Rationalism and Empiricism
Some Notes on Epistemological Strategies and their Implications in Ethics

While the main focus in an ethics course is on ethics and the problems and issues that ethics raises, it is impossible to investigate these problems in isolation, without at least some excursions into the other philosophical sub-disciplines. While all the philosophical sub-disciplines consider what, on one level, are separate questions and issues, there are considerable interconnections, as assumptions in one area will have repercussions in other areas.
    One question that all ethical theories must address is where ethical knowledge arises, i.e., where does the knowledge about general ethical principles or the knowledge that certain actions are moral or immoral originate? These and other similar questions raise issues that are no longer unique to ethics, rather these issues touch upon more general epistemological questions, i.e., questions about knowledge—its sources, nature and justification.
    To some the question ‘Where does knowledge originate?’ might seem rather strange. While knowledge acquisition and manipulation are essential to human beings, the more usual epistemological questions concern some particular idea’s source or some statement’s truth conditions. So, while it is common to inquire into a statement or idea’s source, to inquire into all knowledge’s source seems strange. Some might question whether an answer is even possible. Nevertheless, it is a legitimate (indeed, an essential) philosophical question and though there are difficulties—real difficulties—answers are possible.
    Rationalism and empiricism represent the traditional Western philosophical responses to these epistemological questions. As epistemological theories these philosophical traditions each trace their origins to ancient Greece and the earliest philosophical speculations about the human condition and each also brings unique insights and assumptions to questions about human knowledge’s nature and origins.


Rationalism distinguishes between empirical knowledge, i.e., knowledge that arises through experience, and a priori knowledge, i.e., knowledge that is prior to experience and that arises through reason.
    As knowledge that arises through our experiences, empirical knowledge is about the material universe (and the various entities and phenomena in that universe). Sentences such as ‘Edinburgh is in Scotland’, ‘It is 75o outside’, ‘John Locke was a philosopher’, ‘The average moose weights 1500 pounds’ each express statements about certain entities in the universe and so represent empirical knowledge.
    In contrast a priori knowledge is not about phenomena in the empirical universe or our experiences, though some a priori knowledge is applicable to that universe. The sense in which a priori knowledge is ‘prior to’ experience is logical rather than temporal, i.e., it is possible that one learns some a priori knowledge through experience, nevertheless that knowledge neither requires experience in order to be known, nor is about experience. Perhaps it is easier, then, to consider a priori knowledge as knowledge that arises through reason alone, i.e., it depends upon no experience. Consider, e.g., mathematical knowledge or logical knowledge. The statement ‘All triangles have three sides’ makes no claim about experience or the empirical universe since there are no ‘triangles’ in the universe. There are, to be sure, triangular entities, i.e., physical entities that have a triangular shape, but no triangles themselves. In a similar manner, the statement ‘3+3=6’ makes no claims about the universe as there are no 3s or 6s that one can experience and so possess empirical knowledge about. Again, while it is obvious that some mathematical knowledge is applicable to experience (e.g., ‘3+3=6’ is applicable when one has 3 apples and someone gives one 3 more apples—one then knows that one has 6 apples), this fails to demonstrate that the mathematical statement ‘3+3=6’ is an empirical statement. The logical statements ‘x = x’, ‘All the entities in the universe are either x or not-x’ and ‘No entities in the universe are both x and not-x’ are also statements that while applicable to experience are not about experience.[1]
    There is another difference between empirical and a priori knowledge in addition to their respective sources and content. This difference has to do with their truth conditions. A truth condition specifies under what conditions a given statement can be said to be true or false, i.e., it indicates what one needs to do to prove a statement true or false. Consider the statement ‘It is 75o outside’. Under what conditions is this statement true? It should be obvious that the statement is true so long as the outside temperature is 75o. How would one prove whether the statement is true or false? Again, it should be obvious that one would need to determine, through some procedure or apparatus, the outside temperature. In short, one appeals to experience and the empirical data it provides.
    In contrast to this empirical statement, consider again the statement ‘3+3=6’. Under what conditions is this statement true and how is it possible to prove it? Well, it is true so long as 3+3 does indeed equal 6, this much seems obvious. But, and here is the principal difference between empirical and a priori knowledge, how does one prove the statement to be true? Perhaps the most obvious response is: Well, take three apples and add them to three more apples and then there are six apples. While this demonstration is to the point, does it suffice to prove that 3+3=6? No, at best this little exercise confirms the statement, but it fails to prove it.
    To understand the difference between ‘prove’ and ‘confirm’ consider another illustration. It is a quiet summer afternoon and James decides to rest on the grass beside a river. Some moments later a white swan swims down stream. As James continues to rest seven more swans, that are also white, swim down stream. James considers this experience and realizes that all the swans he has ever seen have been white. So, James formulates the statement ‘All swans are white’ and sure enough the next swan he passes is white. Did this last experience prove that the statement ‘All swans are white’ is true? No, since James has not seen ‘all swans’, it is possible that there is at least one that is some non-white color. James’ experience does, however, provide additional confirmation that the statement is true (at least until James discovers there are non-white swans).
    To prove that 3+3=6 is true then requires that one appeals to more than experience. To be precise, one must appeal to other mathematical knowledge. At this point someone will perhaps take exception with this analysis and point out that since one learns mathematics through experience, so mathematics must also be empirical knowledge! The point is well taken. The source, however, is not the real issue. The real issue is what the knowledge is about and its truth conditions. Moreover, even though some a priori knowledge might arise through experience, it should be obvious that most does not, i.e., while one might argue that one learns basic mathematical truths, e.g., ‘1+1=2’, ‘2+2=4’ and so on, through experience, it seems clear that there are other mathematical truths that it is much more difficult to learn through experience, e.g., ‘3525+2353=5858’ or ‘a2+b2=c2’. The rationalist’s point here is that a priori knowledge is about more than experience and as such it provides knowledge that experience is unable to provide.
    A similar analysis will demonstrate that logical statements such as ‘All the entities in the universe are either x or not-x’ also depend upon no experience to determine their truth. Indeed, since the statement is about all the entities in the universe, the experience one needs to prove it as an empirical claim is impossible. It should be obvious, however, that one needs no experience or empirical data to prove the statement, i.e., whatever characteristic one chooses as x, it is apparent that all the entities in the universe either have x or do not have x. All the entities in the universe are either purple or not purple, bigger than a cat or not bigger than a cat, spherical or not spherical, and so on. One can know that this statement is true even when one has no idea what the characteristic in question is. Thus, one knows that all the entities in the universe are either merbalis or not-merbalis, even though no one else in the universe knows what merbalis is (since I made it up!).
    To rationalists this power to discern and generate universal truths is quite impressive. Indeed, the differences between rationalism and empiricism as to (a) what constitutes genuine knowledge, (b) what such knowledge is about, and (c) its truth conditions, suggest to the rationalists that there is a real qualitative difference between empirical and a priori knowledge. To be precise, most rationalists argue that a priori knowledge is superior to empirical knowledge. The one consideration that is seen as the most decisive in this argument is the difference in truth conditions between empirical and a priori knowledge.
    Most rationalists consider there to be a fundamental problem with empirical knowledge. Empirical knowledge depends upon our senses, senses that, the rationalist wastes no time to demonstrate, are unreliable. Here the rationalist appeals to common sense deceptions and perceptual illusions—when one places a straight rod into water the rod appears to bend, at a distance a square tower appears to be round, parallel lines appear to converge in the distance, and so on.[2] Thus, it is difficult, perhaps even impossible, to ever know that an empirical statement is true. It seems that it is possible to doubt even the most certain sense perceptions. In contrast, a priori knowledge is certain knowledge. While it might be possible to doubt that I see a map on the wall beside the computer (I might have a bizarre optical disease or it might be a hallucination), it seems impossible to doubt that 2+2=4. Furthermore, while empirical knowledge represents conditional knowledge, i.e., knowledge that might have been otherwise, a priori knowledge is universal and eternal. Again, while it is possible to imagine a universe in which the earth’s circumference was 30,000 miles rather than 25,000 miles or a universe in which politicians are honest or a universe in which the Chicago Cubs do win a World Series, it seems impossible to imagine a universe in which 2+2=6 or where triangles have more (or less) than three sides.
    As with most philosophical theories there is some disagreement between rationalists on certain issues. One issue that separates rationalists is the answer to the question where a priori knowledge originates. The more radical rationalists (e.g., Plato and Rene Descartes) argue that a priori knowledge is innate, i.e., the knowledge is in some manner latent within the mind or even built into the mind. At best then experience acts to elicit the knowledge, but the knowledge was there prior to the experience. Plato argues that all genuine knowledge is innate and education is mere recollection or remembrance (see Plato’s dialogue Meno), while Descartes claims that certain critical concepts—God, material substance, and mental substance—are innate. Given these three innate ideas and reason, Descartes argues that other a priori knowledge is derivable.
    The obvious problem that these radical rationalist strategies face is the need to explain where the mind acquires these innate ideas. In Plato’s case the solution is an immortal soul-mind that lives through countless lives (i.e., reincarnations), whereas Descartes argues that God places these ideas in human minds. It is also possible to argue that evolution is responsible, i.e., the mind’s biological structure contains the ideas. While this sounds rather strange, the linguist Noam Chomsky argues this precise thesis. Unless one assumes that certain linguistic structures, e.g., deep grammar, are innate, the argument goes, it is impossible to explain the apparent ease with which human beings learn natural languages.
    Immanual Kant argues a less radical rationalist line. Kant accepts the rationalist claim that reason alone can provide certain knowledge. Nevertheless, Kant also accepts the empiricist claim that all knowledge begins in experience, i.e., without sense experience as the initial data upon which reason can operate, the knowledge acquisition process can never start. Knowledge, as Kant conceives it then is what the mind produces as it orders and structures otherwise chaotic sense data. The rather radical idea here is that it is the mind that imposes the order and structure on the sense data, the implication being that the sense data have no intrinsic order or structure. The main organizational principles that the mind imposes on sense data are its spatial and temporal structure. These considerations led Kant to a metaphysical distinction—the distinction between the noumenal universe and the phenomenal universe. The noumenal universe comprises entities-in-themselves, while the phenomenal universe comprises entities-through-their-appearances (White 1996: 296). This is rather technical so it is best to go through it in stages.
    Suppose someone presents us with a blue glass sphere. It is through our senses that we perceive this sphere. In this case the principal senses are visual and tactile—our visual sense indicates that it is blue and spherical and our tactile sense that it is glass and also that it is spherical. Philosophers call these qualities—being blue, being glass and being spherical—properties or characteristics. All entities have properties—a size, a shape, a color, a taste, a texture, an odor, and sound and so on. Kant’s point is that it is through these properties, and through these properties alone, that all the knowledge we have about the entities in the universe arises. All knowledge about entities comes through their properties (which Kant calls appearances).
    Our commonsense intuitions suggest, however, that there must be some substance or matter that has the properties that our senses perceive, i.e., that the properties cannot exist without some substance that underlies them and possesses them as properties. While the substance that underlies the properties is unseen, nevertheless reason and commonsense insist that it must exist. Descartes suggests that such inferences are rather common occurrences, e.g., when one peers out a window on a cold winter afternoon one might see a person move across the lawn. But does one see a person? No, all that one sees is a cap, a coat and perhaps trousers and shoes. Nevertheless, no one doubts that there is someone under all the apparel. Even though one is unable to see the person one still reasons that there must be one there, since clothes seldom stroll across lawns on their own.
    Kant agrees that there must be ‘entities’ that possess the properties our senses perceive, but argues that while logic necessitates their existence, these entities-in-themselves (which comprise the noumenal universe) are unperceivable and so incomprehensible to the human mind. All that is knowable are the properties (i.e., appearances) that our senses perceive and our mind structures. These appearances are the entities that comprise the phenomenal universe.
    There are no means then to, as it were, move outside our senses to see entities in themselves, to see the ‘real’ universe rather than the universe that our senses communicate to us through perception. Since all our knowledge comes through the senses and reason, these act as filters which order and structure all our perceptions and thoughts. The entities-in-themselves that underlie the perceptions remain forever elusive.
    While perhaps more plausible, Kant’s rationalism imposes limitations on knowledge that more radical rationalists would refuse to accept. Nevertheless, Kant’s approach is rationalist since it is the mind (to be precise, reason), that gives our sense perceptions the structure that changes them into knowledge (White 1996: 297).
    The main point to remember is that rationalists believe that, even though it might require experience to initiate the knowledge process, there is some knowledge that is irreducible to experience, i.e., the knowledge is neither about experience nor is it possible to use experience to demonstrate that the knowledge is true or false.


Empiricism denies the rationalist distinction between empirical and a priori knowledge. All knowledge, the empiricist argues, arises through, and is reducible to, sense perception. Thus, there is no knowledge that arises through reason alone.
    It is essential to be clear here: it is not reason’s existence that empiricism denies, or that reason has a role in knowledge acquisition and manipulation, rather it is that reason has some special access to knowledge over and above the knowledge that experience provides. All empiricists acknowledge that human beings possess reason—reason is the instrument that allows us to manipulate and augment the knowledge that experience provides. Knowledge, however, has its origins in experience rather than in reason.
    Empiricism begins with the distinction between sense data and ideas. Sense data represent the basic information that the senses present to the mind through our perceptual experiences, i.e., sights, tastes, textures, sounds and odors. To illustrate, suppose that one sees a blue sphere. This sense experience is reducible to the visual act and the sense data (i.e., the information that the visual act contains). In this case the information that the visual act contains is that there is a visible ‘blueness’ and a ‘sphericalness’. At this stage there is no conscious recognition that one sees a blue sphere, all there is is the pure sense data that the senses present to the mind through the sense experiences. The mind processes and represents each individual sense datum as an idea, in this case the ideas ‘blue’ and ‘spherical’. The mind then associates and combines the ideas it creates through sense experience to create the conscious idea ‘blue sphere’.
    To the empiricist, sense data represent the basic material that the mind uses to construct the ideas that comprise all our knowledge. Thus, no matter what the idea is, it is possible to trace that idea to some sense experience(s). While the precise details differ, these are the basic cognitive mechanisms that the principal empiricist philosophers—John Locke, George Berkeley and David Hume—all appeal to in order to explain the process through which sense data becomes knowledge.
    Although empiricism denies a priori knowledge’s existence, as knowledge that depends upon no experience, there is still the recognition that some knowledge goes further than experience in the sense that it is not about experience. Nevertheless, empiricism argues that such knowledge is still reducible to experience. Again, this is the crucial notion—that it is possible to trace all knowledge, whether or not it is about experience, to some particular experience or experiences.
    Rather than preserve what is thought to be an inaccurate distinction, empiricism recasts the distinction between a priori knowledge and empirical knowledge into the distinction between analytic knowledge and synthetic knowledge. Through this distinction empiricism denies the rationalist claim that a priori knowledge is superior to empirical knowledge. Indeed, the distinction provides the basis to argue the precise opposite.
    The statements that the rationalists cite as paradigmatic a priori knowledge—‘A triangle has three sides’, ‘3+3=6’ and so on—the empiricist sees as analytic statements. An analytic statement is one where the statement analyzes the concept in question. Thus, the statement ‘A triangle has three sides’ does no more than analyze the concept ‘triangle’, and the statement ‘3=3=6’ does no more than analyzes the concept ‘six’. Moreover, the empiricist argues, these statements never do more than analyze the concepts in question. In a real sense then these statements provide no additional knowledge, all the knowledge that analytic statements contain is given is within the original concept the statement analyzes (White 1996: 280).
    Synthetic statements, in contrast, do provide additional knowledge—knowledge that goes further than the original concept. Consider the statement: ‘The temperature outside is 75o’. This is a synthetic statement since, while it has to be some temperature outside, there is no reason that it has to be 75o rather than some other temperature. The concepts ‘temperature’ and ‘outside’ then have no intrinsic connection to some specific outside temperature, rather what the temperature depends upon are various other environmental conditions. So statement such as ‘The temperature outside is 75o’ provide us with additional (and sometimes valuable) information. All synthetic statements then share the characteristic that, because there is no intrinsic or logical connection between the statement’s elements, these statements provide information about a connection or relation that is unavailable in the original concepts themselves.
    Given that analytic statements reveal no additional insights, while synthetic statements do provide novel ideas and associations, it should come as no surprise that empiricism argues that empirical knowledge is superior to a priori knowledge rather than the reverse (or to be more precise, that synthetic knowledge is superior to analytic knowledge). With the focus on analytic truths rationalism never quite reaches the ‘real’ universe in the manner that synthetic statements are able to do.
    There is, however, a philosophical price to be paid. While the empiricist gains additional insights and knowledge there is a loss in certitude, since the empiricist still must deal with senses that (the rationalist is correct to maintain) are unreliable. The rationalist can be certain that 2+2=4, the empiricist, however, must accept that empirical knowledge is at best probable, never certain. The problem is that the empiricist has no real response to the claim that it is possible to doubt even the most persuasive sense impressions, since it is possible to doubt them without logical contradiction. In philosophical terms, the problem is that our sense perceptions underdetermine their causes, i.e., a given sense perception has more than one explanation. Consider, e.g., that one sees a white rabbit. What might explain this perception? The obvious answer is that one sees a white rabbit because there is a white rabbit there. It is also possible, however, that one has a rare optical disease and the rabbit is some other color, rather than white. It is also possible that one hallucinates or dreams the rabbit. As Alice will attest, these are all logical possibilities and the sense experiences in themselves provide no certain means to decide which explanation is correct.
    This suggests another potential problem that empiricism must address—how to explain mathematics and logic? Remember that empiricism maintains that all knowledge is reducible to experience. Thus, the empiricist must explain how it is possible to reduce sometimes arcane mathematical knowledge to common sense experience. This means that, since mathematical knowledge is thought to be certain knowledge, the empiricist must explain how it is possible to derive certain knowledge through a process—sense experience—that provides knowledge that is, at best, probable. Moreover, the empiricist must also explain how it is possible to prove mathematical statements through experience.
    There have been numerous attempts to demonstrate how it is possible to derive mathematics and logic through experience. Though commendable these attempts all have had serious difficulties and so have met with little general acceptance.
    Even were it possible to reduce mathematics to experience, the questions (1) whether experiences whose truth is probable can produce certain mathematical knowledge and (2) how it is possible to prove mathematical statements through experience, pose rather more serious difficulties.
    Perhaps the easiest, though least intuitive, solution is to argue that there is no certitude in mathematics. This is John Stuart Mill’s tactic. Mill, a radical empiricist, argues that, as with all other all empirical statements, mathematical statements express mere probabilities. All that distinguishes them is that mathematical statements have undergone more extensive con-firmation than other statements (Encyclopedia of Philosophy 2: 503). The disadvantage to this tactic is obvious: one must give up all claims to absolute truth in mathematics. Most philosophers (as well as mathematicians) consider this concession to be as difficult as it is undesirable.
    In contrast to Mill, less radical empiricists, e.g., David Hume and John Locke, still want to maintain mathematics’ certitude. This too, however, comes at a price. To preserve mathematical truths as absolute truths Locke argues that some perceptions, and the ideas that represent these perceptions, can be more certain than others. To be precise, Locke argues that, when reason operates on experience, the ideas, and the associations between ideas, that it produces result in knowledge that is either intuitive, demonstrative or sensitive. Locke maintains that intuitive knowledge and demonstrative knowledge are certain knowledge (Encyclopedia of Philosophy 2: 501). Locke’s arguments here are technical and, to most, less than a complete success. To all intents and purposes, however, what Locke does in order to guarantee certain knowledge is to introduce certain rationalist elements. The consequence is that Locke’s certain knowledge is rather too similar to the rationalist’s a priori knowledge to please most empiricists.
    Since empiricism argues that there is no knowledge that arises through reason alone, it should be obvious that empiricism also denies that there are innate ideas, i.e., ideas that are in the mind prior to experience or that are built into the mind in some manner. The standard argument against innate ideas is that were there such ideas then all rational beings should possess and acknowledge them. Since it is obvious that there are neither universal ideas, i.e., ideas that all human beings possess, nor ideas upon which their is universal agreement, then there are no innate ideas (see John Locke’s Essays on the Law of Nature and Essay Concerning Human Understanding, and David Hume’s A Treatise on Human Nature).
    The empiricist considers the pre-experience mind to be a tabula rasa—a clean slate—and it is through experience that knowledge comes to be ‘written’ on this slate. Thus, empiricism’s credo is that where there is (or can be) no experience there is (and can be) no knowledge.


The debate between rationalism and empiricism continues, and it is quite possible some issues will be impossible to resolve, at least given our finite human intelligence. To the degree that it is possible to determine the correct solutions to these issues, the British philosopher Bertrand Russell concludes that the score is even. Russell argues that while it seems clear that the empiricists are correct that all knowledge must arise through experience, it also seems obvious that there is some knowledge that it is impossible to reduce to experience, i.e., reason is able to use experience to produce knowledge that it is nevertheless impossible to prove through experience (see The Problems of Philosophy).
    The main purpose here, however, is to illustrate that one’s general philosophical assumptions about knowledge’s nature and origins will have consequences in other philosophical investigations, in particular in ethics. And to illustrate that all theories involve compromises, i.e., no matter the initial assumptions, there will be advantages and disadvantages. It is to a philosopher’s credit then to be able to detect and acknowledge the disadvantages as well as the advantages that their positions entail.

John Locke:
Locke’s natural law ethics reveals the same tensions that run through Locke’s general approach to knowledge. The desire to have some knowledge be certain knowledge, even though all knowledge arises through experience, forces Locke to argue that reason is able to combine some ideas in a manner that produces certain knowledge. Such knowledge is irresistible, i.e., it leaves no room to hesitate or doubt (Encyclopedia of Philosophy 4: 497).
    Thus, Locke argues that certain knowledge is possible. Perhaps most important to Locke’s ethics is the conviction it is possible to be certain that God exists. More than this, since Locke bases what is moral on what God wills, it is even possible to know what it is that Gods desires human beings to do, i.e., the divine law. The divine law as discoverable through reason becomes the natural law—the command to preserve human beings.
    The natural law, Locke argues, underlies and governs all human interaction. Thus, through the nature law reason is able to derive all the particular natural rights and moral duties that human beings possess. These are rights and duties that all human beings possess as human beings and that human beings must use as a guide in their behavior. The universal and absolute character is what reason supplies to experience to produce certain knowledge.

Immanual Kant:
While Kant thought there was much to admire in the empiricist philosopher David Hume’s A Treatise on Human Nature, and though he even accepts the empiricist principle that all knowledge arises in experience, Kant is without doubt a rationalist. This rationalism is quite apparent in Kant’s philosophical investigations into ethics.
    Kant believes that the supreme principle that underlies all morals—the categorical imperative—must be absolute and universal. Such a principle can never arise in experience, Kant argues, since all experience is particular (i.e., about particular entities in particular situations at particular times). Neither can experience prove this principle. Experience can at best, Kant insists, confirm the categorical imperative.
    In contrast to the knowledge that arises through experience, the knowledge that arises through reason is abstract and universal. To illustrate the difference consider the statements ‘There are wombats in Tasmania’ and ‘a2+b2=c2’. It is clear that the empirical statement ‘There are wombats in Tasmania’ is about particular entities (wombats) and a particular situation (being in Tasmania). The mathematical statement has no such limitations. This statement is abstract in that it mentions no particular entities and universal in that it applies to all appropriate as, bs and cs.
    It is reason alone then that is able to determine and prove the categorical imperative as the supreme moral principle. Kant distinguishes here between theoretical reason and practical reason. It is theoretical reason that investigates the empirical universe. This is the reason that science uses. Practical reason’s concern is the will, that motive force in human beings that underlies all moral behavior. To be precise, it is practical reason’s role to create a good will. To do this practical reason determines the moral principle that the will must follow, i.e., the categorical imperative.
    The general epistemological limitations that arise because Kant accepts the empiricist principle that all knowledge begins in experience are also apparent in Kant’s ethics. Since it is impossible to know entities-in-themselves there are certain entities and ideas, whose importance to ethics are immeasurable, about which human beings can have no knowledge whatsoever. In particular, it is impossible to have knowledge as to whether (1) God exists, (2) the soul is immortal and (3) that human possess free will. Kant argues, however, that even without certain knowledge, it is still essential to assume that all these are true, otherwise ethics is impossible.

John Stuart Mill:
Mill’s utilitarian ethics incorporates the radical interpretation that Mill gives the empiricist principle that all knowledge arises in experience. Mill interprets the ‘all’ to mean all knowledge. Thus, Mill assumes that even mathematical and logical knowledge are empirical knowledge with all the limitations that such knowledge possesses. Mill manages to overcome, however, the scepticism that characterizes Hume’s empiricism (Encyclopedia of Philosophy 5: 318).
    The Greatest Happiness Principle that underlies utilitarian ethics states that those actions are moral which provide the greatest happiness to the greatest number. What determines happiness is without a doubt an empirical matter, i.e., it is through our experience that we realize what actions cause the pleasures that increase happiness and what actions cause the pains that decrease happiness.
    Reason’s role in this process is to learn through these experiences and to formulate the general moral rules that will, over time, lead to the greatest happiness. It is essential to realize, however, that while these general moral rules are meant to guide behavior, because our experiences change, these rules can and do change over time. There are no certain, or absolute, or universal moral rules. Experience is unable to provide such permanence.
    Mill also acknowledges, that it is impossible to prove that happiness is the ultimate end that drives all human desire and action. As a consequence Mill must concede, and this is a rather radical concession, that it is impossible to provide a logical demonstration that the Greatest Happiness Principle is the fundamental moral law. Logical analysis, Mill argues, has no place in ethics. In contrast to Locke and Kant then Mill denies that ethics is, or can be, a science. In the end, Mill’s normative ethics rests upon psychological observations and arguments, whereas Locke and Kant believe their normative theories to rest upon logical arguments.


   1. Bertrand Russell argues that, more that obvious logical truths, without at least the assumption that these principles are true, rational argument becomes impossible (1912: 72).
   2. There is an extensive discussion about these problems in Rene Descartes’ Meditations on First Philosophy.

Sources and References

Blau, J.L.
   1967 Immanual Kant. The Encyclopedia of Philosophy.
   1967 John Locke. The Encyclopedia of Philosophy.
Descartes, Rene
   1993 Meditations on First Philosophy. Indianpolis: Hackett.
Hamlyn, D.W.
   1967 Empiricism. The Encyclopedia of Philosophy.
Hume, David
   1969 A Treatise of Human Nature. London: Penguin.
Locke, John
   1950 Essays on the Law of Nature. Oxford: Oxford University Press.
   1975 An Essay Concerning Human Understanding. Oxford: Oxford University Press.
   1981 Five Dialogues. Indianapolis: Hackett.
Russell, Bertrand
   1912 The Problems of Philosophy. Indianapolis: Hackett.
Schneewind, J. B.
   1967 John Stuart Mill. The Encyclopedia of Philosophy.
White, Thomas I.
   1996 Discovering Philosophy. Upper Saddle River, NJ: Prentice-Hall.
Williams, Bernard
   1967 Rationalism. The Encyclopedia of Philosophy.