John Stuart Mill (1806-1873) was the dominant figure of political economic thought of his time, as well as being a Member of Parliament for the seat of Westminster, and a leading political philosopher.
John Stuart Mill was a British philosopher, economist, moral and political theorist, and administrator, was the most influential English-speaking philosopher of the nineteenth century. His views are of continuing significance, and are generally recognized to be among the deepest and certainly the most effective defenses of empiricism and of a liberal political view of society and culture. The overall aim of his philosophy is to develop a positive view of the universe and the place of humans in it, one which contributes to the progress of human knowledge, individual freedom and human well-being. His views are not entirely original, having their roots in the British empiricism of John Locke, George Berkeley and David Hume, and in the utilitarianism of Jeremy Bentham. But he gave them a new depth, and his formulations were sufficiently articulate to gain for them a continuing influence among a broad public.
John Stuart Mill was born in Pentonville, then a suburb of London. He was the eldest son of James Mill, a Scotsman who had come to London and become a leading figure in the group of philosophical radicals which aimed to further the utilitarian philosophy of Jeremy Bentham. John Stuart Mill's mother was Harriet Barrow, who seems to have had very little influence upon him. James Mill's income was at first slight, as he struggled to make his living as a reviewer. But his History of India secured him a position in the East India Company and he rose to the post of Chief Examiner, in effect the chief administrator of the Company. In spite of these duties and the work they entailed James spent considerable time on the education of his eldest son. The latter began to learn Greek at three and Latin at eight. By the age of fourteen he had read most of the Greek and Latin classics, had made a wide survey of history, had done extensive work in logic and mathematics, and had mastered the basics of economic theory. This education was undertaken according to the principle of Bentham's associationist psychology, and aimed to make of the younger Mill a leader in views of the philosophical radicals.
At fifteen John Stuart Mill undertook the study of Bentham's various fragments on the theory of legal evidence. These had an inspiring influence on him, fixing in him his life-long goal of reforming the world in the interest of human well-being. At eighteen he spent considerable time and effort at editing these manuscripts into the long coherent treatise that they became in his hands. Guided by his father he threw himself into the work of the philosophical radicals, and began an active literary career. Shortly thereafter, in 1823, his father secured him a junior position in the East India Company. He rose in the ranks, eventually to occupy his father's position of Chief Examiner. A visit to France in 1820 had made Mill thoroughly fluent in the language, and he became a life-long student of French thought and history.
In 1826, Mill suffered a sudden attack of intense depression. This lasted for many months. He continued his work, and indeed even his political activities, but internally he felt that his former goals were without worth. He came to believe that his capacity for emotion had been severely weakened by his father's rigorous training in analysis. His intellect had been educated but not his feelings. In the reading of Wordsworth's poetry he found something of the cure that he needed, and the depression gradually disappeared.
Mill met Gustave d'Eichtahl in 1828. D'Eichtahl was a follower of St. Simon, and introduced Mill to the latter and to works of Auguste Comte. Mill also met John Sterling who was a disciple of Coleridge. Through these thinkers Mill came to appreciate the role of social and cultural institutions in the historical development of human beings. He became convinced of the Comtean view that social change proceeds through “critical periods,” in which old institutions are overthrown, followed “organic periods,” in which new forms of social cohesion emerge and are consolidated. He came to believe with these French thinkers that in his own time society was emerging from a critical period. From Coleridge he came to view the educated class as the vehicle for ensuring social cohesion in the emerging organic period.
Mill now saw his task to be that of helping British society to emerge into the coming organic period. The merely negative polemics of Bentham and his father now seemed very limited. It became necessary not merely to criticize older forms of social organization but to work towards replacing them with something better. Moreover, the defenders of older forms of life should no longer be dismissed as representatives of vested interests. The very fact that the older forms had so long survived meant that they had some good in them, and their defenders should be seen not as reactionaries but as those who continue to recognize that good. If that good is now limited, it still must be acknowledged, and not merely dismissed.
Tactically, the social reformer in critical periods cannot proceed by formulating grand philosophic schemes, however correct they may be in principle. Rather he or she must work for piece-meal reform. Only gradually should general principles be proposed, so that the appearance of radical novelty will be avoided. Mill never abandoned his earlier acceptance of the principle of utility, but now used it positively, not just critically and destructively; he emphasized how it could be deployed constructively, enabling new forms of society to emerge, but ones which incorporate the best of the older forms. He now came to think that the democratic demands of the older radicals had to be tempered with a concern for the dangers which it posed for individualism.
In 1830 Mill was introduced to Harriet Taylor. Her husband was a druggist whose grandfather had once been a neighbor of James Mill. The younger Mill rapidly became intimate with Mrs. Taylor, who came to influence profoundly the rest of his life. She was an invalid who lived apart from her husband. The latter, while he lived, remained remarkably tolerant of the Platonic but very close relationship that Mill and his wife maintained. Mill's father highly disapproved of the connection. When Mill married Mrs. Taylor in 1851 two years after her husband's death there was a complete estrangement from his mother and sisters. Mill reports in his Autobiography that Harriet was of crucial significance to his intellectual and moral development. His father, though an Epicurean in principle, was in fact one who eschewed pleasure. His ends were rather a sort of rigid discipline of work: he never did free himself from the confines of Scots Calvinsim. It was reading Wordsworth that gave the younger Mill a sense of greater human possibilities. But it was Harriet Taylor who kept this sense alive, and continually strengthened his conception of the real end of human being is the progressive development of individuality in all, in women as well as men, in workers as well as aristocrats. It was she who imbued Mill with the sense that if these ends could be developed then it would become clear that at present human beings were enjoying only a small fraction of the happiness that was possible. It was she who gave Mill the expansive idea of the good that was to form the new utilitarian end that replaced the rather rigid and narrow ends described by Bentham and his father. As Mill insisted, she was indispensable to his later thought. He was nearly inconsolable when she died. It was during a trip to Europe in 1858 that she fell fatally ill and died at Avignon, where she is buried. For the rest of his life, Mill spent half a year at a house in Avignon so that he could be near to her grave.
Mill remarks in his essay on the Subjection of Women that:
The most favourable case which a man can generally have for studying the character of a woman, is that of his own wife: for the opportunities are greater, and the cases of complete sympathy not so completely rare. And in fact, this is the source from which any knowledge worth having on the subject generally comes. But most men have not had the opportunity of studying in this way more than a single case: accordingly one can, to an almost laughable degree, infer what a man's wife is like, from his opinions about women in general. To make even this one case yield any result, the woman must be worth knowing, and the man not only a competent judge, but of a character sympathetic in itself, and so well adapted to hers, that he can either read her mind by sympathetic union, or has nothing in himself which makes her shy of disclosing it. Hardly anything can be more rare than this conjunction. (Subjection of Women, Chapter I)
This gives a sense of how Mill felt about his own relationship with Harriet Taylor.
In 1823 Mill had entered the employ of the East India Company as a clerk. India was governed by the Company through correspondence between the Court of Directors in London and the Governors on the subcontinent. This was supervised by the Office of the Examiner of Indian Correspondence. Mill rose through the ranks, and for many years was in charge of the correspondence dealing with the princely states not under the direct rule of the Company. In 1856 he became Chief Examiner, in charge of all correspondence, succeeding another utilitarian, Thomas Love Peacock, who had succeeded James Mill. After the Indian Mutiny, the British parliament proposed the dissolution of the Company. Mill prepared a vigorous defense of the Company and of the government that it provided, but it was unsuccessful. He was offered a position on the new advisory council, but he declined. He retired on a reasonable pension in 1858.
In 1865 Mill was elected to the House of Commons. Given his reputation and his previous seclusion, his work was subject to immense attention. His performance was generally acclaimed, but he found himself at odds with the aims of his electors. He was unwilling to compromise his own principles, and as a consequence he failed in his attempt at re-election in 1868. He continued to work, as he had earlier in his life, for many radical causes. He was particularly concerned for the status of women. His later work was made easier by the cooperation of Mrs. Taylor's daughter, Helen, who in many respects took the latter's place in Mill's life. A number of his important works were published posthumously by Helen Taylor. Mill died in 1873 at Avignon, where he is buried next to his wife.
Mill made his philosophical reputation with his System of Logic, which he published in 1843; this work re-vitalized the study of logic, and provided for the remainder of the century the definitive account of the philosophy of science and social science. This was followed by The Principles of Political Economy in 1848; this defined the orthodox form of liberal principles for the next quarter century. In 1861 he published his only systematic treatise in first philosophy. This was his Examination of Sir William Hamilton's Philosophy, a comprehensive critique of the latter's rationalism and intuitionism. So effective was Mill's critique that this work effectively dated itself and is now unfortunately neglected. His two best-known works in moral philosophy were On Liberty and Utilitarianism, which appeared in 1859 and 1861 respectively. These are of continuing significance. His Considerations on Representative Government, published in 1851, is perhaps now less important than his essay on The Subjection of Women (1869). Mill's partially finished Autobiography was published, with additions by Helen Taylor, in 1873. She also saw for the posthumous publication in 1874 of his Three Essays on Religion.
John Robson has created a monument to Mill in his edition of Mill's Collected Works. All the Mill material is here, admirably introduced by various authors. Besides the standard works, and such things as his newspaper articles, there is also his correspondence. Among the latter there are many gems to be found. For example, there is his remark in his speech on the Reform Bill of 1866, where he proposed and defended universal suffrage and votes for women; this remark is in response to a comment by Conservative MP Sir John Pakington on a remark Mill had made elsewhere:
What I stated was, that the Conservative Party was, by the law of its constitution, necessarily the stupidest party. Now, I do not retract that assertion; but I did not mean to say that the Conservatives are generally stupid. I meant to say that stupid people are generally Conservative. I believe that is so obviously and universally admitted a principle that I hardly think any gentleman will deny it. (Public and Parliamentary Speeches, 31 May 1866, pp. 85-86.)
One finds Mill elsewhere also saying that
Stupidity is much the same the world over. A stupid person's notions and feelings may confidently be inferred from those which prevail in the circle by which the person is surrounded. Not so those whose opinions and feelings are emanations from their own nature and faculties. (Subjection of Women, Chapter I, p. 273)
The volumes are a pleasure to work with. The editing is excellent. The variants are unobtrusively noted, and one can trace with ease the various changes that appeared in later editions of such works as the System of Logic and the Political Economy. Students of Mill cannot but find these volumes invaluable: we are indebted to the editor for a job well done.
2. Language and Logic
In the System of Logic, Mill accepts the traditional doctrine that propositions as used to describe the world are divided into subject and predicate terms, or, as he would say, names, joined by a copula, either affirmative or negative. Among names there are singular and general. All names denote, either individuals or the attributes of individuals. Names are either singular or general. A general name connotes an attribute and denotes all individuals which have that attribute. Thus, ‘white’ connotes the attribute whiteness, and denotes all things that have that attribute. Some singular names only denote; they have no connotation. ‘Caesar’ is such a name. However, many singular names not only denote, they also have a connotation. Thus, ‘the conqueror of Gaul’ is a singular name. It denotes the same individual as is denoted by ‘Caesar’, but unlike ‘Caesar’ it also connotes; it connotes the attribute of being a conqueror of Gaul. In terms of logic, Mill is less concerned that later thinkers would be about the uniqueness implied by the connotation: Caesar is not only a conqueror of Gaul but the conqueror.
In a proposition, names are joined by a copula, either affirmative or negative. The meaning of a proposition—its “import”, as Mill says—is determined by the connotation of its parts, the sole exception being given in the case of proper names, where the meaning is determined by the denotation.
Where the import of a proposition is given by connotation, truth or falsity is determined by denotation. An affirmative proposition is true just in case that the thing or things denoted by the subject term are in the class of things denoted by the predicate term; otherwise it is false. Similarly, a negative proposition is true just in case that no thing denoted by the subject term is a member of the class of things denoted by the predicate term. Things and attributes are always such that any proposition is either true or false and not both. This states the Principles of Non-contradiction and of Excluded Middle. No thing or attribute is such that it can be said to be both wholly itself but also necessarily connected to something other than itself: each thing or attribute is logically and ontologically independent of every other thing or attribute.
Since the holding of these principles depends upon the systematic nature of things and their attributes, it follows that the truth of these principles is in the end a fact about the world and the things in the world—the deepest, perhaps, of all metaphysical facts about the world and the entities in it.
The world we talk about in our propositions is the world that we come to know in our ordinary sense experience or inner awareness. The ontology of the world as reflected in language and logic is the ontology of the world as we know it to be. Knowing the meaning of terms and therefore of the import of propositions is knowing the individuals and attributes which they denote and connote. As we know them in our ordinary experience these individuals and attributes are logically self-contained. The rationalists and the Aristotelians argued that beyond our ordinary experience of things we have intuitions—“rational” intuitions—of ontological connections that structure things in ways not apparent in our ordinary sense experience of the world. Empiricism is the claim that there is no such rational intuition and nothing in the ontology of the world beyond what we know in ordinary experience. The world of the empiricist is one without necessary connections among individuals and attributes. It is this deep fact about the world that is the metaphysical basis of the truth of the Principles of Non-contradiction and of Excluded Middle.
Mill is thus arguing that, while there are no objective necessary connections, there is nonetheless an objective basis for the necessity of logic, but that it is a fact about the ordinary world that forms this basis, a deep fact to be sure, but a fact nonetheless. It is experience which is the test of logic. More recently there have been philosophers of logic and mathematics such as G. Frege who have argued that the objectivity and necessity of logic requires in one's ontology, besides a world of ordinary things known by sense and the world of mental entities known by inner awareness, a third world of objective meanings. Mill clearly rejects, on empiricist grounds, any such third world: for him, the test for logic, as for anything else, is the ordinary world as we ordinarily experience it. In more recent terms, in logic and mathematics, as in epistemology and philosophy of science, Mill must be considered a naturalist. In this naturalism his views are in general similar to those of W. V. O. Quine.
There is no metaphysical necessity. All necessity is verbal, a matter of the import of propositions. A proposition is necessarily true in the case of connotative names just in case that the connotation of the names is by convention the same, as in ‘Bachelors are unmarried’. Since the connotations are the same, the set of individuals denoted by the subject term is identical with the set of individuals denoted by the predicate term, and the proposition is true, simply by virtue of the verbal conventions. In the case of proper names, where the terms do not connote, as in ‘Cicero is Cicero’, the individual denoted by the one term is precisely that denoted by the other, and so its truth is again verbal.
This account of the meaning of propositions is not complete. Mill does not consider the contrasting cases of ‘Cicero is Cicero’ and ‘Cicero is Tully’. Since ‘Cicero’ and ‘Tully’ are both proper names and therefore purely denotative, it would seem to follow on Mill's account that, contrary to fact, the two propositions have the same import. Later logicians would work hard to solve some of these difficulties in Mill's semantics. They would also have to work on problems arising from connotative proper names (“definite descriptions”).
Mill's empiricist thesis, that all necessity is verbal, has important consequences for his account of logic.
Consider the syllogism:
Man is mortal
Socrates is a man
Ergo, Socrates is mortal
It had been argued by Aristotelean and rationalist defenders of logic as a system of necessary science that it is a science that contributed to the growth of knowledge. The inference from ‘Socrates is a man’ to ‘Socrates is mortal’ is mediated, on this view, by the major premise ‘Man is mortal’ which establishes a real necessary connection between the minor premise and the conclusion by virtue of itself recording a necessary connection among the attributes connoted by the terms which it contains. This traditional account presupposes that there are necessary connections among attributes, that, in other words, attributes are not logically self-contained. Given Mill's claim that attributes are logically independent, this is wrong and the truth of the major adds nothing to the truth of the particular propositions, ‘this man is mortal,’ ‘that man is mortal’, etc., whose conjunction it records. Accepting the major as true is simply a way, on the one hand, of accepting that particulars one already knows share the attributes in question, and, on the other hand, a determination that one will continue to affirm this connection of hitherto unexamined particulars.
On Mill's semantics, then, the major premise ‘Men are mortal’ of the syllogism is, in its import, semantically equivalent to the extended conjunction:
Peter is a man and Peter is mortal, & Caius is a man and Caius is mortal, & Cicero is a man and Cicero is mortal, & etc.
It is, in other words a way of asserting an indefinitely long conjunction. Now, an inference
Peter is a man and Peter is mortal, & Caius is a man and Caius is mortal
Hence, Peter is a man and Peter is mortal
is not a genuine inference, one that is ampliative, yielding an increase of truth. It is only an apparent inference. Given the import as a conjunction of the major premise in the syllogism,
Man is mortal
Socrates is a man
Hence, Socrates is mortal
it follows that this too is only an apparent inference. Thus, logic—that is, deductive logic or syllogistic—adds nothing to our knowledge; its rules merely reflect our determination to reason consistently with the ways in which we have reasoned in the past: the rules of formal logic, of syllogistic, are the rules of a logic of consistency.
In contrast, logic as ampliative involves a passage from the knowledge of particulars summarized in the major, or universal premise, to an application of the same rule to a new case. In terms of knowledge, the major premise of a syllogism provides no knowledge beyond what it summarizes about particulars already known. Ampliative inference, that is, inference as a part of a logic of truth, is thus always a passage from particulars to particulars.
Since there are no objective necessary connections among the attributes of phenomena given in ordinary experience, it follows that the only grounds that we have for inferring from a sample to a population or from the past to the future are given by present experience or memory: all ampliative inference is inductive. It follows that such inference can never achieve apodictic certainty. This does not imply, as some have suggested, a scepticism about all events beyond present experience; it does not imply that all such judgments are somehow unreasonable or unjustified. It does imply, however, that our notion of rational justification ought to be adapted to this fact.
Humans find themselves as embodied creatures in the world, making “spontaneous” and “unscientific” inductions about specific unconnected and natural phenomena or aspects of experience. Examples are “fire burns” and “food nourishes”. The satisfaction of our desires, and indeed our very survival, depends upon our coming to ascertain, so far as we can, the truth about the natural world in which we find ourselves and about ourselves too. Ascertaining such truth as best we can is the cognitive means we have available for meeting those ends. No judgment aiming at such truth can, however, ever attain apodictic certainty, and so we ought, as reasonable beings at least, to be satisfied with less than that. In the absence of infallible knowledge, we ought as reasonable persons be satisfied with fallible knowledge. And within that framework we ought to find as best we can those rules of inference that on the basis of past experience form the best—though fallible—guide to matter of fact truth.
The spontaneous inductions, such as “fire burns”, all admit of qualification. Humans soon discover their limitations, and undertake inquiry to try to fill these gaps in their knowledge. The making of inductions we cannot avoid: that is part of our being human in the world, it is what we must do. The judgments at which we arrive are all fallible, and the Cartesian, proposing the cognitive goal of infallibility, would therefore have us reject them all. Since we must continue the practice, it is unreasonable to propose a cognitive goal that requires us to stop. It is therefore reasonable to continue the search after matter-of-fact truth, fallible though it is: that is what we ought to do: must implies ought.
(This principle that must implies ought is the converse of the well-known Kantian principle that ought implies can. It is justified by the argument that where there is something we must be then it is unreasonable to propose that something else be required. This principle permits one to infer an ought from an is, to move from the realm of fact to the realm of justification. Mill appeals to this principle to establish that it is reasonable to make inductive inferences, even though they are all fallible: this is not the only place where Mill appeals to this principle.)
But if the general practice of making inductive inferences is reasonable, some forms of that practice are more reasonable than others. Science is more reasonable than superstition; it is more reasonable than simply sticking with our original spontaneous inductions; and certainly it is more reasonable than the intuitions of the rationalists—including Mill's central opposition, William Whewell, who defended rationalism in science, intuitionism in ethics, the established Anglican religion, and unreformed universities.
Deductive logic keeps us consistent in our search after matter-of-fact truth; the logic of science—inductive logic—provides a set of rules that form the best, though fallible, guide that we can discover for the discovery of new truth. The rules of logic, both deductive and inductive, are rules of the art that has as its end, its cognitive end, the search after truth. All inferences are matters of psychological fact. For this Mill was later criticized by some later philosophers such as Husserl; he was accused of the sin of psychologism. But this is unfair to Mill. The latter is not claiming that the laws of logic are part of the subject-matter of the empirical science of psychology. He is arguing, rather, that the laws of logic, of both deductive logic and inductive logic, are normative, rules or standards about how we ought to reason, or, at least, about how we ought to reason given that we have a concern for matter-of-fact truth.
4. Mill's Empiricism: The Relativity of Knowledge
Mill argues that the apparatus of logic permits us to define predicates such as ‘unicorn’ that connote attributes that are not present in things given in ordinary experience. Such predicates have no denotation, and any proposition such as ‘this is a unicorn’ is false. More difficult are subject terms which connote but do not denote, e.g., ‘the present king of Mexico’. Mill was not alone in having difficulty with the logic of such terms; it was only with the work of Bertrand Russell on definite descriptions that these problems were solved in a way that could fit with an empiricist account of logic and of meaning.
But Mill's basic point is clear enough. Language aims to state matters of fact about the world. There are logical terms, such as ‘is’ or ‘is not’ or ‘and’, and non-logical terms. The latter are the subject terms and predicate terms of propositions. There are certain subject terms and predicate terms which are primitive to this language. All others are somehow defined on the basis of these primitive terms—leaving aside the details, to be worked out by later logicians, how these rules for introducing non-primitive terms are to be specified. Meanings are a matter of subject terms and predicate terms being hooked as it were to things and their attributes. (Mill is hazy, unfortunately, on relations, but in this he is hardly to be distinguished from any logicians prior to Peirce and Russell.) It is Mill's empiricism that these things and attributes to which the primitive terms of language are hooked are presented to us in ordinary experience, either sensory experience or inner awareness of our own states of consciousness.
Subject terms and predicate terms provide the content of propositions. Assuming that thought is propositional, they thereby define the limits of thought. Mill's empiricism thus determines the limits of what is thinkable. In this sense, all knowledge is relative to us, to our consciousness. We can of course have beliefs and even knowledge of things of which we are not conscious; there are parts of the world that we have never experienced. There are parts of things too small to observe by means of unaided sense; there are things too far to be seen by unaided sense; and there are things, such as the inside of unopened oranges, that we do not see but which we would see were we to do certain things, e.g., cut open the orange. The grounds of such knowledge or beliefs are not direct experience, but rather inference from direct experience. But our knowledge of those things, our beliefs about them, are still relative to us in the sense that we cannot think of them except as similar to or resembling things or attributes of which we are conscious in sense experience or inner awareness.
This much Mill takes for granted in his System of Logic. The empiricist framework he defends in detail in the Examination of Sir William Hamilton's Philosophy.
Some philosophers, for example, George Berkeley, have argued that the only grounds that can justify beliefs about ordinary things is direct consciousness. Mill rejects such an idealism: there is nothing in the being of attributes of things that ontologically determines that such things when they exist must be sensed. Mill argues in the Examination that material objects as the permanent possibilities of sensation exist independently of being sensed. But these are capable of being sensed, provided the knower is appropriately situated; though unsensed they are part of the world of (sensory) phenomena.
Mill also rejects the view of other philosophers, e.g., Kant or Sir William Hamiliton, that there are things or entities beyond the phenomenal world. According to these philosophers, phenomena are in fact the effects of such things but those things, as trans-phenomenal or noumenal, are wholly different from those phenomena; unlike material things as the permanent possibilities of sensation, they are devoid of sensory or phenomenal characteristics. These things which are wholly other and wholly foreign to anything we experience constitute the Unknowable cause or causes of the phenomena of which we are ordinarily aware. Among these philosophers, some such as Kant held that all that we can know is that the entities exist as causes of phenomena. Others such as Sir William Hamilton argued that we can know these not only as causes but also as being in themselves characterized by attributes, e..g., the primary qualities, which are also given phenomenally. Finally others, such as Hamilton's follower H. E. Mansel, argued that we can know the Unknowable as having attributes that at once resemble those that are had by ordinary things, e.g., some of the human virtues, but in ways which exceed our human powers to conceive—the suggestion here is that God, the Unknowable, is just but in a way that exceeds all humanly conceivable forms of justice.
Mill objects to these views in the name of logic: they are simply not consistent with the logic that is appropriate to the claim accepted by these philosophers that all knowledge is relative. In the first place, the concept of cause is, contrary to Kant and Hamilton, not an a priori concept. The concept of cause in its basic sense is acquired through our experience of matter-of-fact regularity: it is one that relates phenomena to other phenomena and not phenomena to noumena. The relativity of knowledge includes the relativity of our knowledge of causal relations, and these philosophers therefore have no grounds for supposing the existence of noumenal entities. As for Hamilton's claim that we have in our phenomenal experience knowledge of attributes of noumenal entities as these are in themselves, Mill has no trouble in showing the logical confusion in this claim. Hamilton wants to have it both ways—all our knowledge is relative or phenomenal and within this phenomenal realm we can discover concepts which apply absolutely or non-relatively to the noumenal things-in-themselves. Mansel's views receive Mill's particular scorn: if terms are not to be used in their ordinary sense then they ought not be used at all. A being, no matter how powerful, whose acts cannot be described in terms characteristic of human morality, is not one that we can reasonably worship. Mill made his well-known proclamation that “I will call no being good who is not what I mean when I apply that epithet to my fellow creatures, and if such a being can sentence me to hell for not so calling him, then to hell I will go.” (Examination of Sir William Hamilton's Philosophy, p. 103)
5. Scientific Method
Mill works out the basic principles of experimental science in the System of Logic, Book III.
He argues that the rules of scientific method evolve out of the spontaneous inductions about the world that we make as embodied creatures. As we investigate the world to find the best means to satisfy our natural needs and aims, some patterns maintain themselves, others turn out to be false leads. The former guide us in our anticipations of nature, and in our plans; they enable us to infer what will be and what would be if we were to do certain things or if other things were to happen. These patterns that we accept as guides we come to think of as laws: a law is a regularity that we accept for purposes of prediction and contrary-to-fact inference. Out of these ways of experiencing and coming to understand the world grows our account of explanation: to explain a fact is to locate a law under which it can be subsumed.
(This is what has come to be called the “covering Law” model of explanation or the “deductive-nomological” model. Positivists such as C. G. Hempel came to accept this model, as did K. Popper. But contrary to Popper's claim to originality, Mill is the first to clearly state this model.)
As we proceed in our efforts to understand and explain the world in which we find ourselves, the generalizations we accept begin to accumulate and interweave. We find, moreover, that there are not only generalizations but also at a more generic level patterns among these generalizations. We discover that as we search out such patterns we are often successful in discovering them. This itself is a pattern, a pattern about patterns, or, as Mill put it, a law about laws, a law to the effect that for all sorts of event there are laws, there to be discovered, which explain those sorts of event. This is a generic law to the effect that for every specific sort of event there is another specific sort to which it is regularly or lawfully connected. As we progress in science we generalize this law about laws to include all sorts of events: it is the Law of Universal Causation. It assures us that for any sort of event there are laws, there in the world, which, if we search diligently enough, we will be able to discover.
Guided by this principle, we also discover—fallibly, to be sure—that various rules of inference are more effective than others in generating acceptable causal beliefs. At the specific level, for any sort of event there are a wide variety of alternative possible determining attributes. We make assumptions—revisable—about the possible relevant causes, and identify the actual sort of cause by a process of elimination. Mill provides a detailed study of the various rules of eliminative inference in his well-known “Methods of Experimental Inference” (System of Logic, Bk. 2, ch. 9).
Mill here clearly describes the experimental method of science. One has a range of hypotheses. One searches for data and experiments that eliminate all hypotheses but one. One concludes that the uneliminated hypothesis is true. Clearly, the method of elimination yields a conclusion of truth only if two assumptions are fulfilled. One, it must be assumed that there is a cause, that at least one of the hypotheses is true. This has been called a Principle of Determinism. And two, one must assume that the true hypothesis is among the set one is working with, that the specific cause for which we are searching falls within this genus. This has been called a Principle of Limited Variety. Given that these assumptions hold, then the uneliminated hypothesis must be true.
But why accept the Principles of Determinism and Limited Variety for this area of research? Here one must assume a background theory, a laws about laws, which predicts for this area that these two laws hold. These laws are generic, and predict that there is a law, there to be discovered, that will be a specific law under the genus. The task of the researcher is to discover that specific law, guided by the theory.
The theory itself will be a generalization from laws in other, related areas. Confirmation of these specific laws tends to confirm the theory, and this confirmed theory then supports the Principles of Determinism and Limited Variety that guide the research. If the research is successful and one locates confirmed laws at the specific level, then that confirmation in turn supports the two Principles, which in its turn supports the generic theory from they have been derived. The confirmational support travels up from the specific laws to the generic and then down from the generic to the specific.
The Principles of Determinism and Limited Variety are laws. They predict that there is a cause of a certain generic sort, but do not predict specifically what it is. If that is all we know then it is clear that we would prefer to know the specific cause that they assert is there. In that sense, they are, to use J. L. Mackie's term,“gappy”, and it is the aim of research to fill in those gaps.
Our spontaneous inductions such as “water boils” are gappy in just this sense (“water boils” neglects the role of air pressure). Inquiry aims to fill those gaps. But such inquiry becomes the inductive method science only when those gappy laws are themselves supported, not simply as spontaneous inductions, but there is a background theory, a generic law about laws, that provides for them antecedent inductive support deriving from other, related areas.
Mill's picture of the inductive method of inquiry and the research that it guides is remarkably close to T. Kuhn's picture of “normal science.” What Mill calls a law about laws, Kuhn calls a “paradigm”, but that is a terminological difference. For both, they are theories that guide research: they assert that there is a law, there to be discovered in a certain generically described area, and it is the task of the researcher to discover it.
Mill also allows for something like what Kuhn calls “revolutionary science,” inquiry undertaken when the paradigm or background theory no longer leads to the discovery of specific laws. Failure to find the cause the theory asserts to be there will in general not require the rejection of the theory: the claim is an existence claim, and a failure to find something asserted to exist does not refute that claim, perhaps one has not looked hard enough. Mill, by the way, here agrees with Kuhn about paradigms or generic theories that they are not falsifiable by simple observations: specific hypotheses are so falsifiable but not generic laws about laws. In this they agree that Popper is wrong that falsifiability is the sole criterion of the scientific.
When a theory is not falsified but fails to be a successful guide in research then scientists begin to search for a new theory. But now this research is not guided by a theory: the research-guiding theory is no longer available. A looser form of research is now called for to discover a new theory. When a new theory is located by this research and it replaces the old theory then what Kuhn calls a scientific revolution occurs: the practice of normal science is restored, guided by the new theory.
The new theory will make predictions contrary to the old. When these predictions are successful theory new theory is confirmed. It is confirmed in areas where the old theory has failed. Since the new theory is confirmed perforce where the old theory is confirmed, and is further confirmed where the old theory is not, the new theory comes to be accepted as the one that is true. Since the new theory is contrary to the old, the old is therefore rejected as false. The elimination of an old theory is thus not a matter of a simple confrontation with falsifying observations as occurs as the level of specific laws, and as Popper would have it, but is more complicated, involving not only observational data but also a confrontation of theories. That at least is how Kuhn describes it. Mill would not disagree.
Kuhn is more subtle than Mill in his description of revolutionary science. But the principles Mill allows for the confirmation of generic theories are much of a piece with the looser methods of revolutionary science described by Kuhn. For Mill allows that as one moves up the generic hierarchy to more abstract levels of theory, then the eliminative methods no longer apply, and one must rely on something more akin to induction by simple enumeration. Not only must one do this, but it is reasonable so to do, Mill argues. Mill does not detail instances of theory change, as Kuhn does, but he makes his position clear in his account of the inference to the most generic theory of all, the Law of Universal Causation. That is, he makes clear how looser methods of inference come into play as one moves up the generic hierarchy of laws, how these looser methods are reasonable, and how the generic laws they justify asserting can be reckoned to be confirmed by observational data at the specific level. The terminology is not that of Kuhn, but the ideas are very similar.
Here is Mill's point. At the level of specific sorts of event the rule of enumerative induction—“from all observed As are Bs infer that all As are Bs”—is unreliable, often leading us to accept as true generalizations that later turn out to be false. Inferences in conformity to this rule often overlook relevant factors that are the real causes or mistake necessary conditions for sufficient conditions. Because of the variety of possible factors at the specific level, the eliminative methods lead to judgments that are more secure than the judgements which mere enumerative judgments would yield.
However, for any event, if one takes it generically, simply as an event, then there are at that level only two alternatives, caused or uncaused. Here there are not a variety of alternatives to be eliminated, and the rule of enumerative induction turns out to be more reliable than at the specific level: we have regularly been successful in finding causes at the specific level and this provides grounds for accepting the generic claim, the law about laws, that for all events there are causes to be discovered. Which is the Law of Universal Causation.
But then, as we extend our researches to a new specific area, the thesis that there are causes to be discovered provides inductive support for the lower-level claim that the result of the eliminative inference has really isolated a cause. There is thus an interplay as it were between eliminative and enumerative methods, with inferences at the specific level providing support for generic-level inferences, and the latter in turn providing support for specific-level inferences. Confirmational support rises up a hierarchy from specific laws to laws about laws, and then down from the laws about laws to the specific laws. Inferences in different specific areas mutually support one another through their joint support of the Law of Universal Causation which provides the grounding principle for inductive inference.
Mill's great opponent about the logic of scientific inference was William Whewell, whose History of the Inductive Sciences (1837) Mill had read with considerable care. Whewell argued that acceptance of scientific hypotheses depends first upon their capacity to explain observed phenomena and more specifically upon their capacity to explain phenomena in diverse areas (the “consilience of inductions”). Mill can accept the point about the consilience of inductions, given what he argues about the interplay of enumerative and eliminative reasoning. Further, Mill could accept the importance of hypotheses as working assumptions—“heuristic devices,” in Whewell's terms—, but he could not accept Whewell's claim that the mere fact that an hypothesis accounts for the data provides safe grounds for accepting it as a true statement of law. To be sure, eliminative methods may often show that a working hypothesis is in fact the only one consistent with the facts, and that it is therefore acceptable as true. But Whewell argued on the basis of the history of science that there are cases of hypotheses where the supposed causes have not been observed and which yet seem to yield explanations of observable phenomena. Such hypotheses involving unobserved causes can be found in inferences about areas too small or too distant to be observed—Whewell instances the undulatory theory of light. Mill agrees that there are such cases, and even allows that such hypotheses provide useful analogies for the guidance of future research. But so long as the data do not determine a unique hypothesis, such hypotheses cannot be accepted as yielding a new truth. Whewell's method of hypotheses, or the “hypothetico-deductive method” as it has come to be called by its defenders such as Popper, is often useful in inquiry, but does not justify accepting as true the hypotheses it assumes: for such acceptance to be reasonable, alternatives must be eliminated. It is the eliminative methods that provide the best (though still fallible) test for truth.
For Whewell, consilience is effected by generic hypotheses subsuming under themselves more specific hypotheses. These hypotheses involve generic concepts that, in Whewell's terms, “colligate” the more specific concepts that appear in hypotheses further down the ladder. Genuine progress in science depends not so much upon simple generalization from observed data as from the locating by inventive genius new colligating concepts. Mill does not disagree, but argues, contrary to Whewell, that colligation by itself is no test of truth.
It is Whewell's contention that as new colligating ideas emerge in the history of science, the principles in which they are embedded become necessary. The concepts in these axioms, such as ‘cause’ or ‘force’, are a priori, and research consists in gradually articulating these concepts into principles the necessity of which becomes more evident over time. What Mill would allow to be the free action of creative genius, Whewell construes as the uncovering by the mind of the divine ideas that provide the formal structure in conformity to which the Unknowable Creator constructs the world of phenomena. It is this necessity deriving from the Divine Creator that guarantees the truth of the basic axioms that organize scientific theories and which ensures the consilience of inductions. Needless to say, Mill rejects this account. The claim that some concepts have their origin a priori is inconsistent with the guiding thesis of the relativity of knowledge. Mill does not deny that in the process of scientific investigation, basic axioms become indubitable in the sense that their contraries become inconceivable. But such indubitability is psychological and does not derive from some sort of conformity to divine necessity. The truth of such axioms, if it really does obtain, is a matter of their conformity to the way the phenomenal world is, and mere fact that they are psychologically indubitable to the human mind does not guarantee that: given the relativity of knowledge, even indubitable judgements are fallible.
6. The Science of Psychology: Associationism
Central to both Mill's account of human reason and also to his social projects is his account, deriving from Bentham and his father, of the science of the human mind. This theory of how the human mind derives originally from Aristotle's discussion of associative memory. In Mill's hands it becomes a systematic hypothesis about which regularities govern human learning. Mill himself never wrote a systematic treatise on psychology, but late in his life he reprinted his father's Analysis of the Phaenomena of the Human Mind (1829, second edition, ed. J. S. Mill, 1869), with extensive notes revising and correcting his father's work.
The theory proposes that if f and g are regularly presented in experience as standing in relation R, then the habit forms in the mind, that if we have an impression or idea of f then it is accompanied with the idea of g. If R is the relation of spatio-temporal contiguity, then the ideas are joined to form a judgment of regularity, a causal judgment. If R is the relation of resemblance, then the ideas associated in the mind according to resemblance classes. A few experiences of connection will produce a loose connection in the mind. An increased frequency of experienced connection will produce a stronger association in the mind. And so attributes that are logically separate in experience through being repeatedly experienced in conjunction come to be inseparably connected in the mind.
Mill argues that this theory can account, however sketchily, for our use of language and how it becomes meaningful. If a word ‘w’ comes causally to be associated with a kind f, and f is associated with the resembling kind g, then the presence of f to the mind will call up both the associated kind g and the word w, so that w will come to be associated not only with f but also with g, and in fact with all the kinds that stand in the resemblance relation R to f. In this way words become general, applying to all members of a resemblance class. The mechanisms of association and the relations of resemblance thus come, for Mill, to play the role that abstract ideas played for earlier philosophers.
The habits of causal inference provide ways of anticipating what will occur in the world; that is how we learn what to expect. It is through processes of this sort that we form our spontaneous generalizations such as “fire burns” of “food nourishes.” Cognitively, these are inferences that conform to the rules of induction by simple enumeration. And in terms of our adapting to the world, these inferences are acquired purely passively. If this was all there is to Mill's account to the human mind then the criticism often level by the idealists that it ignores the active element in the human intellect, and presents a simplistic view of human reason.
But Mill's psychology also includes an account of motivation and action. On this theory, pleasure is the prime motivator, the primary end in itself, and the anticipation of pleasure serves as an immediate cause of bodily motions which in turn bring about that pleasure. Through regular success in attaining pleasure, anticipations of pleasure become associated with the sorts of action that bring about that pleasure. When Mill asserts that people seek pleasure, what he is to be taken to mean is that people seek things other than pleasure but that they seek it because pleasure has become associated with it, and that when the desire is fulfilled they experience the pleasure of satisfied desire. In this sense human welfare consists in satisfied desire.
There is one important feature of Mill's psychology in which he differed from his father. On his father's view, a complex idea produced by association is simply a collection is its associated parts. Thus, the idea of a house consists literally of the ideas of bricks, mortar, windows, etc., and the idea of everything consists of the ideas of every thing. This is surely a case of theory over-riding our clear experience. So it seemed to the younger Mill. On the latter's view, as he explained both in the Logic and his introductory notes to the second edition of his father's Analysis of the Phaenomena of the Human Mind, there is a sort of mental chemistry in which the parts fuse, as it were, into a new sort of mental whole. The causation is like that of chemistry, where, for example, water is the product of the fusion of oxygen and hydrogen, and unlike the mechanical causation of mechanics, where (as Mill saw it—not quite correctly [see Section 10, below])the product of several causes is merely the additive sum of the effects of those causes taken separately. These new sorts of mental unity emerge from associational processes and have properties which are not among the properties that appear in the genetic antecedents. Analysis of ideas is still possible, but it is not the simplistic sort of thing, a literal taking apart, that his father would have it be. As Mill explained in his Examination of Sir William Hamilton's Philosophy, the genetic antecedents are not integrant or real parts, as his father supposed, but only (as he put it) metaphysical parts; as metaphysical parts, they are present but only dispositionally. They can, however, through association (under the appropriate analytic set) be recovered, and brought to consciousness.
This new account of psychological association and analysis was important in Mill's thinking about ethics. Thus, where his father (and Bentham) had a simple notion that pleasures are all of a piece, and distinctions among them merely quantitative, one bit added to another bit, Mill came to see that there are qualitative distinctions among pleasures: the “higher” pleasures do result from association but they are different in kind from the “lower” pleasures out of which they arise, and as a matter of fact turn out to be more satisfying forms of pleasure. So Mill could say, where his father could not, that it is better to be Socrates dissatisfied than a pig satisfied.
It is evident that it was during his mental crisis that Mill came to be clear on the existence of, and importance for personal development, of these higher forms of mental unity in our conscious experience of the world. It was reading Wordsworth, it seems, that gave him this sense that there were forms of human being that were hardly part of his father's scheme of things. These feelings, to be obtained through poetry and human intercourse, were subsequently encouraged through his relationship with Harriet Taylor. These feelings, and their cultivation, came to form an important part of Mill's idea of the good that shaped his thought and his efforts towards social reform and progress.
Given the account of association and of action, it is evident that various means to pleasure will become associated with feelings of pleasure. But on Mill's view, this will not be a mere conjunction; to the contrary, as the association becomes strong enough the two parts will fuse into a new sort of emergent whole. The means will not simply be conjoined to pleasure but will become part of pleasure. And so money, for the miser, becomes not just a means to pleasure but for him part of pleasure, a end in itself.
This account of human action presupposes the acceptance of determinism, which Mill vigorously defends in the System of Logic, where he outlines the idea of a naturalistic science of human being. Freedom, Mill argues in Book Six, Ch. 2, which he thought the best in the work, is not the absence of causation but rather the absence of coercion. In fact the whole point of education is to determine the future free actions of the individual: it aims through the associative processes to determine the person's motives and actions.
This much Mill takes over from earlier thinkers such as Hume. But leaving it here has seemed by some again to render the person passive. Mill takes up this point in detail in Examination of Sir William Hamilton's Philosophy. The argument of the critics is that if character and motive are determined by earlier causes, how could a person be said him- or herself to be responsible for his or her actions? Hamilton so argued: the view makes the person a creature of his or her environment. This notion seemed convincing to Mill himself until he came to recognize that among the motives that one could acquire is the motive of self-improvement or self-realization. There are irresistible motives; for these we are not as persons responsible. But there are also resistible motives, and these we can shape and determine. That is, we can shape and determine them provided that we have the desire so to do. One is free if one could have resisted the motives on which one did in fact act, provided there had been good reasons so to do. A motive impairs freedom only if it is irresistible, only if it cannot be blocked by a strong reason against it. The free person is one who is sensitive to good reasons for behaving as he or she does. The second-order ends that lead one to shape one's motives and to develop as an individual became the central feature of Mill's social thinking, and this marks a major break in detail, though, to be sure, not in principle, with the utilitarianism of Bentham and his father. In the Examination of Hamilton's Philosophy, Mill vigorously defends the notion of human beings as active in their own self-determination.
This account of human being also provides an answer to those who argue that Mill's picture of human reason makes persons purely passive rather than active as thinkers. For, among the ends that can come to be associated with pleasure is the end of truth; in this way curiosity becomes an end in itself. And the motive for self-improvement will lead us to find, so far as we can, better ways to satisfy that end. We will so educate ourselves that our reasoning will conform, not to the simplistic rules of induction by simple enumeration, but to the more reliable rules of eliminative induction. The charge that Mill fails to take into account the active side of human reason is thus mistaken, resting on a failure to recognize those parts of the psychological theory that deal with motivation.
Mill has little to say about issue of mind-body relations. He clearly holds that mind, that is, mental events affect bodily events and conversely. These connections are not at all mysterious. Many, such as Descartes, create a chasm between mind and body by taking them to be separate substances of kinds that separable from one another and which therefore lack the necessary connection which is the core of causation on the substance philosophy. Since Mill, on the usual empiricist grounds, rejects substances and objective necessary connections, these problems do not arise. The core idea in causation is, alternatively, regularity, there are regularities with respect to mind-body connections (whenever I will my arm to go up, then it goes up, mostly anyway; when I am in the presence of, say, an orange, I see that orange, mostly anyway), since there are these regularities, there are causal relations: there is no mystery to it.
Mill is certainly not an epipehnomenalist: mind does causally affect body, as well as body affecting mind. Even more certainly, he is not an identity theorist: mental events are clearly distinguished in experience from bodily events: they are therefore not identical. He does hold that the laws of psychology are within themselves gappy, and that the gaps are to be filled by reference to non-mental events. There are, on the one hand, events in the world that evoke as responses mental events, and, on the other had, there are the actions and the upshot of actions which are the effects of mental events. Also relevant are the physiological events that intervene between the sensory stimuli and the bodily response. While Mill recognizes their existence, he has little to say about them, unlike Herbert Spencer who, importantly, deals with them in detail in his Principles of Psychology.
Many now hold that the system of physical or bodily events forms a causally closed system. This has been the standard view since the discovery of the Law of the Conservation of Energy in the middle of the nineteenth century. Like the other great, perhaps greater, discovery in the nineteenth century, Darwin's theory of the origin and evolution of species by natural selection, Mill simply fails to note their significance. In any case, he has a clear solution: parallelism. This was the solution proposed and defended by Mill's contemporary, the physiologist W. B. Carpenter, who found it more reasonable than the epiphenomenalism of T. H. Huxley. Mill could have found, and perhaps did find, this parallelistic solution congenial. Certainly, he was familiar with David Hartley's Observations on Man (1749) which clearly, on the one hand, proposed a parallelism between mental states and bodily states, and, on the other hand, articulated with equal clarity an associationist account of learning.
Source: Wilson, Fred, "John Stuart Mill", The Stanford Encyclopedia of Philosophy (Spring 2009 Edition), Edward N. Zalta (ed.), URL = http://plato.stanford.edu/archives/spr2009/entries/mill/