Related Links in the Library:
Source: Editor's introduction to The Collected Works
of John Stuart Mill, Volume XI - Essays on Philosophy and the Classics, ed. John M. Robson, Introduction
by F.E. Sparshott (Toronto: University of Toronto Press, London: Routledge
and Kegan Paul, 1978).
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 F. E. SPARSHOTT
compilers of collected works must garner the rough with the smooth, and a volume that consists largely of book reviews must be expected to hold much that has turned to no such aureate earth as, buried once, men want dug up again. But Mill did not share this expectation. Such zombies as his massive reviews of Grote were called up to walk the pages of Dissertations and Discussions. The quality of the works reviewed here tends to justify the disinterment: of them all, perhaps only those of William Smith and Gustav Wiggers have quite dropped out of scholarly sight. Grote’s Greece and Plato, Fraser’s Berkeley, and Whately’s Logic hold their places on our shelves, though we leave them there to gather undisturbed their kindred dust.
I propose to introduce this gallimaufry disjointedly, saying a little about the background of each component in turn, though not quite in the chronological order, spanning more than four decades, in which they are printed here. It will be seen that some themes recur; but it seems idle to pretend to impose a systematic order on these mostly occasional pieces.
WHATELY AND FORMAL LOGIC
of all the writings reprinted here, the review of Whately has attracted most attention from commentators, both for its intrinsic interest and as a forerunner of the System of Logic. Alexander Bain called it “a landmark not merely in the history of [Mill’s] own mind, but in the history of logic.” Yet Mill himself exempted it from the general resurrection in Dissertations and Discussions. Why? Partly, no doubt, because it was superseded by the System of Logic: of the works in this volume that Mill did reprint, none falls within the scope of a later treatise. Whately was only one of a series of logicians whose work Mill discussed in 1827 with that “Society of Students of Mental Philosophy” which had begun to meet at the Grotes’ in Threadneedle Street two years before; and it was from those discussions that the opening books of the Logic began. But the Whately review was not merely superseded; in the one place in the Logic where he cites this earlier work, Mill describes it as “containing some opinions which I no longer entertain,” And Kubitz suggests that the main reason for not reprinting the article was that he had recanted its views on the significance of deductive method. The scope of this recantation, which went with a reversal in his views on the possibility of an inductive logic, will occupy us shortly. But the decisive factor could have been one that had little to do with any shift in doctrine. For Mill, perhaps more than for most reviewers even in that polemical age, a review was a political act, serving to encourage or chasten the righteous and to dismay the adversary. In 1831 Whately was “one of the fittest men in the country to hold a high station in a national church such as I conceive it should be”; but by the time the first two volumes of Dissertations and Discussions were published in 1859, though Bailey, Grote, and Bain were still around to be admonished and cheered, Whately as Archbishop of Dublin had long confined his activities to spheres where reviewers could neither help nor harm.
Mill’s failure to reprint the Whately review, however explained, must be regretted by his posthumous friends. His reputation as a logician has suffered among the philosophical laity because the doctrine of the syllogism developed in the Logic has been taken for a general theory of syllogistic logic as such. Mill’s belated care in distinguishing the logic of truth from the logic of consistency, and his insistence that only the former concerns him, has not compensated for his failure to provide a coherent exposition of the latter; in fact, he seems to blend an exposition of syllogistic in terms of consistency with a justification in terms of truth. But the Whately review makes it clear that Mill understood the nature and value of formal logic as a study of the form of valid arguments and a device for testing them. In fact, his vindication of this study against its recent neglect is couched in terms rather like those used by careless readers nowadays against his own Logic. But though this vindication saves Mill’s popular credit by giving meaning to the provisos with which the doctrine of the Logic is hedged, it does not explain that doctrine itself. If syllogism is proper to the analysis of proofs, why should it figure at all in an account of discovery? An examination of Whately’s book yields a possible answer.
Mill’s polemic (5-6 below) against those who supposed there could be a separate inductive logic (as opposed to procedural rules for inductions) is a reflection of Whately’s own arguments. Syllogism, he urged, is the unique form of valid argument. Therefore induction, in so far as it is a form of argument, must be syllogistic; in so far as it is not syllogistic it cannot be a form of argument at all, but a mere process of inquiry that as such must fall outside the scope of logic. An inductive argument is nothing but a syllogism in barbara with the suppressed major premise: “What belongs to the individual or individuals we have examined, belongs to the whole class under which they come.” So now we have two syllogisms, an inductive one:
- Mortality is a property of Socrates and Coriscus
- All properties of Socrates and Coriscus are properties of all men
- Mortality is a property of all men;
and a deductive one:
- All men are mortal
- The Iron Duke is a man
- The Iron Duke is mortal.
But the deductive one merely carries out a decision, or exemplifies a commitment, we made in the inductive one—and the decision is the implausibly sweeping one that whatever is true of the men we know is probably true of all men (and hence of Wellington).
The position implied by this move of Whately’s, and apparently endorsed by Mill, is a very strange one. Inductive reasoning is subsumed under the logic of consistency, and the consistency required is that of abiding by the commitment made in the extravagant major premise of the inductive syllogism. But all this syllogistic machinery is quite useless: if in any case we are going to start with Coriscus and his friends, and end up with Arthur Wellesley, there is no point in making a detour through “all men.” Accordingly, Mill was to write in the Logic (CW, VII, 162) that in “Reasoning or Inference properly so called,” “We set out from known truths, to arrive at others really distinct from them.” Small wonder, then, that he did not reprint the Whately review, in which he proclaimed the impossibility of the condition on which the Logic was to be constructed, and excluded from the proper sphere of logic the whole of “Reasoning or Inference properly so called”!
But it was a serious and genuine impasse that confronted Whately and the young Mill alike. Whately observes: “The justly celebrated author of the Philosophy of Rhetoric, and many others, have objected to the syllogism altogether, as necessarily involving a petitio principii; an objection which, of course, he would not have been disposed to bring forward, had he perceived that, whether well or ill-founded, it lies against all arguments whatever.” Well, perhaps; but the consequence—with which Whately in the ensuing pages vainly grapples—is either to expel reasoning from the province of discovery or to reduce discovery to the rearrangement of terms. It was precisely for freeing logicians from this impasse that the doctrine of the syllogism in Mill’s Logic was to be praised by Whewell. Mill’s final solution, in which all real reasoning is from particulars to particulars, amounts to saying that the appearance of syllogism in the processes of reasoning is only an appearance. What looks like a major premise (“All men are mortal”) is really no such thing: it is only a sort of aide-mémoire, serving two purposes. It reminds us that a number of objects have been examined and found both human and mortal, and registers a decision to let these examined cases stand as sufficient evidence for the inference that, when anything else is found to be human, it can be expected to be perishable. If syllogisms are what formal logic studies, this is not syllogism but pseudo-syllogism, for the warrant for an inference is not the same as a premise in an argument. Unfortunately, Mill had not emancipated himself sufficiently from Whately and company to make this distinction clear. William Kneale seems to be right in tracing the difficulties in Mill’s account to his “failure to realize the incompatibility of a good new insight with a bad old tradition in which he had been educated.” But Kneale seems not to have attended to quite the relevant parts of the bad tradition—his book does not mention Whately at all—and allows himself to be baffled by Mill’s contention that the major premise of a syllogism can serve two purposes: a contention which, we have seen, becomes intelligible when it is seen as the solution to a problem posed by Whately and his peers.
The story I have now unfolded is not the whole story. Mill had begun studying logic ten years before, at the age of twelve, not with Aldrich and Whately’s other predecessors as text-book writers, but with Aristotle’s Organon, accompanied by the scholastics whom he extols in his review, and followed by Hobbes. And it could have been from Aristotle that he learned how different a syllogism in investigation could be from a syllogism in analysis. But he shows no sign of having noted what Aristotle has to say about syllogisms in investigation. To turn from Whately to the Posterior Analytics is to enter a different and saner world, in which the conclusion of a scientific syllogism is not a proposition like “The Duke of Wellington is mortal” but one like “The moon suffers eclipse,” and the inquiry which it concludes does not take the form of discovering classes to which its subject belongs but that of discovering causal relations in which it is involved. The “discovery of middle terms” is not the unfolding of a system of class-inclusions (Wellington is a Duke, is a Briton, is a man, is a mammal, is an animal), but a reference to “the failure of light through the earth’s shutting it out,” involving the discovery that the moon is a body shining by reflected light and the means of that light’s occlusion. And the conclusion is not so much “Thus we may infer that the moon will undergo eclipse” as “So that explains why it is that the moon undergoes eclipse.” The eclipse of the moon, whose occurrence is affirmed in the conclusion, is neither datum nor discovery, but problem. Science is conceived not as observing and classifying individuals, but as probing the workings of systems and mechanisms. Kneale, very reasonably, asks why Mill even ignores the possibility that a major premise might state a connection of attributes rather than record a summary of cases. But apparently he does ignore it. This whole side of Aristotle’s logic must have seemed to him meaningless or hopelessly archaic. Why? Part of the reason appears in what he says in the Logic about propositions: “The first glance at a proposition shows that it is formed by putting together two names” (CW, VII, 21). It is true that syllogistic logic relies on the supposed reducibility of any proposition to the copulation of two terms, but that is a far cry from asserting that every proposition is evidently composed of two names. Why “names”? Why two? In what sense “putting together”? This unintelligible assertion harks back to the theory of language attempted by Hobbes; in Mill’s time it must have seemed very antiquated indeed. But Mill, like many revolutionaries—the men of Thermidor saw themselves as ancient Romans; Mazzini slept with Tacitus under his pillow—was in some things very old-fashioned, using the far past as a lever to unseat the near past, as his passionate Graecophilia and the defiant championing of the school logic sufficiently attest. After all, what made his father the apostle of progress in psychology was his revival of Hartley against the new-fangled Germanism. To this defiant antiquarianism belongs the Locke-like atomism of the doctrine of propositions, with the analogous reductivism that makes all reasoning go from particulars to particulars and also, in the controversy with Bailey, the inability to come to terms with any treatment of the facts of vision that does not reduce them to the association of simple percepts. A recent book argues persuasively that the whole of Mill’s philosophical activity is designed to subsume all subject matters under a single method, analysing them into components that retain their identity and are linked (like “names” in a proposition) in a merely mechanical unity by relations of addition and subtraction. This claim is so far true that, as we shall see later, Mill uses the analogy of chemical combination, in which compounds have properties not derivable from those of their admitted elements, to justify his insistence that such an analysis shall be deemed performable even in cases where it cannot in fact be carried through. If this was indeed his ambition, it seems one more proper to the seventeenth century than to the nineteenth.
The retention of an appearance of duplication in the logics of truth and of consistency, with what looks like the same syllogistic form prevailing in both, is more than merely a hangover from Whately’s theory of induction. Something like it seems to be required by the contention that logic is an art as well as a science. This is a point on which Mill endorses Whately’s position against Hamilton; and in Whately it may well reflect the systematic preoccupations of the compilers of the Encyclopædia Metropolitana, who placed his article in Part I among the pure sciences, separated by a great distance from the arts and applied sciences of Part IV among which Whately conceded that his contemporaries might have expected to find it. But of what is logic the art?
Logic . . . may be considered as the Science, and also as the Art, of reasoning. It investigates the principles on which argumentation is conducted, and furnishes rules to secure the mind from error in its deductions. Its most appropriate office, however, is that of instituting an analysis of the process of the mind in Reasoning; and in this point of view it is, as has been stated, strictly a Science: while, considered in reference to the practical rules above mentioned, it may be called the Art of reasoning.
Mill’s review (8-9) contains the materials for a different account, according to which the science of logic would lie in the analysis of the principles whereby arguments are determined to be valid, and the art would lie in the use of these principles to evaluate arguments; but in the end, while repudiating the “commonsense” tradition that would equate logic with a sort of mental hygiene, he slips like Whately into making the art of logic the art of reasoning. But if that is what logic is, and if true reasoning must lead to new knowledge, and if the rules of an art and the doctrines of the corresponding science differ only in that the art determines the end to which the science establishes the means, and yet the syllogism (on which the science of logic depends) can never lead to new knowledge, we are landed in an impasse from which only desperate measures could free us. Such a desperate measure might be Mill’s device whereby the logic of truth is constructed on the basis of something that looks like syllogism but really is not.
Syllogistic logic, though not so formulated by its inventor, only works smoothly if stated in terms of class-membership and class-inclusion. It is thus especially suited to Mill’s preferred nominalistic metaphysics of juxtaposed particulars. Whately, who preferred to speak in terms of “essences,” is rightly censured by Mill for the appalling mess he made of the doctrine of predicables. As Mill says (3), he was better talking about logic than expounding logic itself, and posterity has assented to Mill’s later judgment that his distinction lay not in any contribution to logical theory, but in doing “more than any other person to restore this study to the rank from which it had fallen in the estimation of the cultivated class in our own country.” His qualities of mind were better shown in training his spaniel to climb a tree before an admiring audience in Christ Church Meadows, and to dive thence into the Cherwell, than in ordering coherently the relations of term, proposition, and argument. In fact, as Mill hints, the offending exposition of the predicables was cribbed word-for-word from the Latin of the wretched Aldrich. Mill’s own account of the matter in his review, confessedly a restatement of traditional doctrine in terms of a purified nominalism, is admirably clear, precise, and consistent.
The saving nominalism that brought light to the murk of the predicables is also responsible for the major positive contribution of Mill’s review, his scouting of Whately’s familiar and superficially plausible but ultimately unworkable distinction between nominal and real definitions (27-8). All definitions, he says, define verbal expressions, but some are and some are not accompanied by the claim that the defined term stands for an existent. This laying of an ancient ghost, the only passage retained and quoted in the Logic, was historically of decisive importance. Even the Archbishop saw the light; by the eighth edition, scarcely a trace of the offending doctrine is allowed to remain. By this time, too, Whately is apologizing for and virtually dissociating himself from the doctrine of predicables, in terms that can be explained only by supposing that he had studied Mill’s animadversions and appreciated their justice.
Whately’s treatment of fallacies departs from his forerunners’ and accords with modern pedagogical practice in seeking to relate logical analysis to the kinds of arguments and subject matters encountered in the world at large. But Mill’s praise of this treatment as abounding with “apt examples and illustrations drawn from almost all the most interesting subjects in the range of human knowledge” (29-30) might mislead a modern reader. As a man of the cloth. Whately takes a third of his examples—43 out of 119, on a rough count—from theological controversy. This may be more than personal predilection. In theology, where observation can do so little, one has to reason a priori, and the place of deduction is assured. In the sciences, where reality keeps creeping in, reasoning a posteriori can hardly be resisted; and the total lack of interest in theology, so untypical of his place and time, that Mill avowed in later life may go far to explain why he never wrote the handbook on traditional formal logic that at the age of twenty-one he showed himself so well equipped to compile.
Mill’s long and powerful vindication of logical analysis (5-14), with which nothing in the Logic is at variance, is followed by an assertion with which the Logic seems more at odds: “The province of reasoning in the investigation of truth is immense.” For by “reasoning” here he means that strictly deductive argumentation which he was later to stigmatize as mere verbal rearrangement and hence not worthy of the name of reasoning at all. This deductive element, described as playing a dominant part in every science but chemistry and physiology (14), is in fact mathematics, which one must admit does not look much like that reasoning based on the dictum de omni et nullo to which Mill’s argument would require its reduction. Alan Ryan ascribes this vindication of deductive reasoning in part to a desire to show the importance of systematic reasoning in economics, in part (following the lead of the Autobiography, 94-7) to a wish to champion his father’s aprioristic Essay on Government against Macaulay’s empiricist critique; but probably the main reason why Mill says it is that it is plainly true. But it does pose a problem, to which Mill felt he lacked the solution. Since geometry deduces unexpected conclusions from its axioms and definitions (33), deduction must be able to serve as a heuristic device. But how can it be so? What is deduced from premises must be contained in them, so that what one seems to discover must be what one in some sense already knew without knowing it. The 1828 solution to this difficulty, which is that one might have failed to put two and two together, seemed to leave “a mist still hanging over the subject” (Autobiography, 109); it was in 1830-31, the Autobiography tells us, that reflection on Dugald Stewart led him to the realization, expressed in the Logic, that in scientific reasoning it is the general propositions themselves that are the heuristic devices, the nerve of the reasoning lying always in the progression from particular cases to particular cases. With this fateful step comes a repudiation of the pure nominalism adumbrated in 1828: class-membership and class-inclusion as the basic relations in syllogism are rejected in favour of the transitive relation of being-a-mark-of, and geometry becomes in effect the methodology of engineering. But, of course, the transitivity of being-a-mark-of is useful in investigation only if there are real kinds in nature. As we remarked in discussing the predicables, formal logic must treat classifications as arbitrary; but arbitrary classifications are heuristically null. It follows that a logic of investigation must repudiate formal logic. And that is what the Logic does.
If contemporary notions of formal logic are correct, Mill had a juster and clearer view of the matter in 1828 than appears in the Logic. The young whippersnapper was justified in the arrogance he showed in his censure of a rival critic: “A good critic on Whately should have laid down as a standard of comparison, the best existing or the best conceivable exposition of the science, & examined how far Whately’s book possesses the properties which should belong to that.” Mill clearly implies that his own recently-published review had shown him able to do what George Bentham had failed to do. And he was right. But it is a mistake to think of the Logic as a giant stride in the wrong direction. The Logic had to turn its back on the logic of consistency in order to devote itself to the logic of truth. Even the notorious account of mathematics becomes less scandalous if one sees it as sketching a mathematics of truth rather than of consistency. The issues remain vexed to this day. Jevons’s much-quoted remark that “Mill’s mind was essentially illogical” was at best a half-truth. He had the talent but lacked the will. At twenty-one he was already in a position to expound, organize, restore, and clarify the traditional formal logic, and intended to do so. A year or two later he decided to do something else instead.
Mill lived to see the beginnings of the great revival of formal logic that has marked so deeply the face of philosophy in the last century. In this revival he took no part, partly for the reasons we have seen and partly no doubt because formal logic is a young man’s game. What is more surprising is that he did not approve of it. It was too complicated. Logic was a necessary art, and therefore should be plain and simple, as Whately’s had been. The elaboration of formal calculi was a distraction from the serious business of the mind. But that, after all, is the sort of thing elderly savants usually say about what the bright young men are doing.
THE PLATO VERSIONS
mill “was, quite as much as Grote, a Greece-intoxicated man”; and, unlike the historian, had twice tasted the intoxicant himself, travelling the country from end to end. Two aspects of the Greek past he found especially heady: Athenian democracy and Platonic philosophy. His first public testimony to the latter infatuation was the series of “Notes on Some of the More Popular Dialogues of Plato” published in the Monthly Repository in 1834-35. The origin of these notes—ill-named, since they are in fact translations with relatively minor omissions and comments —is obscure. He tells Carlyle they were written “long ago,” and says in his Autobiography that they were written “several years earlier” than their publication. But when and why were they written? The paper on which the unpublished ones were written is almost all watermarked 1828 (a few sheets being 1825), which provides a terminus post quem at least for these copies. Richard Garnett, who had access to W. J. Fox’s correspondence as editor of the Repository, seems to date them to 1830, but this appears to be a mistake. Packe (136) puts them “probably after his depression of 1826, when he was attempting to rescue Greek humanism and the Socratic method of analysis from the ruins of his father’s teaching,” but this is only a guess, and if they played an important part in his therapy it is strange that a fact so germane to the theme of his Autobiography should not be mentioned there. He says (42) they were made for his own satisfaction; but why? An accomplished Greek scholar, as Mill was, might make a translation to settle the meaning of a text in his own mind; but even those that were never printed are written in a manner that suggests an intended public other than the writer himself. Mill’s complaint (39-40) about the contemporary state of Platonic studies in England would justify the publication of all nine versions in book form, but if that was intended one does not see why it was not done—there was no improvement in the situation to make the need less pressing. But if they were meant neither for the public nor for Mill himself, for whom? Possibly for his siblings, for whose education he maintained a lively concern. The only other likely person who might profit from such reading and evoke the labour of preparing it would be the fascinating and brilliant but imperfectly educated Mrs. Taylor, whom Mill met in 1830, the date putatively assigned to the translations by Garnett. But Mill says they were already old when she saw them.
Why were these nine dialogues selected for translation? There need be no answer, for Mill may have meant to do more, but he suggests one (adapted from Schleiermacher): they are those dialogues of manageable length in which we observe in action “the service rendered to philosophy by Socrates” in advancing the methodology of the moral sciences (41). Socrates appears in them not as teacher but as debater, or (in the Apology) as champion of his methods in debate, and concerned with political and moral questions rather than with the natural sciences. Such a selection would suit Mill’s lifelong preoccupation, already clearly marked, with the need for a methodical science and philosophy of practice. Like his praise of the unfashionable scholastics in the Whately review, his rescue of the unfashionable sceptic from the fashionable dogmatist and dreamer in Plato belongs to a campaign to resurrect the methodical and empirical side of all western thought. That this was the basis of selection is confirmed by Mill’s practice of omitting or summarizing those passages in which Plato forsakes the presentation of argument for the description of action, and faithfully rendering all the logic-chopping. If I had begun a series of translations thus motivated I would have begun with the Meno and would have done the Crito before the Laches; but omissions mean nothing.
A different ground of selection, however, gives an even closer fit. Mill’s title refers to “Some of the More Popular Dialogues of Plato.” This gives us pause. In what sense can the Parmenides have been “more popular” than, say, the Symposium? The only plausible answer is to be found in Schleiermacher’s attempt to establish both a canon and a systematic order (supposed, with dire effects on the Platonic scholarship of the following decades, to coincide with the order of composition) for Plato’s output. Schleiermacher divided the corpus into three groups, each with an appendix of minor works. With one exception, the dialogues translated by Mill constitute the whole of Schleiermacher’s first, or “elementary” group, plus the Apology, the first dialogue in its appendix. The one exception, the Gorgias, is the first of the second (“preparatory”) group, but its insistence on the distinction between art and mere practice makes it so central to Mill’s concerns as to explain its being taken out of order.
Whatever its source, Mill’s choice of dialogues was probably not based on a direct study of Schleiermacher’s Introduction. Though Mill claims to have learned German around 1825, there is little evidence that he often exercised this skill, and he never cites Schleiermacher otherwise than in Thirlwall’s translations. The introductory note to the Protagoras attributes to Schleiermacher the view that the value of the Socratic dialogues lies in their method of inquiry, and not in any results that the discussion may reach or (more typically) fail to reach (41; compare the note on the Phædrus, 62). But this note, like the corresponding notes to the other dialogues, must have been added to the original translation at the time of publication. For, at the time when they were first written, Mill was oblivious to this methodological possibility. The unpublished versions of the Charmides, Laches, Euthyphro, and Lysis all end with a dismissive remark to the effect that because the dialogue ends with no Q.E.D. it is to be considered as “a mere dialectical exercise” (italics added).
If the translations were made or exhumed with a view to Mrs. Taylor’s edification, their appearance in the Monthly Repository needs no further explanation: Mill’s connection with the magazine, to which he began to contribute in 1832, came about through Harriet’s membership in the congregation and intellectual circle of its editor, W. J. Fox. Nor is it hard to explain the discontinuation of the series, despite its favourable reception: by the summer of 1835 Fox was losing interest in the Repository, and at the same time Mill was becoming more involved with the London Review (the first number appeared in July of that year), the burden of which being made heavier for him by his father’s failing health. In fact, the Plato versions are almost the last things he contributed to the Repository —soon the boot was on the other foot and he was soliciting Fox for contributions. Nor, again, does the order of presentation raise any problem: those left to the last and ultimately excluded are (besides the forbidding Parmenides) those devoted to particular areas of conduct and hence contributing least directly to Mill’s methodological concerns.
Mill’s strictures on the condition of Platonic studies in England at this time have been more often quoted than evaluated, but they appear to be just. Schleiermacher’s translation inaugurated the critical study of Plato, which by this time was in full swing in Germany. Yet the English works mentioned by Mill are virtually all there had been since 1750. Nor did matters improve much. Except for Wayte’s Protagoras (1854), there were no serious English contributions to Platonic studies until the sixties, when a stream of editions and commentaries began to flow that has not yet dried up. Grote’s Plato is in fact one of the first fruits of this revival. For the complete translation desiderated by Mill the English had to wait for Benjamin Jowett, who finally (in Lewis Campbell’s phrase) “succeeded in making Plato an English classic” in 1871. This English Plato was the Plato of the Republic (of which Jowett began to work on his never-to-be-completed edition in 1856), not of the Protagoras; the élitist, not the methodologist. But no doubt this Plato also would have been welcome at India House.
Among the most striking themes in the introductory notes that Mill provides is the defence of the sophists against the strictures of the “Tory perverters of Grecian history” in the Quarterly Review. Mill’s characterization of Plato’s attitude to Protagoras seems eminently just; he is indeed treated as a respectable inquirer with an imperfect technique, not as a disreputable agitator. But the issue is joined on political grounds: the sophists, like the Benthamites, believed in seeking rational solutions to moral and political problems; their conservative opponents believed, like good tory squires, that what was good enough before the war was good enough now, and regarded the sophists as dangerous and subversive meddlers. It is in the latest and longest of Mill’s treatments of this theme, in the Plato review (387-404), that he first fully expounds the contemporary animus behind his defence. The full case the radicals had to meet was that Athenian morals steadily declined from the time of Marathon on; that this decline was a consequence of the rise of democracy and its concomitant, the attempt to ground morals on reason; that the only true source of morality is the intuition of a rustic aristocracy; and that any criticism of the squirearchy is an attempt to make the worse appear the better cause. The student of Greek literature cannot but recognize in this thesis a misreading of Aristophanes’ Clouds (read by Mill at the age of eight) by someone with naïve notions about the methods of comic writers, and it is visibly an apologia for the sort of tory politics Mill had been programmed by Bentham and his father to overthrow.
Mill’s long excursus on the sophists in the Plato review is substantially a rehearsal of Grote’s own account in his History, which Mill extols in his review of that work (328-9). Mill’s estimate of Grote’s achievement seems justified. In his article on Grote in the Encyclopædia Britannica (11th ed.), J. M. Mitchell remarks that Grote valued the sophists more highly than anyone had before or would again; but that, had it proved true, would still have missed the point. What Grote does is present massive evidence on which his reappraisal rests, and with which any reversal of his verdict must reckon; and those who have disagreed with him have for the most part merely repeated opinions whose untenability Grote had conclusively proved. But the battle still continues: the sophists are still endowed with “every virtue under heaven” by all good liberal democrats seeking ancestry. Such historical shadow-boxing has its ironies, however. What to Mill is an imperfect rationalism, valued as a step towards Plato’s profounder analysis and away from the traditionalism that remained the real enemy, is in some modern eyes a praiseworthy empiricism to be pitted against the evil technocratic totalitarianism of Plato.
Besides the attraction of its presentation of the sophists, Mill is drawn to the Protagoras by finding in Socrates a champion of the principle of utility (61). To those who nowadays call themselves utilitarians, and think of utilitarianism as concerned to promote the good of all, this comes as a shock, since what Socrates advocates against Protagoras is a strictly egoistic hedonism: the possibility that a man might consider anyone else’s welfare is not even mooted. The shock was shared by Grote, since we find Mill writing to him (in what connection we do not know): “[P. S.]—As you truly say the Protagorean Socrates lays down as the standard, the happiness of the agent himself; but his standard is composed of pleasure and pain, which ranges him, upon the whole, on the utilitarian side of the controversy.” But in fact Mill seems to be fairly consistent in equating the principle of utility, as he does here, with “the doctrine that all things are good or evil, by virtue solely of the pleasure or the pain which they produce.” An equivalent account is given in “Professor Sedgwick’s Discourse” in 1835, and again in the second chapter of Utilitarianism (where, as here, the names of Epicurus and Bentham are conjoined); that the happiness sought should be that of the “greatest number” is called for not by the “principle of utility” but by the “Greatest-happiness Principle.”
The notes on the Phædrus afford interesting sidelights on what we found to say about the views on logic and philosophy expressed and implied in the Whately review. Plato’s method of collection and division is subtly revamped to become “decomposition and recomposition,” a method of “philosophical analysis” (93) that is proclaimed as the unique method of philosophy and common to all “systems of logic.” This answers to what Ryan identifies as Mill’s own preference in method; but Plato’s own account, which speaks of a classificatory process of collection followed by one of sorting, seems at least to be of a different kind. In the same connection, Mill writes here as if to claim objective backing for a classificatory scheme were necessarily to ascribe to a class some substantive individuality separate from that of the objects classified (94). It is just this equation that Mill gave up when, as we saw, he abandoned nominalism in his Logic. Meanwhile, it has the interesting consequence that Mill defends something like the “ordinary language” philosophy of the late J. L. Austin: since nothing exists save spatio-temporal particulars, investigations of mental and moral phenomena reduce to the clarification of concepts, and this in turn to the examination of how words are used.
The notes on the Gorgias reflect the same bias that led Mill to describe his omission of the dramatic portions of the dialogues as regrettable only on aesthetic grounds (42). As in his review of Grote’s Plato (406-7), he assumes that Plato’s concern is with the intellectual structure of knowledge, and that Socrates is a seeker of definitions rather than an examiner of lives. It is in this purely negative dialectic that Plato’s value for the nineteenth century is mainly to be found (382-3). Many years later, the accident of Grote’s order of composition in his uncompleted Aristotle, which gave the Topics a more prominent place than he probably intended, gave Mill his last and best opportunity to sing the praises of abstract debate as an intellectual discipline, whose value he learned in Grote’s house as a young man among the Brangles (508-10). This love of logic-chopping might seem at odds with Mill’s determined empiricism, but is actually of a piece with it. Rigorous inductive procedures, the discovery of which in the moral and mental sciences was Mill’s chief intellectual enterprise, depend on the unremitting endeavour to overthrow one’s own cherished beliefs; and it is just this that the lost art of dialectic—so curiously prevalent in the dogmatic middle ages, and abandoned in the sceptical revival of letters—sought to encompass. Meanwhile, however, the assumption that Plato’s interest in the epistemology of morals is confined to the use of argument in the service of moral persuasion leads Mill to make heavy weather of the Gorgias, which he finds to be a tissue of fallacies (149, 395). Plato’s version of the Socratic equation of virtue with knowledge, identified by Mill as the key to his thought on these matters (60), had nothing to do with the kind of understanding that is devoid of personal commitment. Mill says, rightly, that no intellectual demonstration can show that “a life of obedience to duty is preferable, so far as respects the agent himself, to a life of circumspect and cautious selfishness” (149), and that Plato might have furnished Socrates’ interlocutors with rebuttals of his arguments. But what happens in the Gorgias is something else. Socrates discovers internal contradictions, not in abstract hedonism, but in the complex values by which ambitious Polus and proud Callicles believe themselves to live. Their lives are shown to be unintelligible to themselves—a point that is evident only when one considers the dramatic settings of the dialogues, which Mill omits. Even in the reading of the dialectic itself Mill’s strictures show a misunderstanding, in their use of such unplatonic language as “a life of obedience to duty.” Mill presupposes that the moral choice is always one between devotion to self and service to others, idealism and virtue being identified with the latter. It is simply a matter of whose interests shall be served. But the Callicles of the Gorgias is an idealist of a sort, a sort that our twentieth-century relativism makes it easier for us to see pervading Greek culture and that Nietzsche was the first in modern times to detect and extol.
Honesty, as Mill says, is not the best policy, and Socrates in the Gorgias is not arguing that it is. But Mill’s normally cool tone takes on an oddly histrionic pathos as he sings the misfortunes of the virtuous in this wicked world. The tone is that of a man who feels that his own virtue has been unjustly despised and rejected, and whose toil has been without reward. Is this perhaps a memory of those dreadful days of 1826 when Mill first found that the life-work for which he was predestined meant nothing to him? Or can it be only that the world’s reception of his innocent liaison with Mrs. Taylor (as of W. J. Fox’s matrimonial ventures) had shown little comprehension and less generosity?
Among the unpublished translations, that of the Parmenides alone incorporates commentary, and its treatment is in other ways untypical. For the opening section, in which Socrates figures in his habitual manner, Mill follows his usual practice of rendering the argumentative bits and omitting the rest—even, rather misleadingly, the passage where Zeno explains why he has done what he did (128a1-e5). But in the notoriously baffling second part of the dialogue, in which Socrates is silent, Mill abandons his usual practice because, as his comments show, he does not take the discussion (and hence the task of translating it) seriously. Reasonably rejecting the would-be profundities of the neoplatonic commentators, for whom it plumbs metaphysical depths, he takes the alternative to be that it is a mass of quibbles. Still looking for the results overtly established at the end of a dialogue, rather than at what is effectively achieved in the course of it, he inevitably finds this one futile: that it might contribute vastly though indirectly to the clarification of such concepts as “one,” whose ambiguity reduced the earlier discussion between Socrates and Parmenides to incoherence, does not occur to him. The result is that Mill’s abbreviations and omissions reflect no opinion of the purpose of the dialogue and prevent the reader from forming his own. Thus Mill omits a large part (136c5-137c3) of the conversation in which Parmenides clarifies his attitude to the “laborious game” he is to play, and substitutes his own denigration of the proceedings as “verbal quibbles.” Equally unfortunate is his handling of the passage in which Parmenides introduces the concept of an “instant” in time that is not a part of time but has no duration, surely one of the most fruitful suggestions in the history of thought. Mill interrupts his translation here (from 156c1 to 157b5), and substitutes a paraphrase in which he suppresses Parmenides’ assertion that when a thing starts to move there can be no period of time during which it is neither in motion nor at rest, misrepresents the claim that the beginning of movement must be instantaneous as saying that the thing “is for an instant neither in motion nor at rest, but between both,” and then sarcastically refers to his own misrepresentation as “This happy idea” (235).
Mill’s actual omissions in this part of the dialogue seem to be due to mounting fatigue and disgust—and, no doubt, a courteous unwillingness to bore possible readers. As far as 147c1 his translation is only slightly condensed, except that he omits (as F. M. Cornford was later to do) the replies of Parmenides’ respondent, in the hope that they would make no difference. At this point he breaks off, with the caustic comment, “It is unnecessary to adduce more than a specimen of this mode of enquiry” (235), and resorts to brief summary. But he starts translating again at 155e2, resorts to summary at 156c1, resumes translation for 157b5 to 159b2, omits a transitional sentence there, gives a slightly shortened version of 159b5 to 160b2—though what he says Parmenides concludes at that point is not what Plato says he concludes!—and from then on merely summarizes what he takes to be the general drift.
Mill’s opening attempt to characterize in general terms the class of theories to which Plato’s theory of “Ideas” belongs is not so much a comment on the purport of the Parmenides as a vain attempt to explain it away. It is not surprising that Mill was baffled. In his day the materials for any sort of comprehension of pre-platonic philosophy were not accessible, as is abundantly shown by his description of Parmenides as “a Pythagorean philosopher” (222) and his supposition that the Way of Truth rested on the assumption that there was a “mysterious virtue in the word one” (223). This is not so much a mistake as a reflection of an ignorance Mill had no ready means of remedying. As for Plato’s dialogue itself, any comprehension of that was virtually precluded by Schleiermacher’s supposition that it was an early work of Plato’s, rather than one intermediate in date between the Republic and the Sophist.
Mill’s final remark, that the second part of the Parmenides is unfit for its purported purpose of “mental gymnastics” because it exploits ambiguities of language rather than removing them, may or may not be true. That judgment probably depends on whether one expects readers to rest content in a mass of mutually contradictory conclusions without seeking to extricate themselves. True or false, the remark leaves one puzzled. Why spend so much time on a work whose upshot one deplores and whose intentions one does not even profess to understand? Why not stop translating at the point where Socrates bows out? But perhaps only the attempt revealed that it was not worth doing.
Whether to expose, to analyse, or to exploit, the second part of the Parmenides is undoubtedly concerned with ambiguities. That fact in itself is a good reason for not translating it, if one is not obliged to, and in any case not translating it without the sort of incidental explication that Mill eschews. For part of the point of the whole affair must be that the initial hypothesis does not mean, as Mill renders it, “Unity exists.” “ Ἑν ἐστι,” unaccented as Plato wrote it, is ambiguous between “One is,” “One exists,” and “It is one,” at least. “Unity exists” would not seem to be a plausible alternative; but if we assume, as Mill assumes, that a mystification of the sort he describes is intended, “Unity exists” would be the appropriately mystifying phrase. Thus do we deceive ourselves.
Our strictures on Mill’s handling of this very odd dialogue imply no criticism of Mill himself. On the contrary. What he did he did for private purposes of his own, and had the good sense not to publish the results. For this abstention he is to be commended. We may deplore the work, but have no right to blame the workman.
GROTE AND GREECE
george grote, twelve years older than Mill, was one of his closest associates and allies. “ ‘Mill, the elder,’ she [Mrs. Grote] would say, ‘had seized him at the most enthusiastic time of life, and narrowed him, under the idea that he was emancipating him.’ ” Scion of a Tory banking family of German extraction, he had been at school with Connop Thirlwall at Charterhouse, and formed there an enthusiasm for classical letters that he never lost. But on leaving school, in the intervals of keeping up the family end of the bank for his squire-playing father, his inquiring mind led him to the study of that modish subject, political economy; and it was in David Ricardo’s house, early in 1819, that he met James Mill. The meeting changed his life by converting him to the radical cause, though it was not until his father’s death in 1830 that he felt able to play an active part in politics.
Grote’s first and greatest service to the radical cause was to embark on a history of ancient Greece. In the culture of the time, classical civilization was paradigmatic, and the available history was that of Mitford, full of errors and fuller of anti-democratic prejudice. Any reasonably accurate and ample history could be sure to supersede it. But in the meantime, as we have noted, an attack on the sophists was an allegorical attack on philosophical radicals everywhere, and a repetition of Aristophanes’ diatribes against Cleon was a blow against any democrat of the day.
The polemical intent to write a counterblast against Mitford is avowed in Grote’s Preface to his first volume. But in a way the work is a natural outgrowth of an earlier study of Greek mythology, inspired by the notion (not then a usual one) that mythology has historical value as, and only as, a revelation of the self-image of the people who made up and preserved the myths. It was some months later that his wife claims to have given him the idea (quite impracticable at the time, in view of his business and family commitments) of writing a history. But nagging has its limitations and, despite Harriet’s scheme of priorities, the passage of the Reform Bill in 1832 inspired Grote to forsake literature for activism. He presented himself as candidate for the City of London in the radical interest, was elected, and for ten years laid the history aside. By 1838 it was plain that the parliamentary radicals were falling apart, and Grote lost interest, but did not yet feel ready to resume the history. He filled in the time reading Aristotle. In June 1841 he did not stand for re-election and planned to resume work on the “Opus Magnum” in the spring of 1842. In 1843 we find him working at it eight hours a day; that summer he retired from the family bank, and from that time on the History and its successors poured forth in a majestic, if sluggish, torrent.
As an expert and respected businessman, a working politician in troubled times, a polyglot and indefatigable master of contemporary scholarship, and (for the later volumes) an expert on the affairs of Switzerland, whose geographical and political divisions afforded such striking analogies with those of ancient Greece, Grote was outstandingly equipped to make realistic sense out of the biased and fragmentary traditions of Greek historiography. His history won immediate acceptance as a standard work. But time was unkind. Within twenty years, Schliemann’s excavations had raised hopes for the reconstruction of early history that Grote hoped he had shown could never be fulfilled; and in the decades to come the deciphering, dating, and interpreting of the inscriptions with which the Greeks had loved to deface their environment was to emancipate Greek history for ever from that haggling over written records to which Grote and his coevals had been confined. Grote lacked the sharpness of mind and style that in a like situation saved Gibbon from oblivion, and his history ceased to be valuable as soon as it ceased to be indispensable.
Grote awaited the reception of his first two volumes with anxiety. George Cornewall Lewis had offered to review them for the Edinburgh, but it was too late: Mill had asked first, and the books were already in his hands. Harriet was disappointed. She had tried to enlist Nassau Senior’s support in ensuring the suitability of the Edinburgh’s reviewer, with a hint that Lewis would be acceptable. It is true that Lewis had a special interest in the mythological questions that occupy most of these volumes, and Mill had none; it is also true that Mrs. Grote tended to value people for their social standing, and Lewis was in line for a baronetcy. But in any case a certain coolness had developed between Mill and the Grotes, partly for personal reasons, and partly because Mill was struggling to free himself from what he saw as the doctrinaire narrowness of their radical orthodoxy. But Mill retained his admiration and affection for Grote as a person and as a hellenist, and his review proved to be “in every sense, a labour of love; love of the subject, love of the author, and admiration of the work.” It took Mill four days to write and three to re-write, “but I had to read and think a good deal for it first.” This reading included the Iliad and Odyssey for his discussion of the “Homeric question.” Mill throws himself into the cultural game of classical philology with adept enthusiasm, and his disagreement with Grote on the Iliad points the contrast between the two men that divided them politically: Grote relies on mechanical criteria of consistency, whereas Mill emphasizes the organic bonds of feeling that unite the whole. This is plainly the Mill who tempered Bentham with Coleridge.
Mill holds the Greek experience to be exemplary not only because our epigonic civilization looks back to it for instances and excuses, but because the Greeks invented what we think of as civilization, raising themselves from barbarism by their own efforts and invention. That is why such a “philosophical history” as Grote was attempting (but Thirlwall was not) was so important. Their secret, if we could find it, might hold the clue of that mental and moral science that should tease out of the smoky squabbles of the nineteenth century the utilitarian millennium of liberty and happiness. Between the lines of Mill’s review we read that Grote had not unveiled that secret. It is just on the crucial point of how a merely traditional theology and mythology give way to rationalizing ones that Mill finds Grote wanting. Of course there is a sense in which the Greek myths were arbitrary inventions, but Grote seemed blind to the way in which a story can be both known to be invented and believed to be true, exemplary rather than allegorical, and showing in the character of its narrative a response to a need of the mind, as yet inarticulate, for explanations that should have a certain pattern. It is to this quest for the explanatory character of mythology, which the modern reader finds reminiscent of Lévi-Strauss rather than of Grote, that Mill is presumably alluding when he tells Bain that his review has “introduced no little of the Comtean philosophy of religion.”
A student of Mill’s ethical theory will note his interest in the institutions of Sparta as illustrating “the wonderful pliability, and amenability to artificial discipline, of the human mind” (302), as well as the limitations of such disciplines as soon as their constraints are removed. This belief is important to the moralist in Mill, concerned with an education that should instil in everyone the artificial motive of “the greatest happiness of the greatest number.” By psychological necessity, men act in such a way as to maximize their pleasures, though the mechanisms of sympathy include among these a delight in the welfare of one’s associates. But this psychological law places no restriction on what may be desired; it only means that men must be effectively trained to take pleasure as they should. The Spartan experience shows what improbable results such training can achieve. But Sparta has a second lesson for us, in the glamour this horrid régime held for Greeks born under happier institutions. How to explain the fascination of evil? Doubtless the admirers of Sparta, second-rate sensitive minds, overlooked its character as a whole and noted only its freedom from those political and social vices from which they suffered in their own cities. The observation, perhaps a commonplace, seems just, though our contemporaries might trace the admiration for totalitarianism to urges buried deeper; it is also familiar to the hellenist, being precisely the explanation Plato offers in his Republic for the dynamics of political degeneration. Also in the Republic, whatever its Comtian ancestry, is a large part of the argument on how myths are invented, propagated, and believed. In all his moral and political thought, Mill shows himself very deeply a Platonist: not in the foolish modern sense that he “believes in” the “theory of ideas,” but in the way in which the very detail and texture of his thought reflects that of Plato.
Seven years later, Mill had another go at Grote in the Edinburgh. When he saw this second review in print he said it read “slighter & flimsier than I thought it would,” but Grote was pleased:
It seems to me executed in John’s best manner. It is (as you say) essentially and throughout a review of the book; keeping the author, and not the reviewer, constantly in the foreground. It is not, certainly, a review of the eleventh volume; so far “Fish” was right in the remarks which he made on it; but I do not think he did anything like justice to its merits, either as a composition or as a review. It is certainly complimentary to me, in a measure which I fear will bring down upon me the hand of the reactionary Nemesis. . . .
One has the impression that Lewis at least, though not Grote, felt that Johannes Fac-Totum was thrusting himself in again. Whatever Grote may say, Mill’s review, for all its compliments and quotations, is a survey of Greek history in relation to Mill’s well-known preoccupations, which could be dressed up as a review only because the political interests of radical author and radical reviewer so largely coincided. Among these common concerns we have already noted the vindication of those philosophic radicals, the sophists; in this review we find its counterpart, an exposition of the military, moral and intellectual failings of the pious and plutocratic Nicias that serves to show the rottenness of squirearchies everywhere—and no doubt, by indirection, of that venerable scandal of the British army, the purchase of commissions.
Grote had to admit that Mill had not really reviewed the volumes he purported to deal with. The situation in this regard is scarcely changed by Mill’s inclusion in the Dissertations and Discussions version of extracts (amounting to some 30 per cent of the whole) from his reviews of earlier volumes. The reason is that the exemplary function Mill ascribed to the Greeks was fulfilled only by the Athenians at the time of their political and cultural supremacy, with which the present volumes were not concerned. Mill’s Athenophilia is shown already by his inclusion of the long quotation from Neibuhr in the otherwise trivial review he printed in 1840 (241-3). The present manifestation of it, with its almost ludicrous encomium on the fun-loving Athenian populace, reveals not only the thrust of Mill’s political programme but the structure of his personal values. Genius and joy, personal freedom and intellectual culture flourish only in an atmosphere that exacts public service but ignores idiosyncracies of word and act, an atmosphere that requires political institutions such as only “a succession of eminent men” could devise. In such a society Mill would be praised for his engagement in public affairs and not censured for his private affair with Mrs. Taylor; in such a society such men as Bentham and the Mills secure freedom and joy for their fellows. Mill shows less sign of misgiving now than he had in the 1840 review that the values modern times have added to the Athenian scheme of functional democracy—internationalism, kindness, mildness of manner, bureaucratic efficiency and the techniques of political representation—might be incompatible with the vivid individualism of the Athenians; but this insouciance is consistent with his methodological atomism. Relations of cause and effect, means and ends, are of course recognized in the world of values, so that we see how (for example) a prerequisite of the flourishing of genius is a lack of inhibition; but the positive goods at which society aims are dealt with in the breathtakingly arithmetical fashion familiar to readers of Utilitarianism. The sum of two goods is a good. Mill seems not to want to admit that, just as when my interests and yours conflict we must compromise or fight, a society may have to sacrifice parrhesia to mild manners or mild manners to parrhesia.
Within the encomium on Athens, a certain tension may be felt. Its democratic institutions are praised as the work of a “succession of eminent men”; yet the failings of its operations are excused on the ground that its policies did not express the sound heart of the Athenian working stiff, since the “conduct of affairs was habitually in the hands of the rich and great” (331). Does this statement mean that eminent men, not popular movements, create democratic institutions; but within such institutions popular movements are good and eminent men are bad? Perhaps it does mean that—and perhaps it is true. In any case, the tension is relieved when we infer from the context that the praised eminence is one of ability, the bad greatness merely one of wealth and family. Yet when we reflect on the individuals thus involved in praise and dispraise—Solon, Cleisthenes, Miltiades, Pericles, Nicias, Cleon, Alcibiades—we may think that this review at least takes no clear stand on that most intractable of political issues, the proper relation of outstanding individuals to a democratic constitution.
One feature of Mill’s praise of Athens may strike the modern reader as strange: his readiness to condone Athenian imperialism, on the excuse that force may be necessary to inaugurate the reign of reason. This apologia—considerably toned down, on Harriet’s protest, from the version in a Spectator review —comes oddly, we may think, from so staunch a champion of the liberties of women and slaves. But Mill’s position is necessitated by that belief in Progress whose absence he was to deplore in Aristotle (505). His attitude is coloured, at least, by his experience as an official of the East India Company, as appears sufficiently from the reference in his earlier review (290) to Sleeman’s Indian observations as a revelation of primitive mentality; but his opinion is articulated more clearly in the remarks on Indian affairs in his personal papers. The local “native” régime may be a set of interlopers or usurping tyrants; but an Imperial government, remote from local squabbles, has no other concern than the welfare of all its charges. Besides, in most cases if not in all, the powers of the central government are those ceded to it by the local authorities in the interests of efficiency, economy, or political advantage. Of the ruling notions of modern anti-colonialism, that all cultures are created equal and that no Indian can be an alien anywhere in the subcontinent, Mill shares neither. It is in fact the gravest charge against Nicias that his failure in the unprovoked aggression against Syracuse betrayed the cause of Athenian imperialism:
If the Athenians had succeeded they would have added to their maritime supremacy all the Greek cities of Sicily & Italy. . . . Even if they had failed & got away safe, Athens could never have been subdued by the Peloponnesians. . . . Perhaps the world would have been now a thousand years further advanced if freedom had thus been kept standing in the only place where it ever was or could then be powerful. I thought & felt this as I approached the town till I could have cried with regret & sympathy.
Although Grote’s work has been left far behind by the advance of scholarship, the patience and amplitude with which he set out all the evidence he did have and teased out the last shred of its significance gave his work some permanent value as a guide to the historiographical tradition. In fact, he goes far to justify Mill’s strong claim “that there is hardly a fact of importance in Grecian history which was perfectly understood before his re-examination of it” (328). This residual value is a function of his patient prolixity. Writing of Macaulay’s history a few years later, Mill summed up: “What a difference between it & Grote’s Hist. of Greece, which is less brilliant, but far more interesting in its simple veracity & because, instead of striving to astonish he strives to comprehend & explain” —provided, of course, that we bear in mind that the simple veracity is that of a proselytizing radical.
The review of Grote’s Plato and the Other Companions of Sokrates is, if anything, less of a review than those of his History. The fault, if fault it be, is not all Mill’s. A comprehensive survey of an author’s work can hardly be reviewed otherwise than by measuring it against the standard view of that work, and we have seen that the state of British Platonism was still such that no standard view could be said to exist. Mill’s opening description of Grote’s achievement suggests an additional reason. Grote’s picture of Plato, though painstakingly traced, is anything but sharply delineated. Years before, Mill had remarked that Plato was taking Grote “a length of time only to be warranted by using the opportunity to speak out very plainly on the great subjects—a thing I rather wish than expect he will be found to have done. . . . ” And if Mill could at last find much in Grote to praise, it was not because these misgivings were unjustified. The three laborious volumes are a remarkable compendium of scholarly opinion and philological lore, full of sagacity, and embodying just such an abstract of the several dialogues as Mill had thought proper to prepare himself. But it is hard to believe that even in their day they can have afforded much moral or metaphysical excitement, and only the loyal eye that discerned the genius of Harriet Taylor could have detected in their author one of the leading metaphysicians and psychologists of his age.
As its scale might suggest, Mill’s review was long premediated, and he warned Grote that he would be using the book as a springboard for his own considered view of Plato’s achievement. He re-read the whole of Plato, in Greek, to prepare himself, and to good effect: he provides a majestic survey, a truly remarkable synthesis even with Grote before him, revealing once more his gift for the even-handed presentation of a mass of fact. “I have seldom given so much time and pains to a review article,” he told Grote. Yet, though steadily enlightened, one is seldom astonished. Mill, now almost sixty, treads the round of his Platonic preoccupations: the merits of the sophists, the value of a negative dialectic, the praise of intellectual independence and a high moral tone, the preference of methods over results and the antipathy to dogma. But he is able to indulge these preferences because Grote shares them. As with the History, Mill thinks Grote’s Plato important as the first systematic treatment of its subject from the point of view of the “experience philosophy”—from which it follows, again as with the History, that since this is the only true philosophy Grote’s treatment is the first that really illuminates its subject. Up to now, Plato’s achievement had been obscured by the orientalizing neoplatonists and German transcendentalists who preferred the obscurantism of his senility to the inquisitiveness of his vigorous youth. These two schools of Platonism, the disputatious and the arcane, are with us still; and, in so far as Anglo-American academic orthodoxy is still wedded to one or another form of the “experience philosophy,” the Platonism of Grote and Mill is substantially that still imparted to most anglophone undergraduates. But in one fundamental way all Platonists of that age are sundered from all those of today, by a crippling defect of which Mill is well aware (385-7). They had no way of dating Plato’s dialogues otherwise than by circular inference from a conjectural development in his thought. Just two years after Grote’s first edition, such a method was discovered and published, and despite many problems its main results are unchallenged. One of the points agreed on is the first to be established, and suffices to undermine Grote’s whole structure: the Sophist and Statesman must be later than the Republic. This makes it impossible to think of Plato as steadily degenerating from dialectical maturity to dogmatic dotage (or, as the opposing school would have it, wading from shallow scepticism into mystic profundity). At the same time, the discovery that the critique of the “theory of ideas” in the Parmenides is later than the dialogues which argue most unequivocally in favour of that theory has made today’s scholars reluctant to accept the view assumed by Mill and his coevals, that the theory of ideas was a simple-minded doctrine in which Plato basked content.
The effect of establishing an unexpected order for the composition of the dialogues is not only to rewrite Plato’s intellectual biography. It makes Plato a much more difficult writer than Mill and Grote took him for: whatever he is up to, it cannot be the straightforward things we used to think, and we have to read him with much greater caution. Mill and Grote acknowledge the difficulty of interpreting the meanings of the dialogues as total compositions, but suspect no difficulty in their parts. It is noteworthy that though Mill affirms broadly that all the arguments of the Gorgias and the Phædo are fallacious he does not specify the fallacies. He does not even analyze the arguments. And the attribution of the “theory of ideas” to an “imperfect conception of the processes of abstraction and generalization” (421) rests on no serious consideration of what the Platonic Socrates says and the reasons he actually gives for saying it. This sort of superficiality, however, was probably inevitable in the then state of Platonic studies, even without the disconcerting results of stylometry: only after the sort of overview established by Grote had become thoroughly familiar would it be feasible for a more penetrating critique to look into the actual fine structure of the arguments. Indeed, much of the work has yet to be done. If classical studies are moribund, they will die in their infancy. Meanwhile Mill and Grote, true apostles of Progress, assume that Plato’s thought belongs to the childhood of the race and that contemporary thought has nothing to learn from him (see 421): he is to be judged by how close he has come to the position reached by nineteenth-century empiricist radicals. For that reason it seemed suitable to make a study of Plato an opportunity for speaking out on the great questions of the day. Mill, just like the transcendentalist interpreters he complains of, will let Plato inspire him but not disquiet him.
Because Mill is not prepared to discover that Plato’s thoughts are other and better than his own, he is apt to say that Plato has “failed to grasp” a point on which they are at odds. Thus he blames Plato for his thinking that techniques of measurement were a sufficient guard against error, and for “overlooking that it is not the act of measurement which rectifies them, but the perceptions of touch which the measuring only ascertains” (420). Plato is not overlooking this belief; he is denying it. The disagreement is radical. Plato’s Socrates is clearly presented as believing that getting one’s sums right can be a significant moral passion, and that it is the moral passion of the just man. The appeal to a method is essential. Indeed, it is rather strange that Mill does not recognize here a reliance on calculation akin to that of a Benthamite legislator. But his remarks on the handling of “justice” in the Republic are full of puzzling things. How can he say that Plato’s ethic allows no place for that paradigm of Athenian justice, Aristides, whose “justice” lay in his unfaltering adherence to the highest convictions of his own place and time? The Republic locates such a man very precisely, as the man of moral courage, in a passage where Plato also affirms another essential point that Mill accuses him of denying, that such tenacity of one’s proposals is a precondition of any “justice” based on independent intellectual comprehension. Again, we wonder how Mill can accuse Plato of ignoring the fact that justice has to do with the rights of other people (419), when the fact that justice is “another’s good” is a key point in the case Socrates is called on to meet. It seems strange that Mill should have missed so much of what is going on, especially as his remarks on the discrepancies between the “mixed modes” of Greek and English thought show him so well aware of the dangers of relying on the customary associations of English terms when discussing Greek philosophy. Part of the explanation may be found in the final chapter of Utilitarianism, where justice figures as a set of entrenched principles and patterns of behaviour exempted from felicific calculation and calling on its own special set of instinctual resources. The whole arrangement of thought is quite alien from Plato’s. Mill is so deeply imbued with the sort of moral psychology inaugurated by Hobbes that he is unable to entertain the very different moral psychology envisaged, on grounds no better and no worse than his, by Plato.
The praise of Athenian democracy in the review of Grote’s History should prepare us for a denunciation of Plato, whose ideal institutions in the Republic seem designed to remove not only what Mill sees as the incidental vices of that polity but, very specifically, each of its merits. It comes as something of a surprise, then, to find Plato praised for recognizing “that the work of government is a Skilled Employment” (436), even though this praise is tempered with censure for having gone too far and denied the unskilled any say in the direction of their affairs. Perhaps the surprise is mitigated when we read in Mill’s diary for 20 March, 1854: “The Reform Bill of the present year and the plan of opening the Civil Service of Government to universal competition, are the most wonderful instances of unsought concession to the democratic principle—the former in its ordinary, the latter in its best, sense—which a reformer had imagined even in his dreams.” In a reformer’s dreams, apparently, the institutions of Plato’s Republic are, in the best sense, democratic.
With Plato finally squared amply away, Grote was free to turn to his beloved Aristotle. It would be an understatement to say that he did not live to complete his task. No lifetime would suffice to write, scarcely to read, a treatise of the majestic proportions implied by the two stout volumes he left for Bain and Robertson to edit. And, as the scale of his writings increased, Grote himself was slowing down. He was in his late seventies and, though he secluded himself faithfully in the mornings, Harriet would find him snoozing over his papers. Though the treatment of the logical works is lucid as ever, his discussion of the principle of contradiction as expounded in the Metaphysics, which is the last completed portion of his work, betrays a gently wandering mind.
Grote himself thought the account of Aristotle’s psychology, written independently for the third edition of Bain’s The Senses and the Intellect in 1868, was the best thing he had done; but the learned world soon consigned the whole of his Aristotelian studies to a common oblivion. Mill’s review is a most delicate act of piety. He praises the book in general terms, without specifying how Grote has advanced the understanding of his subject. He subtly maintains the pretence that almost all he knows about Aristotle he knows from reading Grote, while in taking issue with his views on Protagoras he implies that the old man’s mind was still worth arguing against. In fact, however, there is more of interest in Mill’s review than in all of Grote’s two volumes.
In his review of the Plato, Mill had urged that Aristotle was easier than Plato for modern readers to understand. But many of his own remarks strike us as showing a lack of comprehension of what Aristotle is up to that is not paralleled in his treatment of Plato. In fact, though he knows Aristotle’s logical works intimately, and has studied his political, ethical, and rhetorical writings at first hand, his remarks on the psychological, physical, and metaphysical writings show only such knowledge as one might derive from an article in an inferior encyclopedia. Thus he castigates Aristotle for erecting “chance” and “spontaneity” into independent causative principles (482-3), whereas what Aristotle is doing in the relevant passage (Physics, II, which Mill had not read) is analyzing the nature of the error committed by those who make that supposition. For the most part, his misunderstandings are of two kinds. In the first place, he supposes that whenever Aristotle discusses such topics as “matter,” “form,” or “privation,” he is isolating components of the universe, as though he were carrying out the sort of metaphysical analysis that (we have suggested) Mill himself thought proper; whereas one of Aristotle’s avowed aims is to refute the Platonic thesis that such terms as “form” and “soul” denote separate entities. In the second place, more generally, he pays little attention to Aristotle’s actual arguments, and hence regularly misconstrues the type of explanation that Aristotle purports to offer. The complaint that analyses in terms of matter and form “give no power of prediction” (503) presupposes that Aristotle’s “First Philosophy” is either an abortive attempt at physical theory or a misguided failure even to make the attempt; but to presuppose that is to debar oneself from considering what Aristotle actually says, hence what he might mean by what he says, and hence again what might be the point of saying it. The diatribe against the German “transcendentalists” in the Plato review shows a similar blind antipathy to every dimension of understanding save one, that expounded in the Logic, and attacks just those qualities in German thought that were already enabling German scholars to take the first significant steps in the re-discovery of Aristotle. When he blames Aristotle and the other Greeks for not believing in Progress, the attack is two-edged. Aristotle’s contemporaries had not opened their minds to that form of understanding that might conduce to an endless series of changes in human arrangements, but Mill’s mind is still closed (as in his unregenerate Benthamite youth) to any form of understanding not so directed. Nothing in Mill is so disconcerting as the combination of his air of massive tolerance, the breadth and judiciousness of his surveys of the intellectual scene, with a crippling dogmatic narrowness in metaphysical method.
Some aspects of Mill’s treatment of the logical issues have been noted already. In this area, too, Mill construes Aristotle as addressing himself with greater or less success to Mill’s concerns, without seriously considering whether Aristotle might not legitimately have different concerns of his own. Mostly on the basis of the Topics, Aristotle is said to have shown a reliance on simple induction that was possible for him only because its failures were not yet apparent—it was the failure of simple induction that enabled Bacon to look for something better. But the Topics is not concerned with induction. Aristotle is only laying down a rule for debates: generalizations must be allowed to stand unless actual counter-examples can be produced. And in the Analytics there is no place for a logic of induction at all, but only a process in which one leaps to a conclusion about the way things are. Neither Mill nor Grote alludes to that celebrated image of the rallying of a routed army, in which Aristotle shows how very far his concerns in the matter are from theirs. But perhaps Mill should be understood to mean only that if Aristotle had had a logic of induction, it could only have been one based on simple induction; and that would probably be true, for “Mill’s Methods” are well named.
The discussion of the principle of contradiction reveals a more complex disagreement. Grote’s treatment is confused, and it is not clear what he thought Aristotle’s doctrine really was, but Mill finds in Grote himself the doctrine that the principle can be established only by induction from particular instances in which it is conceded to hold. As Mill points out, the question is a vital one for a convinced inductivist. But Mill misses Grote’s point, which is the perfectly reasonable one that this is how the sort of ad hominem refutation that Aristotle relies on would have to proceed. However, Grote himself seems to be mistaken in supposing that Aristotle is speaking of an argument that one could use against someone who denied the principle of contradiction. Aristotle’s position seems rather to be only that anyone who makes a definite statement must in fact rely on the principle, whatever he may say, because the meaningfulness of his statement rests on the denial of the contradictories of the definitions of the words he uses—if a man abjures statement, he cannot of course be refuted, but he cannot be agreed with either: he forswears communication. The position that Mill takes up against Grote, and in effect against Aristotle, is approximately that which a modern logician would adopt, that the principle of contradiction simply embodies the rule for the correct use of negative terms: a proposition and the negation of its negation are the same proposition. But in going on to say, “the axioms in question . . . have their root in a mental fact which makes it impossible to contravene them” (499-500), Mill courts disaster. If he is to maintain his particularism, a mental fact must be a fact about the mind of some individual on some specific occasion. But about whose mind, at what time, is it a fact that the principle of contradiction cannot be contravened? Worse, to write thus is to imply that what is impossible is not that a proposition and its contradictory should both be true, but that they should both be believed to be true by the same person at the same time. But, since no mental fact prevents me from changing my mind about something I am saying even while I am saying it, our sole ground for saying that such a combination of beliefs is impossible is that they are logically incompatible. This lapse into an indefensible and misplaced psychologism comes strangely from Mill, who as we shall see had elsewhere insisted that a proposition was not proved true by the impossibility of disbelieving it, and it is interesting to find it in such close proximity with the extensive footnote in which Mill fires the last shot in his long campaign against Grote’s subjectivist reading of Protagoras (500n-501n). Mill’s pretence that in this matter his difference from Grote is merely a verbal one is charitable, but untenable: if Grote was unmoved by Mill’s annihilating argument in the Plato review (426 ff.), he was incorrigible. Mill could not reconcile himself to the existence in the world of men, otherwise apparently rational, who believed themselves to believe that whatever a man believes “is true for him.”
In a letter to Pasquale Villari on 28 February, 1872, Mill wrote: “You judged truly that the loss of Mr. Grote leaves a great blank in my life. He was the oldest & by far the most valued of my few surviving old friends.” Four months after his review of the Aristotle appeared, Mill joined the majority.
BERKELEY AND BAILEY
the review of fraser’sBerkeley has an impressive air of omnicompetence. As the reviewer follows his subject through the realms of metaphysics, mathematics, medicine, and political economy, he never admits or reveals himself to be at a loss. He compliments Berkeley on the early completion of his system with the sympathetic respect of one who, like him, had known everything at an age when his contemporaries had yet to learn that they knew nothing. In Berkeley, as in Plato, Mill finds a kindred spirit. All three men had put their intellects at the service of a moral passion; had disguised the subtleties of their arguments in an easy and eloquent style where cogency and sophistry intertwined; had used a single intellectual method and style to bring a wide range of phenomena within the compass of a single system; had preferred the manner of the debater to that of the expositor, though Mill published no dialogues. Nonetheless, one is surprised when Mill claims that Berkeley excelled all metaphysicians in “philosophic genius,” and bases his claim on a revolution he effected in the state of philosophy. Surely Berkeley’s impact was less than Plato’s, who found philosophy brick and left it marble, and no greater than that of the other luminaries Mill enumerates. But one glance at Mill’s list of Berkeley’s major innovations makes all plain. Berkeley made the Mills, father and son, possible, and it is only because this seems a lesser feat to us than it did to Mill that we esteem him less. Bailey wrote shrewdly when he attributed the blind vehemence of Mill’s defence of Berkeley’s theory of vision to filial piety. It may even be that Bailey’s taunt opened Mill’s eyes to the relationship, for the Bailey review states reservations about Berkeley’s technical competence that the Fraser review expressly withdraws.
Fruitful and suggestive as they are, the three theses that Mill singles out as founding “the true analytic method of studying the human mind” seem neither cogent nor clear. The first one states that “the connection between our impressions of sight and the facts they indicate can be discovered only by direct experience” (457). This view seems odd. What could this independent order of indicated facts be? How is it discovered? If we have independent access to it, how do we decide that its status is that of something indicated rather than indicating? Indeed, Berkeley’s point was that there are no such ultimate and privileged facts: all experiences can be related to all other experiences, and the reality to which they are to be referred is nothing but the complex internal structure of the world of experience itself. This point will occupy us later. Meanwhile, the second thesis singled out by Mill is that there can be no abstract general ideas: all ideas are particular, and are ideas of particulars. But what are ideas (Mill speaks here as if they were pictures one stored in one’s head), and how are they to be individuated? Mill takes up this problem in connection with the third thesis, but seems unaware of its urgency in relation to the second. Berkeley, indeed, might think there was no problem. If a man can use the phrase “a horse” with reference to a collectivity of experiences which, though its structure is indescribably complex, is familiar enough to work with, there may be no difficulty in principle in his using the phrase “my idea of a horse” in similar vague reliance on his capacity to find his way around his head. But this is, precisely, not an analytic method. Mill here offers a sketch for an analytic version that relies on his distinction (first expounded in the Whately review, 24) between the denotation and connotation of terms. It is baffling in its brevity. A name is “common to an indefinite multitude of individual objects,” but “is a mark for the properties . . . which belong alike to all these objects” (458). So it seems that only individual substances exist; but properties are identical in all their instantiations, and words are identical on all occasions of their utterance. Plato’s beard, it seems, is far from closely trimmed; and it is not clear just what Mill thinks Berkeley’s denial of general ideas has achieved, or what relation he sees between Berkeley’s contention that every idea must refer to an individual and his own contention that every word must stand for an indefinite number of particulars—a contention which it is easier to harmonize with Wittgenstein’s later notions about following rules than with earlier empiricist theories of reference. But these are tangled issues, and Mill’s own footnote at this point shows that they cannot be effectively broached in so restricted a context as this.
The third thesis is that of immaterialism itself: that the system of appearances is self-sufficient, and that nothing is gained by postulating any unobservable “matter” to sustain it. Mill, like most commentators, thinks Berkeley failed to rise to the height of his own argument when he substituted a divine will and consciousness for the missing material substratum. All we need postulate, says Mill in Heracleitean vein, is the permanent possibility of sensations, “or, to express it in other words, a law of uniformity in nature, by virtue of which similar sensations might and would have recurred, at any intermediate time, under similar conditions” (464). Mill states the fundamental objection to Berkeley’s move with admirable vigour and succinctness, though he recognizes that it was for the sake of this move that the pious undergraduate devised his system: the same argument that shows ideas to be inseparable from minds also shows that each idea is inseparable from the mind that conceives it and from the occasion of its conception; and the “notion” whereby an idea is held to imply the presence of a spiritual force apt to produce it is groundless. But he shows himself rather insensitive to the reasons that support Berkeley’s unfortunate postulate. For we may ask, as many have asked, what could be meant by a “permanent possibility.” Philosophers since Aristotle have agreed that the word “possibility” cannot be understood as standing for any independent reality or state of affairs, but only for the potentialities of some actual system or structure. Nor does one make the idea of a permanent possibility any more plausible by equating it with a law of nature. A law of nature, one might think, could be no more than a description of a sequence of phenomena, unless that sequence itself is a sample of a coherent reality, or a partial manifestation of a coherent will. Mill underrates the stubbornness of this question. But the argument is probably irresoluble, and there will always be four parties to the debate: those, like Berkeley, who boggle at reducing the fabric of nature to a hypothesis about the minute patch of the fabric that someone has perceived or will perceive; those, like Mill, who cannot surmount the impossibility of specifying what such a reality would be otherwise than in terms of what would in certain circumstances be perceived; those who see both parties as victims of obsessive verbal ideologies whose practical consequences must be identical and which accordingly cannot merit allegiance; and those, the happy majority, with neither the will nor the wit to grapple with the issues involved.
In contrast with the special pleading in relation to the three metaphysical theses, the overall assessment of Berkeley’s activity is impressively just and broadly based. The range and keenness of Mill’s mind seldom appear to better advantage. The review of Bailey—written, like the first of the Grote reviews, with astonishing speed —and the reply to Bailey’s rejoinder have the same air of reasonableness, sympathy, and force. They convince one that the unfortunate autodidact, of whom Mill speaks with more sorrow than anger, has met with justice tempered with as much mercy as the case allowed; and Packe, for one, takes his word for it that Bailey was demolished. Yet an uneasy suspicion may enter. The weight of prejudice is now on the other side, and the informed modern reader finds Berkeley’s theory of vision as incomprehensible as Bailey found it. Could it be that he had seen something that Mill overlooked? After all, Mill’s account of what Bailey said is not very full. A reading of Bailey confirms our suspicions.
Bailey sometimes expresses himself loosely, but builds a formidable argument against the Berkeleian position. So far from answering it, Mill seems not to have grasped it; certainly one could not guess from reading Mill what the argument was. The misunderstandings between the two men are far-reaching, and the issues themselves dismayingly complex. I fear that my attempt to clarify the issues will only add new confusions. But an attempt seems necessary.
Since the publication of Berkeley’s Commonplace Book, as Mill recognizes in his 1871 review, it has been clear that the separate publication of the New Theory of Vision was merely a tactical manœuvre. Berkeley was already in possession of the arguments whereby the Principles of Human Knowledge would show that sensations of touch are as certainly “in the mind” as are sensations of vision. But the New Theory of Vision had argued that whereas visual sensations are “in the mind” the sense of touch gives us direct knowledge of an external world, so that by associating our visual sensations with tactile ones we can refer them to the external world that is their cause. This is an absurd position. As Bailey insists, citing Berkeley’s own later works, the five senses are on precisely the same footing in the matter of external reference. Yet, as Mill says (453-4), most philosophers had accepted the theory of vision and rejected general immaterialism, as though the arguments for the former would not sustain the latter. How was this logical monstrosity possible? Partly, as we shall soon see, because of the compelling attractiveness of one particular image; partly, according to Bailey, because Berkeley’s supporters confused the issue of how we judge distances with the quite separate issue of how we initially form the notion of an external world; but partly, no doubt, because if any of the senses is to have the status of sole testimony to reality, it can only be the sense of touch. But the argument demonstrating the priority of touch is an old Aristotelian one from speculative biology, and rests on taking the initial distinction between self and other for granted: an organism must sustain itself by interaction with its environment, and this interaction takes place at the surface of the body, so that the distance senses must refer in a general way to what might impinge on the skin. But this is a far cry from the epistemological reduction in which tangible properties alone are ascribed to reality, and in which the proper objects of vision are held to be subjective phenomena that have the proper objects of touch as their sole objective referent.
The argument of Bailey’s book rests on a foundation that Mill never mentions. This is the distinction between externality, the fact that we locate the objects of vision in a world outside ourselves, and distance, the fact that we locate them within that world at various distances from ourselves. He urges that the distinction between inner and outer is not one that could be learned from any experience, but must be presumed innate: after all, a newly-hatched turtle immediately makes for the water, which it must therefore perceive as “outside.” In any case, the alleged priority of touch over sight in this matter is a myth: “When an object is printed on the retina, the object is seen to be external as directly and immediately as the object is felt to be external.” The location of objects at various distances from the observer, on the other hand, is indeed learned in experience, if not actually from experience (Bailey finds the last point hard to determine, because human infants are less fully developed at birth than the young of other species, and hence are slower in the development of all their faculties). But the relevant experience is not exclusively tactile, and need not be tactile at all, for visual space is a self-contained system that is generated by the complex structure of appearances as the eyes change conformation and viewpoint. Nor is it the case, as some Berkeleians supposed, that change of viewpoint is a tactual matter because it depends on the muscular sensations of perambulation: children have the full use of their eyes at an age when they lack effective use of their legs and are carried everywhere by their mothers. This visual space is of course correlated with tactile space, but if it were not a space in its own right there could be no correlation. And, again, the initial correlations between the spaces must be intuitive, because they could not be learned. Mill counters this last move by citing the experience of people who have recovered their vision by surgery and at first cannot interpret what they see. Unfortunately he relies on the Cheselden case (264), which does not support him: it is that of a boy who has to catch his cat and hold her before he can tell she is not his dog. But, as Bailey points out, the terms in which Cheselden describes this episode show that the boy could already recognize the feline shape as constant through its various occurrences and transformations, and, apparently, that he could use visual cues to help him catch the cat; what he could not do was correlate this recognized form with a familiar tactile form—in Cheselden’s words, he “often forgot which was the cat, and which the dog.”
Mill’s appeal (262) to the structure of the eye as conclusive evidence for the Berkeleian hypothesis tends to confirm Bailey’s claim that the hypothesis rests essentially on a single argument: that distance must be invisible, because the three-dimensional world projects on the retina a two-dimensional image. This image “painted on our retina,” in the significant phrase used by Mill (253) and unwisely accepted by Bailey, is the true object of vision. It is as though there were a second organ of vision gazing at the retina, which is envisaged as a tiny screen set up inside the head, like a camera obscura. And the image on this screen is thought of as like the image of a pinhole camera: flat, simultaneously clear and detailed and sharp in all its parts. It is astonishing that this entirely fictitious notion of the facts of vision should have dominated the psychology of the senses for so long. It is plainly derived, in Mill’s give-away phrase, from “as much of optics as is now commonly taught in children’s books” (253), and in particular from a simple diagram showing how light-rays pass in straight lines from solid objects through the pupil to their images “in the fund of the eye”: in these diagrams, distance is indeed represented by what Berkeley so unintelligibly says it is, “a line directed endwise to the eye” and projecting a single point on the retina. This is gibberish. Optical diagrams do not represent the processes of vision. Bailey points out that only a material line (a thin wire, or something of the sort) could project anything in the fund of the eye, and then only if it stopped short of the eye itself; and if it did project anything in the fund of the eye it would have to be visible. If what Berkeley is talking about is the line we draw in our diagrams when we are explaining the laws of optics, this is not the sort of thing that can either project or fail to project anything. As Bailey says, “The distance of an object from us is not a line presented endwise to the eye: distance is not represented on the retina by a point. These are phrases which describe no real facts.” And to Mill’s unfortunate contention (254)—“the distances of objects from us are represented on our retina in all cases by single points; and all points being equal, all such distances must appear equal, or rather, we are unable to see them in the character of distances at all”—he responds, unanswerably: “If distances are seen, and seen to be equal, and yet not seen in the character of distances at all, will the critic be obliging enough to say in what character they are seen?” Mill does not help himself by saying (267) that objects, the spaces between objects, and the distances between objects and the eye, are all projected on the retina in the same way: it only becomes more obvious that he is using a language that is quite inappropriate to the phenomena. Yet Bailey wins only a debating victory. For the reason why Mill is so bewildered by Bailey’s obtuseness in failing to see the obvious is that he is not thinking about the implications of the language he is using: he is simply reminding Bailey of characteristics that the pinhole-camera image and the optical diagram really do have. And Bailey has no right to object, since he himself accepts the language of “painting on the retina” that makes Mill’s remarks seem apposite.
Bailey seems always about to discover the delusive nature of the optical diagram as a representation of the facts of vision, but never quite succeeds. Because the retinal image is literally inside the head, and because visual imagery is in a sense subjective and in that sense “in the mind,” the Berkeleians wrote as if the visual world were originally seen to be inside the head. Bailey sees that this is absurd: “If an external object can be perceived by sight as such, it must be perceived also to be distant; to stand apart, or occupy a different portion of space from the being which perceives it.” But he cites the Berkeleian argument in a way that shows he has missed the essential point, that if there are such things as primordial and uninterpreted data of sensation they must be seen neither “as” inside nor “as” outside, since it is from such data that the concepts of inside and outside must themselves be constructed or derived. Similarly, though the retinal image is two-dimensional, our visual data can be neither two-dimensional nor three-dimensional, since those terms derive their meaning from reference to a space within which both solid and flat objects can be distinguished. Bailey recognizes and remarks that within this visual world flat things may be mistaken for solid ones and solid things for flat ones, and that in both cases (whatever Mill may say) the correction is less likely to be supplied by the sense of touch than by a closer look; but he quotes Berkeley’s own fundamental observation, that visibilia must in themselves be neither plane nor solid but have figure and shape in another mode, only to reject it as involving a contradiction.
The crucial consideration, that the clear and distinct retinal image is a fiction, is hidden from Bailey and Mill alike. The closest they come to recognizing it is in their discussion of Wheatstone’s recent (1838) demonstration of the stereoscopic mechanisms of binocular vision. It is amusing to see how the two theorists deal with these facts. Bailey attributes the muscular sensations derived from the accommodation of the two eyes to the sense of sight; Mill, for whom only what can be read from the optical diagram can be visual, associates the same sensations with the sense of touch. If Mill’s position seems ridiculous, this is not because Bailey is right, but because this analytical method of dividing labour among the senses and the intellect and other “faculties” cannot be made to work. Bailey claims that the stereoscopic phenomena themselves show that the perception of “geometrical solidity” does not depend on inference; but Mill has only to reply that he never denied there were distinctive visual phenomena associated with distance, he only said that the fact of their association must be inferred. And after all, he says, Wheatstone’s researches show no more than that we have to do with two retinal images, not one. Neither theorist shows any awareness that because of the indistinctness of peripheral vision, the shallow field of the lens of the eye, and so on, a viewer must be restlessly active in constructing a visual field that would have the characteristics they attributed to the retinal image. Seeing is a complex activity, not a passive reception of stimuli. Bailey comes a little way towards realizing this fact; Mill does not even begin.
The issue between Mill and Bailey resists clarification, not merely because they are victims of a common delusion, but because each is involved in a hopeless confusion of terms. What do they mean by “distance”? Mill speaks as if it were a sort of visible object, the ghost of a dimension; Bailey as if it were a homonymous term referring to disparate properties of visual and tactile spaces. What makes their controversy so hard to follow is that they both assume that the term “distance” must function as the name of an object, of which it might make sense to assert or deny that it was visible; whereas in actuality the word functions in a much more various and elusive way, alluding to a complex and pervasive aspect of our experience, and to the mass of heterogeneous procedures that go by the name of “measurement.”
One aspect of this terminological vagueness and confusion was clear to Mill. He complains of Bailey’s question-begging use of the term “perception,” a term that combines a clear reference to sensory experience with a quite indeterminate claim of some sort of real status for the objects of that experience. Bailey makes facile fun of this legitimate and serious complaint by quoting James Mill—an authority, as he maliciously observes, that every Westminster Reviewer is bound to respect—as saying, “I believe that I see distance and form; in other words, perceive it by the eye, as immediately as I perceive the colour.” “Who does not see that the word thus employed has a precise meaning?” he asks, and explains: “When I speak, without any qualifying adjunct, of perceiving an object by sight, I simply mean seeing it; when I speak of perceiving an object by touch, I simply mean feeling it.” But distance, form, and colour, which are what James Mill says he perceives, are not objects at all in what seems to be Bailey’s sense. The whole question turns on what one may properly be said to “see,” and what one should mean when in the context of a psychological discussion one says that a subject “sees” something. And on this subject Bailey complacently wallows in a slough of confusion. Mill, he insists, is wrong to say that when we mistake a plane for a solid “our error consists in inferring that it is solid,” for “The perception of solidity, or, if the phrase be preferred, the undoubting belief that we see a solid object is, in both cases, equally an impression produced at once upon the mind through the eye, without any process of reasoning or suggestion”; it is an affection of the optic nerve that produces “a certain affection of the mind called seeing an object,” so that “the third dimension of space is seen” whether one is looking at a real solid thing or being deceived by a trompe-l’œil—“The only difference is, that, in one case, the solidity is real, in the other illusory.” But how can a belief be an impression produced through the eye? And how can “seeing an object,” which presupposes that there is an object to be seen, be called an affection of the mind? Even Bailey sees that this is going too far, and adds a cautionary—but, unfortunately, nonsensical—footnote: “We cannot, of course, in the common acceptation of terms, say that we see what does not really exist. . . . It is scarcely necessary to warn the reader that in this discussion geometrical solidity alone is intended”; and adds on the next page that such illusions cannot be dispelled simply by the information that they are illusions, “although we no longer infer that the appearance before us is attended with the usual accompaniments of solidity.” One sees what is intended—that the discrimination of shapes by the eye is a purely visual skill, learned by exercising the eye, and that inference is involved only when we suppose that other sorts of sensory experience would be correlated with the visual data—but Mill is hardly to be blamed if he treats such bumbling with contempt. Mill, for his part, uses the word “see” in a programmatic way that is consistent both with itself and with his project of analytic reduction: one can be said to “see” only what can be distinguished on the retinal image in the optical diagram that purports to show how the eye works. Everything else is attributed, on principle and without further ado, to the inferential and associative activities of the intellect, even though (as Bailey justly remarks) no inference can be detected, the grounds and conclusion of the alleged inference cannot be isolated, and no known or conceivable process of association could have the results claimed. Never mind. What the retinal image cannot contain must be contributed by the intellect. It is as though Mill were committing an aggravated form of the error he was later to stigmatize in his father’s Analysis, supposing that the word “see” contained a reference to the physical organ of sight as part of its meaning.
Although Mill is right when he says that Bailey’s use of the word “perception” fogs all the issues, Bailey seems justified in rejoining that to adopt Mill’s vocabulary and speak of “sensation” and “inference” would be “to adopt the theory which I controvert,” since the whole question at issue is whether the alleged distinction can be made. And he might have added that the word “sensation” itself is fraught with ambiguities. Berkeley, as quoted by Bailey, uses “sensations” to mean “objects purportedly perceived,” for he gives the sun and the stars as examples of sensations; Bailey uses “sensations” rather as some later philosophers have used the phrase “sense data”; Mill uses the word to mean “acts of sensing.” The trouble is not that the wrong words are used, but that there is no consistent use of any set of words to make all the necessary distinctions.
On the main issue, Bailey has much the best of the argument. If “distance is in reality a mere tactual conception,” it seems impossible that this could be “mistaken for a visual perception.” Surely “We cannot believe we have any particular sensation, unless we either have it or have had it at some prior period.” It is absurd to argue, as Mill does (259), that our neglect of the tactile content of our visual percepts is like our failure to attend clearly and distinctly to the meanings of familiar words: for, of course, we can distinctly recall the meaning of any familiar word, if we choose to attend to it, but we cannot by any analogous feat of attention recover the alleged tactile content of our visual impressions of distance. Mill admits that our notions of tactile space are much vaguer and less consistent than our notions of visual space. This admission in itself gives his contention that three-dimensional space is fundamentally tactile a wildly paradoxical air. Unfortunately, he also accepts that “the mind . . . does not dwell upon the sign, . . . but rushes at once from the sign to the thing signified” (257). But then, asks Bailey in triumph, “In what state must the mind be when we are looking at external objects? What is it that the understanding is engaged with? A neglected sign and an indistinct idea, between which the mind is thus bandied about, must assuredly produce a very obscure and unsteady discernment, while, in point of fact, nothing can be clearer or firmer than our perception of space in all directions, when we look round the room or out of the window.” But Bailey is still using that word “perception” in a way that begs the question at issue. The confusion is hopeless. Mill’s protégé Alexander Bain managed to clarify some of the issues, as we shall now see, but it is doubtful whether Mill fully appreciated his contribution.
BAIN AND PSYCHOLOGY
samuel johnson took care, when writing his parliamentary reports, “that the Whig Dogs should not have the best of it.” Mill and his associates took a similarly functional view of the periodical press, and the review of Bain is no more a work of dispassionate judgment than the reviews of Grote. Bain supplied Mill with up-to-date scientific data for his Logic, and reviewed it in 1843 in the Westminster, to which he had begun to contribute in 1840. In 1846 he was a summer guest of the Grotes at Burnham Beeches; Mill recommended him for the Examinership in Logic and Mental Philosophy at the University of London; Grote got him appointed to the new chair of English and Logic at Aberdeen, and supplied an account of Aristotle’s psychology for the third edition of The Senses and the Intellect as well as a history of ancient psychology and ethics for Bain’s Mental and Moral Science of 1868; Bain and Grote joined Mill in annotating James Mill’s Analysis. Bain wrote a life of one Mill and an appreciation of the other, and edited Grote’s posthumous works. It seems a small world these intellectual radicals came to move in. In the circumstances, Mill’s suggestion (342n) that he only decided to review Bain after carefully weighing the respective merits of his work and Spencer’s seems disingenuous—especially when we learn that Mill had advised Parker to publish Bain’s first volume, and joined Grote in guaranteeing him against loss in publishing the second. In the Autobiography, issued posthumously under Mill’s own name, the pretence of impartiality was dropped.
Mill’s review makes much of a distinction between a priori and a posteriori schools of psychology. Bain does not mention this distinction, and it is a puzzling one. Psychologists of both persuasions seem equally a posteriori in their methods: they seek to uphold their views by citing facts in approximately the same amounts—though not always the same facts. If anything, for reasons that will appear, it is the supposed apriorist who is more ready to appeal to experience, the self-styled aposteriorist who relies on dogma. Basically, as Mill insists they must, both follow the same method: that of reducing the complex operations of a living organism to the development in experience, in accordance with regular and predictable processes, of the simplest possible original operations. All and only what cannot be acquired must be assigned to instinct. Where the two schools differ is in what they say when confronted by a complex phenomenon of which neither can demonstrate the analysis. They then dogmatize in different directions. The apriorists, instead of acknowledging a pragmatic limit to analysis, announce the discovery of an ultimate and forever irreducible intuition or instinct; the aposteriorists invent a spurious analysis in terms of whatever entities their method postulates. Mill indeed recognizes (350) the existence of this temptation and the importance of resisting it; but in fact neither he nor Bain shows any scruples in the way they invoke the “principle of association” which Mill claims “extends to everything” (347).
As an example of the divergent dogmatisms of the two schools we may consider the alleged infinity of time and space. Apriorists, Mill says, claim that the mind’s belief in this infinity is instinctive, on no better ground than that nothing in our experience can be infinite; but the true explanation is not far to seek. Because we have no experience of a spatial or temporal point without neighbours, whenever we imagine such a point we imagine it (from force of habit) as neighboured; hence, we can imagine no limits to time and space, and therefore find their finitude unthinkable and call them “infinite” (345-7). But, we may ask, from what experience does the alleged association proceed? We know places where there are things, and places where there are no things; times when things happen, and times when nothing much happens. But in what sort of experience do moments of time and points of space, as such, form elements? What is supposed to be the difference between a time when time ends and a time after which there is infinite time in which nothing happens? The alleged extrapolation from “experience” seems plausible only if one allows virtually any relation between any sort of real or ideal units to count as a case of “association.”
Bain, though this hardly appears from the review, was not prepared to fudge his psychology as Mill did. His attempt to anchor his associationism to the physiology of the nervous system effectively prevented him from doing so. The “chemical union” which Mill praises Hartley for introducing (347) allows one, as Bailey complained, to use the term “association” of almost any form of explanation that relates an experience to previous experiences or alleged constituents. Bain’s speculative account of the processes of the nervous system eschews such vagueness. Knowledge is produced by the accumulation of patterns of electrical discharges, each of which records something known and figures in memory simply by being repeated. The patterns can combine mechanically, but cannot fuse. Bain is thus committed, as Mill was not, to the programme of actually discerning and disentangling the elements whose association is postulated. In the end, this scrupulous atomism makes associationism implausible by multiplying the required number of brain traces beyond credibility; but at least we can guess what form an associationist explanation should take, which with Mill remains forever mysterious.
It was because Mill admitted “chemical” unions that he could with a good conscience invoke the “complete Baconian induction” whereby the apparently visual phenomena of distance are shown to be ultimately tactual in purport. As early as Leibniz’s Nouveaux Essais apriorists had complained against the aposteriorist assumption that an “innate” faculty must be one manifested in infancy: the point, they said, is not the moment in time at which an ability is first displayed, but whether it admits of being analyzed without remainder into elements previously given in experience. But to say that a “chemical” union has taken place is to admit that such an analysis is in principle impossible; the issue can then be settled, if at all, only by appealing (as Mill does) to the circumstances in which an idea is first manifested.
It is true that Bain, who lacks Mill’s excuse, himself sometimes makes the assumption against which Leibniz complained. But Bain’s logical acumen was rather blunt. In fact, though a man of great learning and industry and a strong sense of fact, he had little gift for philosophical analysis. The pieces of his work are generally sober and well-informed, but are not always consistent among themselves; and this is nowhere more evident than in what he says of visual perception. In the Book on “The Senses,” where his account is firmly linked to the physiology of eye and brain, the Berkeleian doctrine of the priority of touch has no place, and colour appears among visibilia as merely one of the means of differentiation of visible objects. But in the Book on “The Intellect” colour is back in its old place as the unique visibile, and statements requiring the Berkeleian doctrine are interspersed with others more compatible with the doctrine worked out earlier.
Yet there can be little doubt as to what the overall theory is to which his account tends, and it differs far more widely from Mill’s than Mill is aware, even though the difference is less plainly marked in the first edition than in the later revisions. First, the “retinal image” as the static quasi-object of vision vanishes, and with it vanishes the independent significance of the findings of optics. “The optical sensibility does not give even visible form”; the visual presentations at any moment “are but the hint to a mental construction” to which we carelessly attribute the qualities of a static picture; in fact, temporal and spatial distinctions are revealed by movements, those involving vision being not parasitic upon but parallel to those involving touch. The “suggestion of locomotive effort” is at the heart of our sense of real distance; but the notion of extension “when full grown is a compound of locomotion, touch, and vision, any one implying and recalling all the others.” Thus “extension, or space, as a quality, has no other origin and no other meaning than the association of these different sensitive and motor effects.” Here is Berkeley’s Berkeley, restored to intelligibility, with the opticians’ Berkeley relegated to limbo at last; and Bain is able to recognize without embarrassment that our spatial sensibility incorporates such ineluctably visual elements as a sense of expansive compresence.
Another vast area of confusion vanishes on the very first page. The old controversy had assumed that there was a problem about how we get from subject to object, from inner experience to outer world. But now we read that subject and object, outer and inner, are concepts that can only be acquired in contradistinction from each other; and the baffling talk about “inside the mind” and “outside the mind,” as though the mind were the skull, is set aside just as firmly, though rather less clearly. This done, we are free to return to the commonsense view that the notion of an external world rests on the experience of resistance to our bodies —a factor which, Bain noted with mild surprise as soon as it was safe to do so, Mill was “almost singular” in overlooking.
Mill notes as one of the merits of Bain’s work that it rests on a solid account of neural activity. But his version of the theory of vision, in which he virtually claims that Bain agrees with him despite some over-emphasis on the activity of the eye muscles, shows how far he is from appreciating the difference this makes. For example, Mill still feels able to talk about the retinal image as a picture. But, whether Bain was aware of it or not, the ground rules for such discussions had changed. From now on, one had a choice. Either one took account of the central nervous system, in which case the old-fashioned compartmentalization of the senses became irrelevant, or one confined oneself to epistemology and phenomenology, in which case Mill’s style of generic analysis became inappropriate. Psychology could never be the same again.
Mill remarks shrewdly on the different levels on which the different sections of Bain’s work proceed. The Book on “The Emotions,” as he justly observes (361), is no more than a natural history; that on “The Will” is a sustained effort at reductive analysis in the old style of James Mill and the eighteenth century generally. But the Book on “The Senses” belongs to a new age, in which psychology would be turned into a positive science by recognizing that its first task was to establish what is in fact the case. Because of this disparate character of its parts, Bain’s work could be regarded equally as a late production of speculation or an early product of science. Mill, naturally enough, can see it only as a continuation of his own work with new aids. Yet we should be careful not to make too much of the differences between the two men. There is one essential point in the theory of vision that is common to both: that visual data are originally “signs” whose interpretation must be learned in experience and whose meaning is to be explicated in terms of experience. This point holds true whether or not new-born animals have inborn tendencies to react to stimuli, of whatever kind.
It is in the Book on “The Will,” as Mill suggests (354), that Bain shows his originality as a psychologist of the old school by trying to reduce all the phenomena of animal action to the terms of a new and very simple model: all skills are acquired by the modification of an original entirely random and generalized activity of the nervous system and hence of the muscles, and the modification is effected by simple reinforcement or inhibition through pleasure or pain. This implausible model seems to rely excessively on the singular helplessness of the neonate human. It is curious to see the enthusiasm with which Mill seizes on the description of the new-born lambs (358-9), a description which is made to support Bain’s case only by the observer’s gratuitous insistence on the randomness of the motions he describes. One wonders how many lambs would survive if their lives depended on such a series of chance contacts as is here supposed, without any initial tropism or IRM’s. And one wishes Bain had indicated how he would have accounted for Bailey’s new-born turtles, trekking to the sea. It is precisely in this sort of model-building that Bain is weakest. It is his combination of unimaginativeness and implausibility (together, of course, with the obsoleteness of all old science) that explains why we no longer read him but still read Hobbes, who knew so much less but suggested so much more.
Mill gravely understates (364) the oddness of one aspect of Bain’s account of volition, his version of the development of the moral ideas. This is very different from Mill’s own. Bain thinks of morality as wholly negative, a system of inhibitions built up in the first instance entirely by corporal punishment. His dourness seems appropriate to the reputation of the calvinist and granitic city where he spent most of his life. From this point of view Mill’s utilitarianism is not a theory of morals at all, but of something else. The phrase “moral approval” is explicitly called a misnomer, on the grounds that only disapproval can be moral.
I have already noted Bain’s account of the origin of our sense of the externality of the world. He ends his work by citing an account of the physical world tantamount to Mill’s notorious formula, “permanent possibility of sensation”—cited, not from Mill, but from the Idéologie of Destutt Tracy. This fits in with the basic principle of Bain’s work, played down in Mill’s review: the principle of relativism, that consciousness can only be consciousness of differences and changes. It follows immediately from this principle that the mind can have no knowledge of any “absolute.” This relativism goes naturally with the discovery of the ceaseless activity of the nervous system, and it is plainly hard for Mill to adjust to it. The older philosopher is hampered by the empiricist traditions of atomism and reification, which turn the mind into a warehouse of ideas, and knowledge into an assemblage of separate facts about separate things. Thus he has little to say (beyond a faint protest) about Bain’s doctrine of belief. “As, in my view,” says Bain, “Belief is essentially related to the active part of our being, I have reserved the consideration of it to the conclusion of the Treatise on the Will.” To believe anything is to act as if it were the case; hence, by extension, to have a propensity to act so; or, in cases where (as in believing that one would have enjoyed living in ancient Rome) no prospect of action arises, to be in a disposition that would have led to action had the occasion arisen. Mill is understandably puzzled to understand how such a position could be consistently developed and defended, and it must be admitted that (like much of Bain’s work) it raises no fewer problems than it purports to solve, but it is at least clear that it forms part of a philosophy of process in which Bain feels so much at home that the details of its statement do not trouble him much. Precisely the same difference in mental set appears in Mill’s later exchange with Bain on the subject of “potential energy.” Mill, with impeccable logic, and citing Hamilton for his definitions, points out that what is called “potential energy” is really potential motion. If it is anything at all it is a real force; but it seems to be postulated only as a fiction, to reconcile the observed phenomena with the dogma of the conservation of energy. But working scientists are notoriously insensitive to considerations of this sort.
Mill’s support (365-7) of Bain’s determinism also conceals a difference in approach, though not one that Bain emphasizes. Bain, like Mill, allows no validity to the “consciousness” of freedom, and for the reasons that Mill gives. But what is hidden from consciousness for Bain is not best described, as it is by Mill, as a hidden law obeyed by our volitions. For Bain, mental and neural phenomena run in parallel and do not interact. To every mental state answers a brain state. And the brain is an electrical machine, whose later states are accordingly a function of its earlier states and inputs. In fact, as many later writers were to point out, consciousness in Bain’s theory is fundamentally misleading.
Mill’s comment on Ruskin’s inadvertent aposteriorism in Modern Painters, apparently a casual aside in his treatment of the classification of emotions, is more important than it looks. Aesthetic feelings and artistic practice have been strongholds of apriorists at least since Hutcheson published his Inquiry, and perhaps since the neoplatonists and Plato himself: a sense of beauty seems to resist derivation from or analysis into any other mental phenomenon. Bain spends a surprising amount of space on various attempted reductions, but he shows little aptitude for these topics, and his later editions rely with relief on the authority of Sully. Ruskin, though some disparage his taste and reasoning power, has never been accused of deficiency in the amount of his aesthetic sensitivity. Perhaps Mill (himself found deficient in such sensibility—by Bain!) is hinting that it is not only philistine Scots who are prepared to reduce the aesthetic sense to a more general form of susceptibility.
What Mill says about the necessarily negative nature of the evidence for apriorism (349) sheds some light on a puzzling argument in Utilitarianism. Mill there argues that questions of ultimate ends do not admit of proof in the strict sense, but that the fact that each man desires his own happiness affords the sole possible proof that the happiness of all men is desirable. Apparently, then, Mill is speaking of “proof” in some restricted or metaphorical sense, but it is not clear just what this sense should be. In the present passage, where Mill uses almost the same language, the nerve of the argument is exposed: that the failure to disprove a thesis for which there is prima facie empirical evidence, though not proving the thesis true, must be allowed to serve in lieu of demonstration in all cases where it is logically impossible that anything better could be found. That would make good sense in the Utilitarianism passage. The things that ought to be desired must be among those things that can be desired, and the only logically possible way of showing that a thing can be desired is to show that it is in fact desired. “Ought” implies “can,” and possibility is parasitic on actuality. It always remains logically possible that someone should discover an actual, and hence possible, and hence possibly proper, object of desire that is not reducible to a component of happiness; but no one has yet, despite all endeavours, managed to do so. Until they do, some form of utilitarianism must hold the field. It is not, admittedly, clear that Mill means to argue to this effect in Utilitarianism, but his general aim of making the moral sciences truly scientific would lead us to expect him to follow the same lines in ethics as in psychology generally. And, just as in general psychology his analyses are rendered nugatory by his admission of canons of association according to which anything may be “chemically” analyzed into anything, so in ethics his argument becomes trivial because (as Bain complained) even the most self-abnegatory actions are interpreted as self-seeking through an analogously magical sort of transformation.
TAINE AND THE UNDERSTANDING
taine was forty-two when Hachette published his grande pâtée philosophique in April, 1870, but he had been meditating it for twenty years: a theoretical underpinning for the historical works by which he is better remembered. If one can show that all knowledge comes from experience, differences in style should reflect differences in experience: literary, artistic, and social histories should be explicable in terms of cultural traditions that could in turn be ultimately explained by such factors as climate and terrain—Taine had nothing but contempt for George Grote, whose history treated politicians as free agents.
Like Mill, Taine was something of an outsider in relation to the cultural establishment of his country. But whereas Mill and associates could use the forces of Scottish irredentism and northern nonconformism, not to mention the private empire of India House, and set up University College in Gower Street to be a counterweight to the port-sodden churchmanship of the ancient universities, Taine was up against a more tightly knit and centrally controlled cultural empire. Outside the official establishment there was nothing: one had to choose between taking it over oneself, and finding a niche in which to pursue (as so many French intellectuals have done since) one’s private intellectual aims in the abundant leisure its ample rhythms afforded. As a youth, Taine was denied the prizes and professorships he sought, being thought too flashy in his brilliance and too unstable in his politics: but at thirty-five his growing literary fame won him appointment first as Examiner to Saint-Cyr and later (in succession to Viollet-le-Duc) as professor of aesthetics at the École des Beaux-Arts—though even then an attempt by the military authorities to terminate the controversial appointment had to be circumvented by the interposition of the Emperor himself. Nonetheless, the two men shared a feeling of being in an embattled minority. Thanking Taine for his series of articles on the Logic in the Revue des Deux Mondes, Mill says that when he began to publish he was almost alone in his views, and that even now the empiricist philosophers were outnumbered twenty to one; while Taine predicted that his own psychological work would find only a hundred readers in France and a hundred in the rest of Europe.
Both in his review and elsewhere, Mill treats this split in the philosophic community as following national lines, suggesting that both Cousin’s idealism and Hamilton’s apriorism, against which Taine and he were pitting themselves, were Germanic in inspiration. But his letter to Taine repudiates this ascription of national affiliations to schools of thought. The French think of empiricism as peculiarly English, the English as typically French. In reality, intuitionism and empiricism are related dialectically: the dominance of either calls forth the other as its antithesis. Which of them happens to prevail in any particular milieu at any particular time is quite fortuitous; at the time of writing, Germany itself is swinging towards the empiricist pole. And Taine himself at one time spoke of Mill’s philosophy as a re-working of Kant.
Mill’s review of Taine, like that of Bain, is not the first meeting of the two minds, but an episode in a long relationship. At first, Taine had not been deeply impressed: he found Jowett more progressive. “On vante beaucoup ici,” he wrote in 1860, “la Logique de Stuart Mill et la Psychologie physiologique de Bain. Il y a du mérite, mais ce ne sont pas des génies.” But in 1861 he devoted to the Logic a series of articles which he later published as a monograph, and in the preface to the latter version he sings to another tune: “En ce moment, la scène est vide en Europe. . . . Dans ce grand silence, et parmi ces comparses monotones, voici un maître qui s’avance et qui parle. On n’a rien vu de semblable depuis Hegel.” Mill acknowledged the accuracy of Taine’s account of his views, which appeared yet again as part of the History of English Literature; and much of it was incorporated, sometimes with little change even in the wording, in Taine’s own account of induction in De l’Intelligence. It is therefore not surprising that Mill finds little in this part of the book to quarrel with.
Mill’s review, with its reference to the foundation of knowledge on images, might mislead the uninitiated into thinking that Taine’s background in associationism is the antiquated French ultra-Lockianism of Condillac, rather than the North-British neo-Hartleianism of the Mills and Bain. Taine himself thought otherwise: his original Preface acknowledges a debt to Condillac for one point only, and claims Mill, Bain, and Herbert Spencer as his chief creditors. The fourth edition of 1883 supplements this general avowal with three specific acknowledgements: to Condillac, for the theory that all general ideas “se réduisent à des signes”; to Mill, for the theory of induction; and to Bain, for the account of the perception of space. A letter of January, 1873, gives the reason for this explicitness: the British had treated his book as a mere re-hash of their own work. He virtually accuses Spencer of plagiarizing his views for the revised edition of Principles of Psychology and falsely claiming that Taine got them from him. He continues:
Il dit dans sa seconde préface que L’Intelligence “a fait connaître en France quelques-unes de ses maîtresses conceptions.” Cela est inexact. Ceux à qui j’ai emprunté sont John Stuart Mill et Bain (Induction, sensation musculaire donnant l’idée de l’étendue), et je les ai cités tout au long. Je n’ai emprunté à Spencer qu’une phrase. . . .
Pardonnez-moi ces revendications; je me suis aperçu en lisant les Revues anglaises que l’on faisait de mon livre une simple imitation, une transcription française des théories anglaises.—M. Stuart Mill, dans un article de juin 1870, a bien voulu que mon travail était entièrement original, et, à mon sens, cela se voit par la méthode employée, par les théories de détail et par les théories d’ensemble.
The “article of June 1870” is of course the review included in this volume, from which, unlike Taine, I would have gathered that Mill was less impressed by the book’s intrinsic merits than by its significance as portending a possible change in the climate of French opinion. Certainly he specifies no respect in which the book has advanced the study of its subject, treating its chief departure from his own views as a mere abandonment of the book’s own principles. But were the abandoned principles Taine’s, or Mill’s? A letter of 1872 suggests that Taine’s intentions were far from empiricist, for one of the chief matters in which he claims originality is his metaphysical reduction of the individual to a mere series of events, “tous les événements de la nature n’étant que des formes diverses de la pensée.” And his earlier essay on Mill had strikingly contrasted Mill’s approach with his own: “This theory of science is a theory of English science. . . . The operations, of which he constructs science, are those in which the English excel all others, and those which he excludes from science are precisely those in which the English are deficient more than any other nation. He has described the English mind whilst he thought to describe the human mind.”
Acknowledging Taine’s thanks for his review, Mill apologizes both for its brevity and for its uninformativeness. “Je sais combien cette notice est insuffisante mais j’ai voulu, au premier moment possible, attirer l’attention des hommes éclairés sur un livre dont la publication en France me paraît destinée à faire époque. Votre livre n’a pas besoin d’être interprété. Il suffit qu’on le lise, car vous possédez parmi tant d’autres qualités, le génie de la clarté.” And he goes on to explain more fully where he differs from Taine about the status of axioms. His account of this doctrine had indeed been compressed to the point of unintelligibility, and seems in fact to have been derived from the earlier and cruder version in the monograph on Mill (according to which “abstraction” affords “an intermediate course between intuition and observation, capable of arriving at principles, as it is affirmed that the first is, capable of arriving at truths, as we find that the second is”) rather than from the more refined version adumbrated in De l’Intelligence. Taine’s mature doctrine seems to be as follows. The empirical concepts and generalizations reached by induction, even when based on intelligible relationships and not merely on observed regularities, can never be extrapolated to remote situations with more than probability (449). But the concepts that figure in the axioms of the exact sciences are not so much abstractions from experience as anticipations of experience, ideals to which experience can never be shown to conform (414). The laws of the exact sciences are disguised analytic statements, depending for their truth on the analyses and reconstructions on which the concepts contained in them ultimately depend (485). The laws of geometry and mechanics therefore have to do not with actual but with possible things. Their axioms depend not, as in Mill’s empiricism, on likenesses recognized through an associative process, but on the identity of formal properties (480-6). In explicit contrast with Mill, Taine opposes the perceived likeness of two geometrical figures to the recognized identity of a geometrical construction. The repetition with which science deals is identical recurrence and not repeated likeness: we can thus be certain that identical causes will have identical effects, and in this sense the principle of induction is proved. But it is for experience to decide whether what we are confronted with is the same cause (540); scientific laws are universally applicable, but it is for observation to decide when they are exemplified (484-6). This position is indeed, as Taine claims, very far from Mill’s. He agrees with Mill against the Germans in going from the particular to the general, instead of starting with a Weltanschauung and hoping that there will be somewhere for the chips to fall; but his work cannot be brought within the boundaries of associationism. It seems to foreshadow the more sophisticated empiricism of such theorists as Nagel, for whom a scientific theory has the “necessity” of a mathematical equation but needs to be supplemented by less formal understandings as to how far any real situation may be deemed to conform to its specifications. In particular, Mill seems to be wrong in accusing Taine of exploiting the ambiguity of the concept of sameness: on the contrary, his theory rests on contrasting resemblance with identity. But, although Mill may have missed the point of Taine’s main argument, what he says is perfectly true of some of the incidental discussions. In a passage on geometrical proofs, Taine does indeed confuse identity with exact likeness, and derives the mathematical concept of equality from just this ambiguous notion of “the same.”
Perhaps from sheer incredulity, Mill disregards Taine’s point that axioms about triangles are always valid and would always be applicable even if nothing came near enough to being triangular for this applicability to be very useful. The difference between Taine’s language and Mill’s is instructive: Mill, in his letter, speaks of the concepts of the exact sciences as idealizations of experience; Taine calls them anticipatory constructions. It is this seemingly trivial difference in terminology that enables Mill to say, “if the concept itself is the product of experience, the truth of the properties comes to us from the same source” (446). Taine, like most later thinkers, regards concepts as constructs rather than as products; and, if “product” were indeed the right word for them, whether the consequences alleged by Mill would follow must depend on the manner of their “production.” Here again, however, Taine’s carelessness or inconsistency lays him open to an objection that Mill makes more clearly in his letter than in his review. Even if one admitted the a priori character of such concepts as that of a straight line, he says, so that its properties were revealed to intuition rather than drawn from experience, “on peut dire que cette observation directe ne pourrait nous révéler que les propriétés du produit regardé comme conception mentale, c.à.d. des faits psychologiques, et qu’elle ne nous dit rien sur les lois générales de l’univers.” For Taine had written, “The propositions of these sciences are not merely probable but certain beyond our little world; at all events, we believe it to be so, and, moreover, are unable to believe or conceive that it is otherwise” (450). Is this not just that “inconceivability of the opposite” whose adequacy as a test of truth Mill had challenged fourteen years before? Not quite, perhaps, for what Taine says we cannot doubt is not that something is true but that its truth is necessary. But what sort of necessity is he really invoking? Logical, or merely psychological? Mill would concede the latter but deny its relevance. A mere habit of expectation has no evidential force, and Taine’s programme had been to substitute something stronger. Nor is this a momentary lapse of Taine’s pen. Years later we find him affirming that Kant’s question about the possibility of synthetic judgments a priori is a psychological one, to be settled by observation in the manner of Bain and Mill (not to mention himself), and that such observation shows them to be of two sorts. Some are disguised analytic statements: “Les autres ne sont pas valables; ils ne sont que des généralisations ou des anticipations de l’expérience; a priori, ils sont dépourvues [sic] de toute autorité; l’autorité qu’ils ont leur est conférée toute entière a posteriori par les expériences qui les confirment.” What sort of psychological test could show whether a statement is a generalization from experience or an axiom in a deductive science one cannot imagine, and Taine’s attempt at a novel solution to Kant’s problem breaks down after all in total confusion. If Taine really did wish to found his epistemology on psychology, Mill was right after all: the claim of unrestricted validity for the axioms is as false to Taine’s principles as it is to Mill’s, though not exactly in the way Mill has in mind.
MILL AND THE OPEN MIND
Scattered and occasional as they have been, our remarks seem to have tended after all towards one general conclusion. Mill prided himself on his open-mindedness, and Bain concurred. But on the topics covered in this volume this claim seems hardly justified. We saw him missing the main points in Bailey, misrepresenting Bain, using Grote as a peg to hang his own pet notions on, scrutinising Taine merely for possible agreements and disagreements, and professing, at the start of his review of Bain, an impartiality between schools of psychology that the associated correspondence belies. Again, though early a champion of traditional formal logic against the psychologizers, he was so far from seeing the significance of the transformation of logic that began with Boole and was already under way in his middle years that Jevons could see his prestige as the main obstacle to logical reform. This judgment casts no discredit on Mill. A man of his precocity cannot be expected or required to be an innovator in old age, and the head-start of twenty-five years that he claims his father’s forcing methods gave him could end by leaving him with too much to unlearn. Besides, open-mindedness is not soft-headedness. A man, unlike a government, is not called on to condone manifest errors, and all the incidental blindnesses and dogmatisms we have noted stem from his resolute opposition to a doctrine he believed to be fraught with immediate moral and political dangers. All the same, a tension remains between the dogmatism he shows and the receptivity he claims. That this claim is so widely conceded is partly to be accounted for by the marvellous, almost hypnotic, breadth and equanimity of his expository style: his unexampled air of unruffled comprehensiveness and imperturbable reasonableness. Bain, a dull writer, completely missed this quality: “The language faculty in him was merely ordinary,” he says. But Mill himself knew how much he owed to the discipline of the civil service, which taught him so to cast a controversial minute that its recommendations would seem acceptable and even inevitable to his reluctant masters. Alan Donagan has commented on the perfect expressiveness of Mill’s controversial style, in which passion never appears as a fatty layer over the sinew of argument; but in taking this wiry force as index of a sincere heart he fails to note that it may represent a dexterity that distracts the eye from the workings of a devious mind. When we consider the great speed at which some of these pieces were written we can only be astonished at the smooth force with which facts and arguments seem to conspire together in a natural order to draw Mill’s conclusions for him. Only an independent reference to the books reviewed and the facts alleged can reveal the strong acids that were needed to blend such heterogeneous nutrients.