mansel, like Mill, saw as Hamilton’s most distinctive contribution to philosophy (109). The rest of this Introduction will take its cue from the combatants, and confine itself to the piecemeal treatment of some major issues. The most interesting of these would seem to be the following: Mill’s phenomenalist analysis of matter and mind; his demolition of Hamilton’s account of causation, which is perhaps a major curiosity rather than a major issue; his account of conception, judgment, and inference, and his assessment of Hamilton’s contribution to logic; and, finally, his analysis of the freedom of the will.
MATTER AND MIND
Mill’s account of matter and mind begins with what amounts to a hostile review of Hamilton’s own hostile review of Thomas Brown’s Lectures on the Philosophy of the Mind. (Hamilton’s article appeared in the Edinburgh Review in October, 1830, and was reprinted in his Discussions.) Hamilton declared that it was a striking proof of the low state of intellectual life in Britain that Brown’s Lectures had not hitherto received their just deserts:
The radical inconsistencies which they involve, in every branch of their subject, remain undeveloped; their unacknowledged appropriations are still lauded as original; their endless mistakes, in the history of philosophy, stand yet uncorrected; and their frequent misrepresentations of other philosophers continue to mislead. In particular, nothing has more convinced us of the general neglect, in this country, of psychological science, than that Brown’s ignorant attack on Reid, and, through Reid, confessedly on Stewart, has not long since been repelled;—except, indeed, the general belief that it was triumphant.
Hamilton claimed that Brown played fast and loose not only with the testimony of consciousness, a vice to which all philosophers are liable to succumb, but with the testimony of Reid. Brown was what Hamilton called a cosmothetic idealist, and Hamilton was at pains to insist that between the testimony of consciousness—which is all on behalf of “Natural Realism” or “Natural Dualism”—and the inferences of idealism there is a great opposition. Reid, on Hamilton’s view, was a realist and dualist, where Brown falsely makes him out to be an idealist of the same kind as himself.
Mill devotes a chapter to showing not merely that Reid wavered in his convictions on the question, but that when he was plainly committed to any view, that view was cosmothetic idealism. Moreover, very few of Hamilton’s arguments against Brown hold water, and when Hamilton adduces, to attack Brown, general principles, such as the impossibility of representative perception, the result, on Mill’s account, is to leave Brown untouched and most of Hamilton’s own argument in ruins (164). Mill distinguishes, with Hamilton, three views about perception which have been held by those he lumps together as cosmothetic idealists: the first is the view that what is really perceived is not a state of the perceiver’s mind, but something else, whether a motion in the brain as in Hobbes or an Idea in the mind as in Berkeley; the second is the view that what is perceived is a state of mind, but that it and the perceiving of it are distinguishable. These two doctrines, says Mill, really are doctrines of mediate or representative perception, as Hamilton says they are. There is a something which is the direct object of perception and which represents the external object. The third view, however, and the view which Brown held, is not a theory of representative perception at all, for there is no tertium quid, no object of direct perception from which the existence of some other object is inferred. The object of perception here is “a state of mind identical with the act by which we are said to perceive it” (155). There is here no very clear distinction between a certain sort of phenomenalism on the one hand and outright realism on the other, indeed—a point which Mill does not make, but which some current versions of a “sense data” theory of perception do.
Brown’s account of the perception of external objects is invulnerable to the objection that there is no way of knowing whether the object of perception resembles, or truly or faithfully represents, the external object itself. For Brown does not claim that it bears any such relationship to anything external. The relation is causal, not pictorial. In effect, to perceive something in the outside world just is to be in a certain sensory state and to conclude non-inferentially that the cause of this state lies in something external to oneself. And this, says Mill happily, is the only rational interpretation to be placed on the views of Reid as well. Indeed,
if Brown’s theory is not a theory of mediate perception, it loses all that essentially distinguishes it from Sir W. Hamilton’s own doctrine. For Brown, also, thinks that we have, on the occasion of certain sensations, an instantaneous and irresistible conviction of an outward object. And if this conviction is immediate, and necessitated by the constitution of our nature, in what does it differ from our author’s direct consciousness? Consciousness, immediate knowledge, and intuitive knowledge, are, Sir W. Hamilton tells us, convertible expressions; and if it be granted that whenever our senses are affected by a material object, we immediately and intuitively recognise that object as existing and distinct from us, it requires a great deal of ingenuity to make out any substantial difference between this immediate intuition of an external world, and Sir W. Hamilton’s direct perception of it.
Brown, on Mill’s account, gets the better of Hamilton by consistently denying that some properties of things are known as they really are in the (unknowable) object and some not; Brown genuinely held the doctrine of the relativity of knowledge in an unconfused form (167). In this Brown was on the opposite side to both Reid and Hamilton, but it was an issue on which not even Hamilton was willing to suggest that Brown was unaware of the differences between his own views and those of Reid. Brown’s theory of perception explains all our knowledge of the attributes of matter in terms of the sensory promptings of an external cause, while Reid’s, like Hamilton’s, allows us “a direct intuition of the Primary Qualities of bodies” (176). Mill, of course, thinks that Brown’s view is the only one consistent with his premises; certainly, as Mill argues both earlier and later in the Examination, Hamilton can hardly hope to keep his half-way house. Either he must be a thoroughgoing vulgar realist and agree that what we see just are things, endowed with the attributes we see them to have, the plain man’s view; or else, if he is to allow himself such corrections of consciousness as are required when he says, for instance, that no two people see the same object, or indeed that each of us sees two “suns,” say, because we receive an image through each eye, and in so saying departs very widely from what any plain man believes, then he must adopt a much more wholesale subjectivism.
Mill’s own account of what we believe when we believe in the existence of the outside world is the best known part of the Examination. It is hard to know whether to be more surprised by the confidence with which he puts it forward or by the contrast between that confidence and the diffidence, so reminiscent of Hume, with which he confesses that it will not yield a plausible analysis of mind. Mill’s account of matter seeks to analyze it in terms of possible sensations. In effect, the requirements of something’s being a material thing, distinct from our sensations of it, are the following: it must be public in the sense that it can be perceived by many different people, whereas each of them alone can have his actual sensations; it must be “perdurable,” that is, it must exist unperceived, and must outlast the fleeting experiences of it which those who perceive it may have; and it must retain the same properties even if these make it “look different” in different circumstances.
We mean, that there is concerned in our perceptions something which exists when we are not thinking of it; which existed before we had ever thought of it, and would exist if we were annihilated; and further, that there exist things which we never saw, touched, or otherwise perceived, and things which never have been perceived by man. This idea of something which is distinguished from our fleeting impressions by what, in Kantian language, is called Perdurability; something which is fixed and the same, while our impressions vary; something which exists whether we are aware of it or not, and which is always square (or of some other given figure) whether it appears to us square or round—constitutes altogether our idea of external substance. Whoever can assign an origin to this complex conception, has accounted for what we mean by the belief in matter.
The question is, of course, whether an appeal to “possible sensations” can account for all this. Perhaps the first thing that should be said is that Mill is oddly reticent about employing the fact that human beings are embodied consciousnesses in any of the argument; later, he employs the sensations of muscular effort and resistance as part of the primitive data which he suggests the mind works on in arriving at a conception of space. But it is on the face of it odd to begin arguing about the belief in an external world without raising any question about what external can mean unless “external to me,” and how it can mean that, unless we are spatially located from the beginning—and how, if we are so located, it can make any sense to begin to construct a world whose existence we seem to have to assume in order to talk about the constructive task in the first place. Mill can, of course, retort that he is not talking about spatial externality yet. What he is talking about initially is permanence; it is a second part of the case to show that a permanent object in sensation has to be construed—or is naturally to be construed—as a spatially external object. That is, so long as we do not insist on publicity, and do not have too many qualms about whether something could be round or square except in a spatially extended world, we could perhaps break up the belief in a material world into a belief in something permanent which holds together the objects of sense and into a second belief that it is located in space as well as in time. If we think of the percipient as a non-spatial ego in which subjective experiences inhere and which has a history as the history of one such being, we might think of the non-ego as the objective correlate of the percipient self. It is not at all clear that Mill had any such possibility in mind, and it is quite clear that we shall not get very much out of Mill’s account by pressing it; nonetheless, to the extent that Mill takes over the terminology of Hamilton, in which we are said to be conscious of an Ego and a non-Ego, the question whether the non-Ego is an external—that is spatially external—world is evidently an open one. The first step establishes a non-Ego as a deliverance of consciousness, if we side with Hamilton, and as an inference if we side with Mill; only subsequent steps can establish its nature.
Mill at any rate is eager to show that so long as the mind is credited with a capacity to form expectations, we can see how the mind would move from having had experiences in certain circumstances in the past, to believing in possible experiences realized by similar conditions in the future. These, Mill says, are not bare possibilities but conditional certainties—by which he merely means to insist that he does not suggest that, in the everyday sense, it is only “possible” that when we look at a chair we shall have the appropriate sensations. He means that we shall quite certainly have the appropriate sensations, but, of course, only in the appropriate conditions. The mind, then, faces the fact that its experiences occur in various determinate ways; it constructs the hypothesis that this orderliness will be found in all sorts of other areas, and finds it confirmed. The content of the hypothesis is that the world contains permanent possibilities of sensation, and the world turns out to do so. Mill is eager not to turn the Permanent Possibilities themselves into mental constructions; in a footnote replying to a critic who had complained that Mill had offered “no proofs that objects are external to us,” he says that he had never attempted any such proof:
I am accounting for our conceiving, or representing to ourselves, the Permanent Possibilities as real objects external to us. I do not believe that the real externality to us of anything, except other minds, is capable of proof. But the Permanent Possibilities are external to us in the only sense we need care about; they are not constructed by the mind itself, but merely recognised by it; in Kantian language, they are given to us, and to other beings in common with us.
It is their givenness which explains the sense in which they are objective rather than subjective; whether this makes them external in a sense which would satisfy the plain man as well as the philosopher remains to be seen.
That there is an external world is a sort of hypothesis, then. It is formed entirely unconsciously, of course, but the awkwardness is not its genesis but its meaning. Mill seems unworried by this, and given the remark quoted immediately above, it is easy to see why. He could share Brown’s view of what the belief in an external world amounted to—namely belief in an underlying cause of our sensory experience—since his interest lay not in disputing the adequacy of the analysis, but in accounting for the fact thus analyzed without invoking anything like an original conviction of the existence of an external world. Not for nothing did Mill call his account the psychological theory of the belief in an external world; he thought that Hamilton, Reid, and for that matter Brown, too, had erred by adopting the “introspective” method of analysis, by which he meant that they were too ready to infer from the present existence of a belief in their own minds that it was part of the mind’s native constitution. The psychological theory was in principle no more than a genetic hypothesis, a hypothesis about how the belief could have grown up. As such, it seems to be a rather difficult one to bring to empirical test, although such a test seems appropriate for it; the difficulties are too obvious to be worth dwelling on, but they make one wonder why Mill did not make more of the question whether there was any way of averting them. Would he have regarded infantile efforts at focussing on remote objects as evidence one way or the other? Would a new-born baby’s recoil from what looks like a sheer drop be evidence about how original a sense of spatial location might be? In the absence of more discussion in Mill’s work, speculation is fruitless.
Whether Mill’s analysis of matter would satisfy the plain man’s notions about matter is a question to which he does devote some attention. He has two rather different stances. The first is that the belief in matter goes beyond the belief in the permanent possibility of sensation: we move from believing that we shall have certain sensations under certain conditions to believing that the whole series of possible sensations has an underlying cause. Now, on this view, we are at any rate inclined to ask whether this belief in an underlying cause actually means anything—since it makes no observational difference whether or not there is such a cause, there is some difficulty in knowing what difference is made by its affirmation or denial. Believers in parsimony, Occam’s Razor, or other austerities of thought will perhaps incline to reject it on the grounds that we should believe as little as we must to account for the facts; Mill thinks that Hamilton’s “Law of Parsimony” should cause him an analogous embarrassment, but makes nothing of it in this context—he is concerned to reduce the number of our primary intuitions, rather than to purge the plain man’s ontology. This being his aim, he is quite content to argue that
Whatever relation we find to exist between any one of our sensations and something different from it, that same relation we have no difficulty in conceiving to exist between the sum of all our sensations and something different from them. . . . This familiarity with the idea of something different from each thing we know, makes it natural and easy to form the notion of something different from all things that we know, collectively as well as individually. It is true we can form no conception of what such a thing can be; our notion of it is merely negative; but the idea of a substance, apart from its relation to the impressions which we conceive it as making on our senses, is a merely negative one. There is thus no psychological obstacle to our forming the notion of a something which is neither a sensation nor a possibility of sensation, even if our consciousness does not testify to it; and nothing is more likely than that the Permanent Possibilities of sensation, to which our consciousness does testify, should be confounded in our minds with this imaginary conception. All experience attests the strength of the tendency to mistake mental abstractions, even negative ones, for substantive realities.
On the whole, this argument suggests that the generality of mankind hold mistaken views about matter, though its intention may only be to suggest that they hold unverifiable views. But Mill also suggests that he and the plain man may not be at odds.
Matter, then, may be defined, a Permanent Possibility of Sensation. If I am asked, whether I believe in matter, I ask whether the questioner accepts this definition of it. If he does, I believe in matter: and so do all Berkeleians. In any other sense than this, I do not. But I affirm with confidence, that this conception of Matter includes the whole meaning attached to it by the common world, apart from philosophical, and sometimes from theological, theories. The reliance of mankind on the real existence of visible and tangible objects, means reliance on the reality and permanence of Possibilities of visual and tactual sensations, when no such sensations are actually experienced.
This view, in contrast to the first one, suggests that the plain man qua plain man believes in Permanent Possibilities only; the belief in an unknowable underlying substance is either imposed on him by philosophers, or adopted by the plain man only qua amateur philosopher.
The argument between phenomenalists and their opponents has, of course, continued unabated ever since. It is not only the plain man who feels uneasily that Mill’s “permanent possibilities of sensation” moves awkwardly between an account of matter which stresses that it is permanently and objectively available to be sensed, and one which dissolves that objective existence into the fact that minds are permanently available to sense—but not necessarily to sense anything other than their own contents. It is at the very best difficult to feel that a possible, but non-actual sensation is more solid, more material, more firmly part of the furniture of the world than an actual sensation is.
Before turning to Mill’s attempt to provide a phenomenalist account of personal identity, therefore, we should look to Mill’s expansion of his analysis of matter in the shape of his account of our knowledge of its primary qualities. Mill’s analysis is devoted to several different tasks, of which the most important is to show that the “psychological theory” can deal with the generation of the idea of Extension, which
has long been considered as one of the principal stumbling blocks of the Psychological Theory. Reid and Stewart were willing to let the whole question of the intuitive character of our knowledge of Matter, depend on the inability of psychologists to assign any origin to the idea of Extension, or analyse it into any combination of sensations and reminiscences of sensation. Sir W. Hamilton follows their example in laying great stress on this point.
But Mill also wants to explain two other things, firstly, the difference between what we treat as subjective feelings as distinct from what we treat as perceptions of something in the object and, secondly, why we group the objective properties of bodies together as their primary qualities. These did not cause much controversy among Mill’s critics, but the attempts at generating the idea of extension along the lines laid down in Bain’s treatise on psychology did. The fundamental complaint was always the same, that all attempts to explain where we might have acquired the concept of extension presuppose that we have it already. As Mill says in the footnote in which he replies to them:
A host of critics, headed by Dr. McCosh, Mr. Mahaffy, and the writer in Blackwood, have directed their shafts against this chapter. . . . The principal objection is the same which was made to the two preceding chapters [on the Psychological Theory of the belief in an external world, and its application to mind]: that the explanation given of Extension presupposes Extension: that the notion itself is surreptitiously introduced, to account for its own origin.
The distinction between sensations referred mostly to the subject of perception and those referred mostly to the object, Mill explains fairly casually. That we can refer the experience to an outer object is the major difference between sensation and other mental phenomena; so, the pleasure of a man eating a good meal can be said to inhere in the meal, but is more readily ascribed to the man than the meal, because pleasure and pain are part of a class of “sensations which are highly interesting to us on their own account, and on which we willingly dwell, or which by their intensity compel us to concentrate our attention on them.” The result is that in our consciousness of them “the reference to their Object does not play so conspicuous and predominant a part . . .” (212). Mill does not appeal to the way in which the pleasure and, to a lesser extent, the pain caused by a given object varies from one person to another as a reason for distinguishing the pleasure and pain from what causes them; nor does he suggest that there is anything problematic in treating secondary qualities like colour in the same way as pleasure and pain. The distinction he is interested in is really that which his opponents see as a distinction between the essence of matter, and all else. If we can imagine a thing losing its colour without ceasing to exist, and losing its capacity to give pain or pleasure without ceasing to exist, then colour and pleasure lie on the side of the secondary qualities; if we cannot imagine an object losing its extension or impenetrability without ceasing to exist, then these are its primary qualities. That we in fact agree in thinking of resistance, extension, and figure as the primary qualities of matter, indeed think of matter as consisting of these attributes “together with miscellaneous powers of exciting other sensations” (214), Mill readily admits. That we group these together he explains by the fact that sensations of smell, taste, and hearing do not cohere directly, but “through the connexion which they all have, by laws of coexistence or of causation, with the sensations which are referable to the sense of touch and to the muscles; those which answer to the terms Resistance, Extension, and Figure. These, therefore, become the leading and conspicuous elements. . . .” (213.)
So the question eventually comes to that of whether the associationist psychology can explain our conception of things as being spatially extended, with the implications that this property suggests, that they must have boundaries or figure, if we are to tell one thing from another, and that they must be less than wholly interpenetrable. Resistance, or relative impenetrability, Mill explains as an inference from the experience of obstructed muscular movement when this is combined with appropriate sensations of touch. The combination assures us that the impediment to movement is not internal paralysis or something similar. Figure, Mill deals with rather casually as the conjoined information of sight and touch; he invokes a good deal of not very persuasive psychological evidence to suggest that a blind man either has a different conception of figure from that of a sighted man or no conception at all, and even toys with the less than obviously coherent claim that a blind man might think the external world was composed entirely of one object. But it is evidently the analysis of extension that is crucial to his case. He makes it at second hand by way of an extended quotation from Bain. The gist of the case is simple enough. We have certain sensations connected with the contraction of our voluntary muscles, and these are different according to the extent of such contraction, so that we can discriminate half, wholly, or very partially contracted muscles; these are associated with the sweep of a limb or other bodily movement. Now it would obviously be putting the cart before the horse if Mill and Bain were to employ the idea of a limb sweeping a certain amount of space in explaining the origins of our idea of space. Most of Mill’s critics, as we have seen, said that this was just what they had done. Whether the charge can be rebutted is very difficult to decide. In a sense, Mill is between the devil and the deep blue sea. Any notion of the sweep of a limb which is distinctively non-spatial looks inadequate to generate a conception of space at all, while any notion adequate to the generation of a concept of space seems to get there by starting with some notion of space already. If we make the sweep of a limb purely temporal—that is, if we say that the non-spatial notion is simply one of the length of time it takes for sensations to succeed each other—we escape the charge of paralogism, but we do not get very close to the usual idea of space. Mill does not make this admission; on his analysis, the blind man’s conception of space is temporal not spatial, and even the sighted majority have a conception which is basically temporal:
a person blind from birth must necessarily perceive the parts of extension—the parts of a line, of a surface, or of a solid—in conscious succession. He perceives them by passing his hand along them, if small, or by walking over them if great. The parts of extension which it is possible for him to perceive simultaneously, are only very small parts, almost the minima of extension. Hence, if the Psychological theory of the idea of extension is true, the blind metaphysician would feel very little of the difficulty which seeing metaphysicians feel, in admitting that the idea of Space is, at bottom, one of time—and that the notion of extension or distance, is that of a motion of the muscles continued for a longer or a shorter duration.
The temptation remains to say what is shown here is only that a man who has our conception of space can measure distances by the time it takes to cover them; it does nothing to suggest that time alone can convey that conception of space to one who does not have it. Just as Mill’s analysis of the external world provides us with “possibilities of sensation” external to our actual sensations only in the same way that the number six is external to the series of numbers from one to four, so here he seems to offer us extension in one dimension when we want it in another.
The point at which Mill himself admitted to defeat was in the analysis of mind rather than matter. The general line that he saw himself obliged to pursue was what we should expect; if matter was a permanent possibility of being sensed, the “Ego” should be amenable to analysis as the permanent possibility of having sensations. Mill’s first concern is to show that there is nothing in such a phenomenalism to justify charges of atheism or all-embracing scepticism. If the mind is a series of mental states, there is no bar to immortality in that: a series can go on forever just as readily as a substance can. No doubt metaphysicians have been eager to argue that we must be immortal, on the grounds that the soul, being a substance, is indestructible, but such arguments, says Mill, are so feeble that philosophers have increasingly given them up. The existence of God is equally untouched: “Supposing me to believe that the Divine Mind is simply the series of the Divine thoughts and feelings prolonged through eternity, that would be, at any rate, believing God’s existence to be as real as my own” (192). And the existence of other minds is as well vouched for on phenomenalist as on substantialist premises. We know in our own cases that between bodily effects and their bodily causes there intervene mental events—sensations, motives, and so on—and we infer inductively that the same thing is true in other cases; we see bodies like our own and believe on excellent evidence that there are minds associated with them. “I conclude that other human beings have feelings like me, because, first, they have bodies like me, which I know, in my own case, to be the antecedent condition of feeling; and because, secondly, they exhibit the acts, and other outward signs, which in my own case I know by experience to be caused by feelings” (191). Mill thus concludes that Reid’s accusation, that the phenomenalist ends as a solipsist, fails.
But this is not to say that the phenomenalist position is freed of all difficulty. The pressure in favour of phenomenalism is the same in the case of mind as in the case of matter; we have no knowledge of mind as it is in itself, only of its phenomena. Just like Hume, Mill holds that what we perceive are the mind’s modifications, such as thoughts, sensations, desires, and aversions. What we have in the way of evidence is a stream of experience; is the mind or the self more than such a stream, therefore? Mill answers that it seems that it must be more. The reason lies in the nature of memory and expectation. In themselves memories and expectations are simply part of the stream of consciousness, but their oddity is that they essentially involve beliefs, and beliefs of an awkward kind. When we expect a future experience, we expect something to happen to us, and when we remember a past experience, we remember that something happened to us.
Nor can the phænomena involved in these two states of consciousness be adequately expressed, without saying that the belief they include is, that I myself formerly had, or that I myself, and no other, shall hereafter have, the sensations remembered or expected. The fact believed is, that the sensations did actually form, or will hereafter form, part of the self-same series of states, or thread of consciousness, of which the remembrance or expectation of those sensations is the part now present. If, therefore, we speak of the Mind as a series of feelings, we are obliged to complete the statement by calling it a series of feelings which is aware of itself as past and future; and we are reduced to the alternative of believing that the Mind, or Ego, is something different from any series of feelings, or possibilities of them, or of accepting the paradox, that something which ex hypothesi is but a series of feelings, can be aware of itself as a series.
In essence, Mill’s problem is that if matter is a hypothesis that a mind formulates to account for the regularity of its experience, a unitary self must be presupposed to do the hypothesizing, and a unitary self that, furthermore, can view its experience as something regular enough to need explaining by such a hypothesis. But if my construction of my experienced world depends on a prior identification of the data of experience as my sensations and so on, there seems no hope of accounting for me in the same terms—for, out of what would I construct me? Mill insists in a long footnote that he merely intends to leave open the question of what the mind’s nature really is, neither, as some of his critics have alleged, adopting the “psychological theory” in spite of the objections, nor accepting the common view of the mind as a substance (204n-7n). Indeed, says Mill in the main text,
The truth is, that we are here face to face with that final inexplicability, at which, as Sir W. Hamilton observes, we inevitably arrive when we reach ultimate facts; and in general, one mode of stating it only appears more incomprehensible than another, because the whole of human language is accommodated to the one, and is so incongruous with the other, that it cannot be expressed in any terms which do not deny its truth
This abstemiousness about putting forward any explanation of the inexplicable did not save Mill from Bradley. In his Ethical Studies Bradley did his best to kill off the psychological theory with a famous joke: “Mr. Bain collects that the mind is a collection. Has he ever thought who collects Mr. Bain?” and went on to say of Mill that when he had “the same fact before him, which gave the lie to his whole psychological theory, he could not ignore it, he could not recognize it, he would not call it a fiction; so he put it aside as a ‘final inexplicability,’ and thought, I suppose, that by covering it with a phrase he got rid of its existence.” This judgment is transparently unjust, but there is something extremely unsatisfactory about Mill’s agnosticism all the same.
One cannot do the subject justice here, but we may at any rate agree that Mill could have done more. He could, for example, have explored the idea that the self can be a serial self, without needing a non-serial percipient self to give it unity, or that it is a logical construction which does not require a constructor; he could have pressed the “error theory” implicit in what he says about the way ordinary language favours one view of personal identity, and attempted to pull apart the implications of the language from the bare facts of the world. The fact remains that he did not.
Although there are grounds for treating Mill’s attack on Hamilton’s account of causation in conjunction with discussion of free-will—namely, that Mill discusses the “volitional” theory of causation while he is attacking Hamilton, and in the process commits himself to the view that we have no direct power over our own volitions (298-9)—there is more to be said for tackling it briefly and on its own. For on causation Mill adds nothing to his own account in the Logic, whereas on the subject of the freedom of the will he supplements what he says in the Logic, and in addition fills out the theory of punishment and the conception of justice that we find in Utilitarianism and On Liberty. His attack on Hamilton’s theory of causation is brief and dismissive. The issue was what we might expect: Hamilton appealed to the innate structure of the mind, and Mill thought the appeal quite illicit. On this topic Hamilton’s case was an odd one. For he did not appeal to a positive intuition of the connectedness of events, nor to anything like Kant’s synthetic a priori principle of the rule-governed succession of events. Rather, he appealed to an incapacity of the mind. The incapacity in question was the mind’s inability to conceive of what he called an “absolute commencement.” This incapacity, as Mill says, is on Hamilton’s account not entirely reliable as a guide to how things are, for acts of the free will are cases of just such an absolute commencement. It does seem at first, however, the sort of thing on which one might found a view of causation. That is, we cannot regard any event as an uncaused happening, because we cannot conceive of any such thing; we must, therefore, look for the cause of it. The difficulty lies in Hamilton’s explanation of the nature of the incapacity. Hamilton does not make any claim for its fundamental status. He explains it is a case of the general incapacity to imagine that there could be an increase or decrease in the quantum of existence in the world. This is, of course, a sort of relative of the principles of the conservation of energy or the conservation of matter; so read, Hamilton might be saying that the aim of causal explanation is to show how a fixed quantity of matter undergoes changes of form. The reason why he put the problem in this odd way was very probably his scholastic enthusiasm for the Aristotelian four causes, but Mill was surely right to say that the only one of the Aristotelian causes which corresponded to the modern conception of cause was the efficient cause. Hamilton went on to claim that the effect is the very same thing as the cause, presumably meaning only that effects must be made out of the same fixed quantum of matter. This was to ignore the efficient cause in favour of the material, and, in thus deciding to leave out of account the changeable element in causation, Hamilton simply left out causation. “Suppose the effect to be St. Paul’s: in assigning its causes, the will of the government, the mind of the architect, and the labour of the builders, are all cast out, for they are all transitory, and only the stones and mortar remain” (292). In any case, says Mill, it is plainly absurd to suppose that the law of the conservation of matter is an original endowment of the mind; until they are taught otherwise, men believe that when water evaporates, it is annihilated, and do not think that when wood is reduced to ashes, the missing wood must be somewhere in some shape or other, even if only as smoke. It therefore looks as if Hamilton’s interpretation of our incapacity to conceive an absolute commencement is suicidally ill-adapted to provide a theory of causation. Had he employed the principle in its most natural sense, as referring to the inconceivability of an uncaused event, it might have been bald, though it would have been addressed to the right topic; however, to employ it, not as a principle about the effects of events upon each other, but as a principle about the unchangeable quantity of existence in the world, made it simply irrelevant to the topic in hand.
Mill declines to provide a positive account of causation, on the entirely proper grounds that he has done more than enough in that line in the Logic. Instead he turns to Hamilton’s views on logic. Anyone who wearies of Mill’s hounding of Hamilton through the questions of how we form concepts, what it is to judge something to be the case, and so on, will wish that Mill had declined the chase on the grounds that here, too, he had done enough in the first two books of the Logic. The question, what is a concept, resolves itself for Mill into the familiar question whether there are any abstract ideas; he offers a thumbnail sketch of the three possible views on universals, declares that Realism is dead beyond hope of revival, and proceeds to set out the rival attractions of Nominalism and Conceptualism. The view of the nominalists was that “there is nothing general except names. A name, they said, is general, if it is applied in the same acceptation to a plurality of things; but every one of the things is individual” (302), and this is the view of the mediaeval nominalists’ successors such as Berkeley. The conceptualists, of whom Locke is representative, agree that “External objects indeed are all individual” but maintain nonetheless that “to every general name corresponds a General Notion, or Conception, called by Locke and others an Abstract Idea. General Names are the names of these Abstract Ideas.” (302.) Mill complains of Hamilton that he will not settle for one or other of these positions, but seems to swing between agreeing with Berkeley that we simply cannot form ideas of, for example, a triangle which is neither isosceles nor scalene nor equilateral—in which case he would be a nominalist—and a manner of talking about “Abstract General Notions” which is only consistent with conceptualism. Mill himself settles for nominalism, by explaining that we may have abstractions without having any abstract ideas.
General concepts, therefore, we have, properly speaking, none; we have only complex ideas of objects in the concrete: but we are able to attend exclusively to certain parts of the concrete idea: and by that exclusive attention, we enable those parts to determine exclusively the course of our thoughts as subsequently called up by association; and are in a condition to carry on a train of meditation or reasoning relating to those parts only, exactly as if we were able to conceive them separately from the rest
Attention is fixed by naming the respect in which we are to attend to whatever it is. Mill insists that words are therefore only signs, and there can be such things as natural signs; anything which will direct the attention in the appropriate way will form the basis of classification and conceptualization. “We may be tolerably certain that the things capable of satisfying hunger form a perfectly distinct class in the mind of any of the more intelligent animals; quite as much so as if they were able to use or understand the word food” (315).
Mill’s eventual aim is to vindicate against Hamilton the doctrine that there can be a logic of truth as well as a logic of consistency. In the process he sets out to criticize Hamilton’s account of what is involved in judgment and reasoning. The two basic complaints that Mill levels against Hamilton are that his account of judgment appears to make all true propositions analytic, and that his account of reasoning makes it impossible to see how one can ever find out something by reasoning. Here again we are in a much-trodden field, and one where there has since Mill’s day been a continuous effort to disengage questions of logical implication from questions about the novelty to any particular reasoner of the conclusion he reaches by deductive inference. In the matter of judgment, Mill had an interest in insisting on the importance of belief, and thus of the idea of truth. In editing his father’s Analysis of the Phenomena of the Human Mind, he had remarked on the imperfections of the associationist analysis of belief in terms of the association of two ideas. To believe that the grass is green and to deny that the grass is green, we need to have the same propositional content in mind; it is the judgment we make of its being true to fact or false to fact that is different. In so far as associating ideas is supposed to be mentally analogous to depicting a state of affairs, it leaves out what is distinctive about judging that something is or is not the case; for a picture to become an assertion or a denial it needs to have something else added to it, namely the judgment that it is or is not how things are.
Mill takes up the theme against Hamilton with additions. Hamilton had rashly suggested that judgment was a process of seeing whether one concept was part of another, though he also claimed that in judgment we looked to see if two concepts were capable of coexistence or were mutually repugnant. But this argument he glossed in such a way as to suggest at any rate that such an inspection yielded what we should normally think of as a synthetic judgment. We put together such concepts as water, rusting, and iron, and if they are congruent, reach the judgment that “water rusts iron.” Mill comments pretty sharply on this fearful muddle. It confuses judgments about the compatibility of our concepts with judgments about the coexistence of attributes in the world, and in any event does not make the necessary move from contemplating a state of affairs as possible to asserting that it is actualized.
The discussion is complicated to some degree by the psychological overtones of any discussion of concepts. Hamilton at times seems to be wanting to say that an established truth is analytic, in that our concepts embody everything we associate with that of which they are the concept; so, only new truths would be synthetic, and they would make us revise our concepts in such a way that what had been synthetic now became analytic. This cannot be said to be an attractive doctrine in general, nor can Hamilton be said to have showed much sign of really wishing to articulate it; it would mean that a statement such as “all men are mortal” would be speaker-relative both in meaning and in epistemological status. For somebody whose concept “man” included “mortal” it would be analytic, and for somebody whose concept did not, it would be synthetic. Even then, in Hamilton’s account, we are not much further forward, for if concepts are congruent when propositions are possibly true, and if they are related as part to whole when they are necessarily true, how are they related when something is said to be true only contingently? As Mill complains, the necessary reference to a belief about the world seems to have been omitted.
Take, for instance, Sir W. Hamilton’s own example of a judgment, “Water rusts iron:” and let us suppose this truth to be new to us. Is it not like a mockery to say with our author, that we know this truth by comparing “the thoughts, water, iron, and rusting?” Ought he not to have said the facts, water, iron, and rusting? and even then, is comparing the proper name for the mental operation? We do not examine whether three thoughts agree, but whether three outward facts coexist. If we lived till doomsday we should never find the proposition that water rusts iron in our concepts, if we had not first found it in the outward phænomena.
Mill’s chapter on reasoning is concerned with the problem which had haunted the Logic, that is, how can reasoning give us new knowledge? Mill requires a theory of reasoning which accounts for the way in which we can, by bringing judgments to bear on each other, learn what we could not know by inspecting them separately. The conventional complaint against Mill to the effect that he habitually confuses psychological and logical questions really does seem warranted here, for most of his objections to Hamilton boil down to the claim that if we move from “all men are mortal” via “Socrates is a man” to “Socrates is mortal” by seeing that a concept comprehended under a concept is comprehended under any concept that comprehends that second concept, then it is impossible to see how we could move from premises to conclusion. Did we once have the greater concept clear in our mind, subsequently forget part of it, and then recall it (343-5)? Mill produces what he takes to be a conclusive refutation of the “conceptualist” view that reasoning is eliciting the implications of concepts, when he offers geometrical reasoning as a plain case of achieving new knowledge of things rather than merely of concepts by a process of reasoning alone.
Here are two properties of circles. One is, that a circle is bounded by a line, every point of which is equally distant from a certain point within the circle. This attribute is connoted by the name, and is, on both theories [that is, Nominalism and Conceptualism], a part of the concept. Another property of the circle is, that the length of its circumference is to that of its diameter in the approximate ratio of 3.14159 to 1. This attribute was discovered, and is now known, as a result of reasoning. Now, is there any sense, consistent with the meaning of the terms, in which it can be said that this recondite property formed part of the concept circle, before it had been discovered by mathematicians? Even in Sir W. Hamilton’s meaning of concept, it is in nobody’s but a mathematician’s concept even now: and if we concede that mathematicians are to determine the normal concept of a circle for mankind at large, mathematicians themselves did not find the ratio of the diameter to the circumference in the concept, but put it there; and could not have done so until the long train of difficult reasoning which culminated in the discovery was complete.
This discussion, of course, ties in with Mill’s account of geometry in the Logic, with its insistence that geometry was not about definitions but about the things picked out by the definitions.
Mill goes on to criticize Hamilton’s account of logic in terms which the preceding discussion would lead us to expect. Hamilton intended, so far as one can see, to describe logic as a purely formal science, and to explain the domain of what we should now call philosophical logic as that of the analysis of the mental operations necessary for valid thinking and inference—concept formation, definition, and so on. But this is notoriously an area in which the absence of an adequate notation hindered all efforts at distinguishing clearly between formal and material considerations. Mill, moreover, was an unabashed primitivist in such matters. He complained in the Examination that Hamilton’s attempt to explicate the law of noncontradiction by such formulae as “A=not-A=0” or “A-A=0” was merely a “misapplication and perversion of algebraical symbols” (376), and his letters reveal that he had no inkling of the importance of the work of Boole. In the absence of an adequate notation, it is difficult to develop a coherent account of what is meant by restricting the notion of logic to formal considerations. Mill is wholly successful in showing that Hamilton made a fearful chaos of it. What everyone since has found less convincing is Mill’s positive account of a logic which should be wider than the logic of consistency. It is not that his fundamental position is incoherent, though it is loosely stated.
If any general theory of the sufficiency of Evidence and the legitimacy of Generalization be possible, this must be Logic κατ’ ἑξοχήν, and anything else called by the name can only be ancillary to it. For the Logic called Formal only aims at removing one of the obstacles to the attainment of truth, by preventing such mistakes as render our thoughts inconsistent with themselves or with one another: and it is of no importance whether we think consistently or not, if we think wrongly. It is only as a means to material truth, that the formal, or to speak more clearly, the conditional, validity of an operation of thought is of any value; and even that value is only negative: we have not made the smallest positive advance towards right thinking, by merely keeping ourselves consistent in what is, perhaps, systematic error.
Here, evidently, Mill divides general logic into what one might call the realm of inductive support on the one hand, and the realm of deductive implication on the other. The general principle that deductive arguments are conclusive because there is no way to affirm their premises and deny their conclusions without self-contradiction is one which Mill seems to adopt for himself. The so-called principle of non-contradiction, says Mill, “is the principle of all Reasoning, so far as reasoning can be regarded apart from objective truth or falsehood. For, abstractedly from that consideration, the only meaning of validity in reasoning is that it neither involves a contradiction, nor infers anything the denial of which would not contradict the premises.” (378.) Yet Mill does not want to draw such a sharp line between inductive and deductive arguments as either his opponents at the time or his successors now would do. The suggestion, even in the quotation immediately above, is that where objective truth or falsehood is in question, there is a sense of “validity” other than that employed in deductive reasoning. And that in turn suggests another heretical doctrine, that Mill thinks of the relation between premises and conclusions as relations of evidential support; some evidential support is so good that when we see plainly what we are saying we see that we should contradict ourselves by simultaneously asserting the premises and denying the conclusion. But instead of concluding that induction and deduction are wholly different operations, Mill inclines to the view that there is no real inference in deductive arguments.
The twentieth-century reader’s unease at all this must be a good deal increased by two passages which betoken the same unwillingness to give any weight at all to the formal/material distinction. Mill seems at first to see that there is something odd about the so-called law of identity which, he agrees, lies at the basis of all reasoning, though it is not clear what it is that he dislikes. At one point he suggests that the law of identity amounts to saying that a statement true in one form of words remains true in another form of words bearing the same meaning. To elucidate the law, says Mill, we need very much more than a statement like “A is identical with A.” We need, indeed,
a long list of such principles as these: When one thing is before another, the other is after. When one thing is after another, the other is before. When one thing is along with another, the other is along with the first. When one thing is like, or unlike, another, the other is like (or unlike) the first: in short, as many fundamental principles as there are kinds of relation. For we have need of all these changes of expression in our processes of thought and reasoning.
If the law of identity is fundamental in reasoning, it must be a general licence “to assert the same meaning in any words which will, consistently with their signification, express it” (374). This suggests that Mill does not think that identity is a property of things, but wishes to gloss it in terms of the equivalence of propositions. But he ends by admitting to some uncertainty whether the fundamental laws of logic are really necessities of thought or merely habits which we have acquired by seeing that these laws apply to all phenomena. That they do apply to phenomena, Mill certainly says here. Speaking of the laws of identity, contradiction, and excluded middle, he says,
I readily admit that these three general propositions are universally true of all phænomena. I also admit that if there are any inherent necessities of thought, these are such. I express myself in this qualified manner, because whoever is aware how artificial, modifiable, the creatures of circumstances, and alterable by circumstances, most of the supposed necessities of thought are (though real necessities to a given person at a given time), will hesitate to affirm of any such necessities that they are an original part of our mental constitution. Whether the three so-called Fundamental Laws are laws of our thoughts by the native structure of the mind, or merely because we perceive them to be universally true of observed phænomena, I will not positively decide: but they are laws of our thoughts now, and invincibly so. They may or may not be capable of alteration by experience, but the conditions of our existence deny to us the experience which would be required to alter them.
Mill’s last encounter with Hamilton on the logical front concerns two doctrines on which Hamilton very much prided himself. These are the claim that we can and should distinguish between syllogisms taken in “extension” and taken in “comprehension,” and the doctrine of the quantification of the predicate. Mill is very fierce against the first, but mostly because he thinks Hamilton failed to see that the extension of a class is no clue to the meaning of a class name. Thus the meaning of “table” is explained by the attributes in virtue of which tables are such; anyone who knows what they are knows what “table” means and what a table is. The number of things which happen to be tables is neither here nor there; to know that they are tables requires that we know the attributes of tables already, and once we know that, we know all there is to be known about the meaning of the word “table.” Whether this view entails that there is no light to be cast on the syllogism by treating it in terms of the calculus of classes is debatable. Mill follows Hamilton into a fog of visual imagery. According to Hamilton, says Mill, we should think of “all oxen ruminate” as meaning “If all creatures that ruminate were collected in a vast plain, and I were required to search the world and point out all oxen, they would all be found among the crowd on that plain, and none anywhere else. Moreover, this would have been the case in all past time, and will at any future, while the present order of nature lasts.” (387.) Mill’s objection is not that this is not implicit in the proposition, but that such a claim is not what is present to the mind. What is present to the mind is that two attributes are conjoined.
Hamilton is now best remembered for his doctrine of the quantification of the predicate. This is not to say that he is kindly remembered for it; it is little more than a curiosity of the history of logic, and Hamilton’s own version of it has been described as presented with “quite fantastic incompetence.” The most that anyone now tries to do is rescue Hamilton from such charges. It is, however, hard to see quite what Hamilton was trying to add to the traditional theory of the syllogism, the more so because his later elucidations of the doctrine, produced in the heat of controversy with De Morgan, not only diminish the claims of the doctrine in respect of the number of new forms of proposition added to the traditional square of opposition, but, as De Morgan pointed out, render invalid syllogisms he had earlier claimed as valid. Mill does not tackle Hamilton on these technical issues. Rather, he challenges him on his claim that the quantification of the predicate is a principle of mental hygiene. Hamilton appeals to “the self-evident truth,—That we can only rationally deal with what we already understand, determines the simple logical postulate,—To state explicitly what is thought implicitly.” The postulate is a fairly ludicrous piece of advice; conversation would be impossible if we said everything we thought.
The true place of the doctrine of the quantified predicate lies in the theory of the syllogism, and particularly in the area of Aristotle’s claims about the permissible and impermissible forms of proposition. Hamilton’s claim that we can quantify the predicate makes good sense in the case of affirmative propositions like “all x is y” or “some x is y,” where we can give clear meaning to “all x is some y” and “all x is all y,” and again to “some x is all y” and “some x is some y.” Even here there is trouble lurking, since “all x is all y” may be interpreted either as “every x is every y”—which is true if there is only one x, only one y and x is y—or as a class-proposition to the effect that everything in x is in y and vice versa. Hamilton plainly wanted to read it as a class proposition, and only so could it give the required meaning to what he called “parti-partial negatives” like “some x is not some y,” where he wanted to admit as possible propositions even “some A is not some A” as in “Some animal (say, rational) is not some animal (say, irrational).” Then when pressed by his critics, he added the doctrine that some meant, not some at least, but some only, and this move collapsed the particular affirmative and particular negative propositions of the traditional square of opposition into each other, so destroying the claim that with the quantified predicate we achieve eight distinct forms of proposition, which can be put into four pairs of contradictories in the usual way.
The whole subject of how to interpret the quantification of the predicate in the case of negative propositions is bedevilled by the awkwardness of the verbal formulae involved, and it is no wonder that Hamilton and De Morgan argued at cross-purposes for the better part of twenty years. However sympathetic to the quantification of the predicate one may feel, it seems clear that most of what Hamilton hoped to achieve is much more readily achieved by resorting to Euler circles. With the aid of these and the predicate calculus it is possible to spell out several versions of what is implied by Hamilton’s claims. No point which can readily be related to Hamilton’s thought is served by so doing, and, because syllogistic logic is of interest to most modern logicians for what it suggests about the capacity of mediaeval logicians to anticipate twentieth-century controversies, rather than for more directly instructive reasons, Hamilton’s muddles, late in the day, are unexciting stuff. One can say on Hamilton’s behalf that the theory of the quantification of the predicate opens up an interesting area of logic, which remained largely inaccessible until a more adequate notation was developed. The later history of the subject runs through De Morgan’s speculations about the “numerically definite” syllogism and on to twentieth-century work on “the logic of plurality.” But to all this Mill had no contribution to offer, and Hamilton rather a small one.
On the issues as he saw them Mill’s demolition of Hamilton’s claims for the doctrine is brief, lucid, and complete. He objects to Hamilton’s rewriting of some as “some only”; although Hamilton may be right that there is a sous entendu of conversation to the effect that if I have seen, and know that I have seen, all your children, I should not remark merely that I had seen some of them, this fact is no reason to clutter up the theory of the syllogism (400-1). “Some A is B” is a single judgment, says Mill, and the predicate calculus would no doubt be thought to be on his side in formalizing it as ∃x(Ax & Bx), but “some only of A is B” is a compound judgment, and here, too, the modern formula would give Mill comfort, for it would be ∃x(Ax & Bx) & ∃x(Ax & -Bx). The same doubling up is required also when we attempt to quantify the predicate in the case of universal affirmatives. So, says Mill, Hamilton is not asking us to make explicit what is already implicit, since what he says is implicit (that is, in our minds already) is nothing of the sort. The Hamiltonian rewritings merely substitute two judgments for one. Mill adds a footnote to explain that we individuate judgments by way of seeing what quaesitum we answer, and he quotes one of Hamilton’s own authorities to the effect that the “cause why the quantitative note is not usually joined with the predicate, is that there would thus be two quæsita at once; to wit, whether the predicate were affirmed of the subject, and whether it were denied of everything beside” (400n-1n). Mill’s conclusion is what one would expect:
The general result of these considerations is, that the utility of the new forms is by no means such as to compensate for the great additional complication which they introduce into the syllogistic theory; a complication which would make it at the same time difficult to learn or remember, and intolerably tiresome both in the learning and in the using. . . . The new forms have thus no practical advantage which can countervail the objection of their entire psychological irrelevancy; and the invention and acquisition of them have little value, except as one among many other feats of mental gymnastic, by which students of the science may exercise and invigorate their faculties.
Given that Hamilton’s claims had been for the psychological and theoretical merits of the doctrine, it is hard to blame Mill for not going out of his way to find a more plausible and persuasive version of the doctrine to criticize.
FREEDOM OF THE WILL
The last issue on which we shall see how Mill takes Hamilton to task is that of the freedom of the will. As we should imagine, the Philosophy of the Conditioned found the questions of how the will determined action, and how the will was itself moved (if not determined) to act, the occasion for a riot of declared nescience. Mansel, whose commitment to the unanswerability of ultimate questions was stronger than Hamilton’s, placed the question whether and in what way the will was free on the list of topics where philosophy proceeded by denying the intelligibility of the claims of reductionists, materialists, and necessitarians, rather than by defending an articulated account of the nature of the will and its free operation. But it was, if anything was, the central issue on which he proposed to stand and fight. For Mansel, the two opposing armies were those of the philosophy of Personality on the one side and those of Necessity on the other, and, although he did not do anything to defend this view of the nature of the battlefield or his own place in the ranks of the personalists in The Philosophy of the Conditioned, the opposition itself appears plainly enough almost throughout his Bampton Lectures. Mill attacks some of the obiter dicta in Mansel’s Prolegomena Logica, but in criticism he sticks pretty closely to Hamilton. However, for most readers, Mill’s positive views provide the interest of the chapter, for Mill commits himself to a number of views on punishment, the nature of justice, and the analysis of responsibility which outraged his critics at the time, and which still are live philosophical positions.
Mill says, rather plausibly, that Hamilton’s account of the freedom of the will is central to the whole Philosophy of the Conditioned. Hamilton brings the supposed incapacity of the human mind to conceive an “absolute commencement” into head-on conflict with our apparently intuitive conviction that we are free agents, whose acts of will are indeed absolute commencements. Hamilton’s Philosophy of the Conditioned, moreover, denied the teachings of common sense on the freedom of the will. Where Reid had come close to Dr. Johnson’s famous assertion that “we know our will is free, and there’s an end on’t,” Hamilton thought we knew nothing of the sort. Even Reid had agreed that people act from motives; a motive must in some fashion determine the action—even if the motive was not a direct cause of action, it was surely one of the co-operating causes which determined the will, and the will in turn was the direct cause of the action (444). Mill gratefully acknowledges Hamilton’s assistance in repudiating Reid’s common-sense position, though he does so in a somewhat barbed fashion: “Sir W. Hamilton having thus, as is often the case (and it is one of the best things he does), saved his opponents the trouble of answering his friends, his doctrine is left resting exclusively on the supports which he has himself provided for it” (445). But the freedom of the will is central to Hamilton’s metaphysics in more than providing a paradigm of the conditioned nature of thought, and in more than providing a point at which Hamilton’s distinctive views emerged clearly by contrast with those of Reid. For Hamilton’s theology rested on human freedom. In effect, he held that the existence of a non-natural origin of action was the chief ground for supposing that there was a personal Creator, rather than, say, a material First Cause or a Platonic Form, at the origin of the universe. It is not just that the human personality provides, and has to provide, the model in terms of which we imagine God to ourselves—this was the burden of Mansel’s case—it is that unless human agency is somehow outside the ordinary natural course of events, there is no reason why the universe should not be thought of as having a wholly natural origin.
Mill does not so much argue against this view, though he does do so, as complain about the wickedness of resorting to such arguments at all:
the practice of bribing the pupil to accept a metaphysical dogma, by the promise or threat that it affords the only valid argument for a foregone conclusion—however transcendently important that conclusion may be thought to be—is not only repugnant to all the rules of philosophizing, but a grave offence against the morality of philosophic enquiry
The only thing about Mill’s attack on Hamilton’s theology that is of much philosophical interest is negative. Mill does not suggest that a (really or only apparently) contracausal freedom of agency could have appeared in the world by purely natural processes. He insists instead that Hamilton’s argument for the existence of God is a poor one compared with his own favoured argument, that from design (439). And he argues against Hamilton that a necessitarian or determinist could believe in God as a First Cause with no more difficulty over the First Cause’s own origins than the libertarian had. But he does not suggest anything like the kind of theory of emergent properties which might explain the way in which a sufficient degree of, say, neurological complexity and brain capacity causes a change of kind in the determination of action without introducing supernatural causes. The fact has a certain historical interest in showing how little Mill had absorbed of the evolutionary theory which would so naturally have provided him with just such an explanation.
All this, however, is almost by the way. For Mill’s aim is to present the positive case for necessitarianism or—since he rejected the idea of any “must in the case, any necessity, other than the unconditional universality of the fact” (446)—what he preferred to call determinism. The determinist holds no more complicated a belief than that human actions are not exempt from the causality in terms of which we explain all other phenomena. He hold that “volitions do, in point of fact, follow determinate moral antecedents with the same uniformity, and (when we have sufficient knowledge of the circumstances) with the same certainty, as physical effects follow their physical causes” (446). Mill encourages us to test the belief against evidence, both individual and social, and assures the reader that it is confirmed by the predictability of people’s behaviour. Mill, like empiricists before and after him, assumes rather readily that all prediction rests upon knowledge of physical causes. There is no such thing as real unpredictability, no genuine indeterminacy in the facts; all there is is the residual ignorance of the observer. “The cases in which volitions seem too uncertain to admit of being confidently predicted, are those in which our knowledge of the influences antecedently in operation is so incomplete, that with equally imperfect data there would be the same uncertainty in the predictions of the astronomer and the chemist” (446). Such uncertainties do not induce the scientist to abandon his belief in the universal reign of causality, and they ought not to induce anything of the sort in human affairs: “we must reject equally in both cases the hypothesis of spontaneousness . . .” (446).
Hamilton had expressed uncertainty about the revelations of consciousness on the subject of free will. Mill thinks that this is proper, because the only unchallengeable deliverances of consciousness are those where there really is no room for error—whatever I now feel, I really do now feel, and cannot think I do not. But freedom is not a matter of current feeling; it is a hypothesis, namely, the hypothesis that I could have done something other than what I actually did do. As a counterfactual, its content is ex hypothesi not present to consciousness; so consciousness simply cannot tell us that we are free. Although Mill half credits Hamilton with this realization, he argues that Hamilton sometimes lapses into saying we intuit our own freedom—inconceivable though it is on his own account to do so—and argues that, more interestingly, Hamilton holds that what we intuit is not our freedom but rather our moral responsibility, in which freedom of the will is implicit. This introduction of the concept of responsibility gives Mill the opportunity to leave Hamilton’s case on one side, and to return to the argument with the Owenites which dominates the discussion of freedom and necessity in Book Six of the Logic. Mill wishes to distinguish his own, determinist doctrine from two species of Fatalism. The first is pure or Asiatic fatalism, which “holds that our actions do not depend upon our desires. Whatever our wishes may be, a superior power, or an abstract destiny, will overrule them, and compel us to act, not as we desire, but in the manner predestined.” (465.) The second doctrine is that of Owenite fatalism, or “Modified Fatalism”:
our actions are determined by our will, our will by our desires, and our desires by the joint influence of the motives presented to us and of our individual character; but that, our character having been made for us and not by us, we are not responsible for it, nor for the actions it leads to, and should in vain attempt to alter them
The doctrine Mill held against both varieties of fatalism was not fatalist, merely determinist: that
not only our conduct, but our character, is in part amenable to our will; that we can, by employing the proper means, improve our character; and that if our character is such that while it remains what it is, it necessitates us to do wrong, it will be just to apply motives which will necessitate us to strive for its improvement, and so emancipate ourselves from the other necessity
The Owenites had argued from their position of modified fatalism that it was unjust to punish people, or, which was in their eyes, though not in everyone’s, the same thing, that punishment was ineffective as a means of social control and therefore amounted to gratuitous cruelty. The reason why their views on punishment mattered to Mill in the Examination was perhaps rather different from the reason why they mattered when he was writing the Logic. In his youth, Mill had obviously been very vulnerable to the accusation that his character had been made for him, and not by him, and that he was an artefact of James Mill’s designing. The argument in the Logic is directed almost entirely to showing that we can improve our characters, that we are not the helpless slaves of antecedent circumstances, and can choose to become something other than we have so far been brought up to be. The discussion in the Examination is less passionate. It takes off from the fact that, on Mill’s analysis, the idea of responsibility is wholly bound up with the idea of punishment. To show that there is an analysis of responsibility consistent with determinism is, in effect, to show that there is such a thing as just punishment in a determinist world.
Mill accepts that it is unjust to punish people for what they cannot help, or when they could not have acted otherwise than they did. But his analysis of what we mean when we say that a person could have acted otherwise rephrases the statement, in the classical empiricist mould, as a claim that the person would have acted otherwise if he or she had so chosen. That all else could have remained unchanged, and that the person in question should have acted differently, is what Mill denies. When Mansel says that we know that we could have acted differently, even if everything else had been the same, Mill agrees, “though the antecedent phænomena remain the same: but not if my judgment of the antecedent phænomena remains the same. If my conduct changes, either the external inducements or my estimate of them must have changed.” (448n.) We cannot act against our strongest motive, so freedom must consist in being able to act according to it. Mill goes on to claim that this kind of freedom is entirely consistent with determinism—as it evidently is—and that it is entirely consistent with holding ourselves and others responsible for their actions. Mill begins by insisting that “Responsibility means punishment” (454). He distinguishes at once between two different ways in which we may be said to be liable to punishment.
When we are said to have the feeling of being morally responsible for our actions, the idea of being punished for them is uppermost in the speaker’s mind. But the feeling of liability to punishment is of two kinds. It may mean, expectation that if we act in a certain manner, punishment will actually be inflicted upon us, by our fellow creatures or by a Supreme Power. Or it may only mean, knowing that we shall deserve that infliction.
Mill sees that it is the idea of deserving punishment which needs explaining. Expecting to suffer is very obviously consistent with a complete absence of free will.
Mill, in essence, provides a naturalistic theory of punishment. If a society has some sense of right and wrong, then those who cultivate anti-social dispositions, and threaten the security and well-being of everyone else, will naturally be thought to be behaving wrongly, and will be objects of fear and dislike to everyone else. They will therefore be left out of the distribution of common benefits and will have whatever measures of self-defence others think necessary employed against them. The wrongdoer
is certain to be made accountable, at least to his fellow creatures, through the normal action of their natural sentiments. And it is well worth consideration, whether the practical expectation of being thus called to account, has not a great deal to do with the internal feeling of being accountable; a feeling, assuredly, which is seldom found existing in any strength in the absence of that practical expectation.
Now it is noticeable here that Mill introduces a consideration which haunts the subsequent discussion of punishment much as, with its contractual overtones, it haunts Mill’s account of justice in Utilitarianism and much as it haunts On Liberty. This is the suggestion that society is founded on some sort of implicit agreement about the reciprocity of good and evil; we get security against the attacks of others in return for our forbearance, and we are punished when we break this agreement. Being practically held to account is a way of having the reciprocal nature of social agreement brought home to us. People who never enter into egalitarian relations cease to have notions like “fair play” in their moral lexicon. The importance of some such conception of justice as fairness is not much developed anywhere in Mill’s work, though it emerges in Mill’s interpretation of what utility requires. Here it emerges in what he says about the retributive element in punishment, and in a rather Kantian interpretation of the connection between punishment and the good of the criminal himself.
The main aim of Mill’s account, however, is to show how punishment is not shown to be unjust on determinist interpretations of it. After arguing, rather neatly, that even if we believed that the “criminal” class consisted of creatures who had no control at all over their noxious behaviour we should endeavour to control them by measures very like what we now call punishment, he confronts head on the opponent who says that all this is beside the point. The root of the difficulty is a question of justice: “On the theory of Necessity (we are told) a man cannot help acting as he does; and it cannot be just that he should be punished for what he cannot help” (458). Mill’s first response to this is at least odd, at worst catastrophic. He says that the claim that the criminal could not help it needs qualification; if he is of vicious temperament, the criminal cannot help committing the crime, but if “the impression is strong in his mind that a heavy punishment will follow, he can, and in most cases does, help it” (458). On this view the threat of punishment is a countervailing motive, which so to speak pushes the criminal in the opposite direction to that in which his criminal character pushes him. Mill’s critics all saw that there was something very wrong here, but nobody seems to have pointed out that, on Mill’s analysis, anyone who commits a crime can always make precisely the claim that Mill is trying to rebut. If he cannot help doing wrong when he is not threatened, the proper conclusion to draw is that when he is threatened and still offends, those who have threatened him have not done so effectively. If he could not help it, unthreatened, how can he help it, inadequately threatened?
Mill’s great concern to show that we are responsible for our characters may be thought to indicate some awareness of the trouble he had caused himself. The criminal who explains to the court that it is unfortunate that he has such a bad character, but that once he had it, it overwhelmed all the threats the law was prepared to utter, could be told that he had no more business going around with a bad character than he would have had going around with a loaded revolver. The retort, however, will not do much to save Mill’s case. Anyone who is faced with that argument can simply respond by saying that without a sufficient motive to improve his character he could not improve it; given the initial badness of his character, it was no use looking to any internal motive for change; and as for the absence of an external motive, how could he be blamed for that? Mill, indeed, does not linger on the question of the agent’s motives. He turns rather to the question of what makes punishment just. In explaining this, he gives hostages both to fortune and to Kant. Punishment has two proper goals, the good of the criminal and the defence of the just rights of others. If punishment is not inflicted to protect the just rights of others, it is mere aggression on the individual punished. But, many of Mill’s readers might wonder, how can he argue that a proper purpose of punishment is to do the offender good? Is not On Liberty devoted to denouncing precisely such a claim? And when Mill says: “To punish him for his own good, provided the inflictor has any proper title to constitute himself a judge, is no more unjust than to administer medicine” (458)—is this not in flat contradiction to his attacking Whewell for suggesting that the law on quarantine was for the sufferer’s own good? Mill responds to this charge in a long footnote. He seems to see only part of the point, for he begins by saying that of course we punish children for their own good, and we may treat “adult communities which are still in the infantine stage of development” in the same way; but he seems to draw back a little over adult offenders. “And did I say, or did any one ever say, that when, for the protection of society, we punish those who have done injury to society, the reformation of the offenders is not one of the ends to be aimed at, in the kind and mode, at least, of the punishment?” (459n.) There is here, perhaps, a suggestion to the effect that Mill accepts Kant’s view that nobody can be punished simply to do him good, but that once he forfeits his right to immunity from all punishment, we may properly consider how to reform him when we consider what punishment to inflict.
The same awkwardness emerges when Mill talks of the legitimate defence of our just rights as a ground of punishment. Looked at from society’s point of view, it is just to punish offenders who transgress the rights of others, “as it is just to put a wild beast to death (without unnecessary suffering) for the same object” (460). To say this seems precisely to ignore the whole question of the distinction between punishment applied to free moral agents and mere measures of social control applied to non-human creatures. But then Mill moves on to the question of whether the criminal can complain of being treated unjustly, and says that the crucial element in holding ourselves responsible for our actions lies in our recognizing that other people have rights. Doing so is, in essence, placing ourselves at their point of view, and if we do so we shall see that there is no injustice in their defending themselves against any disposition on our part to infringe those rights. Once again, the importance of equality emerges in the observation that we shall more readily recognize the justice of their defending their rights by punishing offences against them, the more often we have ourselves stood up for our own rights in this way. Something much nearer an appeal to fairness than to simple utility is evidently at stake.
Thereafter, Mill’s account is very like Hume’s or, indeed, one may say, like most empiricist accounts. Mere retribution is of no value, and would amount to gratuitous cruelty; something like retribution is warranted, as a way of satisfying the natural hostility and outrage which criminal acts arouse in us, but such a justification is instrumental, a case of means-ends argument, and not an appeal with arithmetical overtones to fitness or to an eternal justice. The means-ends arguments for punishment reinforce the determinists’ case, for it would evidently be both silly and cruel to inflict punishment where it could not modify behaviour, or to threaten it where it could not do so in prospect. Mill appeals to the same considerations to explain why we should punish only the guilty. If we are aiming to deter people from committing crimes, there is no point in punishing those who have not committed crimes, since there is then no basis for an association of ideas between the crime on the one hand and the punishment on the other.
It goes without saying that Mill raises all sorts of issues that have not been tackled here. The general implausibility of his analysis of responsibility has been argued at length in various other places, and almost every point he makes about motivation, about the justification of punishment, and about the compatibility of freedom and determinism has been the subject of exhaustive, but still quite unexhausted controversy for the past hundred years. A review of these arguments is not necessary here. Two negative points will suffice. It is worthy of notice that Mill does not seem to see that his opponents are groping, even if only dimly, towards the crucial point that what we call punishment is very far from being a means of social control of an obviously utilitarian kind. Why, for example, do we not endeavour to remodel the characters of those who have not yet offended, but who are likely to? Why do we not set penalties for offences for maximum deterrence at minimum cost? So effective would capital punishment be if threatened for parking offences that it is doubtful if more than one or two persons a year would be executed in the whole United States, yet the idea seems absurd. Mill has nothing to say about this issue, perhaps because he takes for granted constraints on the utilitarian calculus which are of rather doubtfully utilitarian origin. Secondly, it is worth noticing that the two places where the Examination is at its most interesting and least persuasive are where Mill discusses personal identity and where he analyzes individual responsibility. The reason is easy enough to point to, and extremely hard to explicate. In essence, Mill’s epistemology requires us to treat our own selves and our own behaviour as if they are external objects and the behaviour of external objects. We can, of course, treat other persons in this “external” or third-person fashion; we can treat some parts of our past in this way, and, up to a point, our own distant futures. The wholesale assimilation of the first-person and third-person view of the world looks much more problematic. If it is essentially an incoherent project, we should expect the incoherence to appear just where it does in the Examination, that is, when our view of our own identity is being assimilated to our view of the identity of other persons and objects, and when our control over our own activity is being assimilated to the control we may exercise over things and over other persons. If readers of the Examination are unlikely to find it quite such an exemplary work of empiricist self-criticism as Hume’s Treatise of Human Nature, it will, at least in these respects, stand the comparison.