For about the first 200 years after the publication of the Treatise, critics had viewed Hume as the consummate skeptic, and they had labored to refute his skeptical arguments. Thomas Reid and Immanuel Kant, for example, each tried in his own way to overcome Hume's skeptical doubts. After the publication in 1941 of Norman Kemp Smith's The Philosophy of David Hume, however, scholars have come to appreciate Hume's positive philosophy. The stated goal of the Treatise is no less than to produce a "science of man," which describes the inner workings of the human understanding, the passions, and the sentiments of morality.
Hume was deeply influenced by the success of Newtonian natural science in discovering the principles according to which the physical universe operates. His goal was to use what he saw as Newton's method as a tool for the discovery of the principles that systematically describe the workings of the mental world. Specifically, these principles would have to be based on "observation and experiment" (Treatise, Introduction). Throughout the Treatise, Hume supported his claims by appealing to the observations of human behavior that have been accumulated over the ages, as well as observations of himself. The kind of "experiment" to which he refers is purely mental. Hume produces, or tries to produce, a given mental state, in order to see what happens as a consequence. Sometimes he invites the reader to try the same experiment.
Newton had deplored the use of "hypotheses" in science. He regarded hypotheses as mere possibilities dreamed up by the imagination and not supported by observation and experiment. The mere possibility of a contrary case should have no force against principles to which an exception has never been observed. Thus in a section of Principa Mathematica entitled "Rules of Reasoning in Philosophy," he stated:
In experimental philosophy we are to look upon propositions collected by general induction from phenomena as accurately or very nearly true, notwithstanding any contrary hypotheses that may be imagined, till such time as other phenomena occur, by which they may either be made more accurate, or liable to exceptions.Hume was averse to any use in the study of human nature of hypotheses that are "extravagant" (Introduction) or "vulgar" (Book I, Part III, Chapter 12). Again describing his own work, he wrote, "He proposes to anatomize human nature in a regular manner, and promises to draw no conclusions but where he is authorized by experience. He talks with contempt of hypotheses" ("Abstract"). Nonetheless, Hume advanced explanations that he himself described as hypotheses. He laid down strict rules for their use.
A scrupulous hesitation to receive any new hypothesis is so laudable a disposition in philosophers, and so necessary to the examination of truth, that it deserves to be complied with and requires that every argument be produced, which may tend to their satisfaction, and every objection removed, which may stop them in their reasoning. (Book I, Part III, Section 9)Hume credited Locke and others as "founding their accurate disquisitions of human nature entirely upon experience" ("Abstract"). The empiricist rejection of ungrounded hypotheses in favor of explanations based entirely on observation has the consequence that we are ignorant of many things, but this was a price that Hume was willing to pay to avoid error.
Hume took the science of man to be the foundation for all the other sciences. Most directly, it is the basis for logic, morals, aesthetic criticism, and politics. (Here Hume was operating with the traditional conception of logic as a science of human reasoning, not of necessary logical truths, as it is now conceived.) Indirectly, the science of man is the basis for mathematics, philosophy of material nature, and natural (as opposed to revealed) religion. The reason is that that the conduct of these sciences depends on assessments made by the human mind. We need to know how the mind operates before we can evaluate the success of scientific practice.
Book I of the Treatise, which is the subject of these lectures, is devoted to the human understanding. Like Locke and Berkeley before him, Hume began his investigation of the understanding with a discussion of the objects, of the understanding which they (following Descartes) called "ideas." Hume used the term "perceptions" to signify the objects of the understanding (Book I, Part I, Section 1). Perceptions are distinguished into impressions and ideas. The fundamental difference between them is that impressions are livlier and more vivid than ideas. We may observe this in our own mind by comparing our present perceptions of objects (say, the bright sun) with memories of such perceptions. Although a strong memory may approach a weak sense-perception, generally they are quite different from one another--enough to mark of two distinct types.
There is something else that distinguishes impressions from ideas: the order in which they occur. A basic Humean principle is that impressions are original and ideas are mere copies of them (what we now call the "copy-principle"). That every idea is a copy of some prior impression and every impression is copied by some idea is something we can observe. This constant conjunction between the impression and idea leads Hume immediately to
conclude that there is a great connection between our correspondent impressions and ideas and that the existence of the one has a considerable influence upon that of the other. Such a constant conjunction, in such an infinite number of instances, can never arise from chance, but clearly proves a dependence of the impressions on the ideas or of the ideas on the impressions. (Book I Part 1, Section 1)The fact the the impression comes first decides the issue: it is the cause and the idea is the effect. A supporting reason for the causal dependence of idea on impression is that people without the use of one of the senses can never frame an idea proper to that sense. Someone who was never able to see cannot form an idea of red, for example. The thesis that impressions cause the existence of ideas was taken by Hume to be central to his treatment of the understanding. "The full examination of this question is the subject of the present treatise" (Book I, Part 1, Section 1).
There may seem to be an exception to the copy principle. People have imagined a "New Jerusalem," with streets paved with gold, without ever having had an impression of it. In the other direction, Hume has had impressions of Paris, where he used to live, and the idea of Paris he forms now is not a very good copy of the layout of the city. This puzzle is overcome, though, by recognizing that the copy principle applies only to simple impressions and ideas, not to complex ones.
We first receive impressions through sensation. Hume says that they arise from "unknown causes" (Book I, Part I, Section 2). Any attempt to discover the causes of impressions of sensation would be the work of anatomists and physical scientists. (Locke had similarly declined to give a detailed account of sensation, though he did think some kind of corpuscularian account holds the most promise.) Impressions of sensation give rise to ideas of sensation. But in a curious reversal, Hume allows that ideas of sensation can cause the existence of impressions of reflection. I may have an impression of pain, which produces an idea of pain. This idea in turn can give rise to various passions, such as "desire and aversion, hope and fear, which may properly be called impressions of reflection, because derived from it" (Book I, Part 1, Section 2) Hume takes this causal power of ideas as a reason to make ideas the first, and indeed the central, focus of his examination of the understanding.
Aside from the ideas produced directly from sensation, we have ideas of memory and imagination. Hume postulated that there is a faculty of the mind for producing these two kinds of ideas. The ideas themselves are distinguished in two ways. Like impressions and ideas, they can be distinguished by relative vivacity. Ideas of memory are more vivid than those of the imagination. Moreover, ideas of memory preserve the order found in the ideas which they revive. The imagination, on the other hand, freely serves up ideas in any order whatsoever, as can be seen from tales of dragons and giants found in "the fables we meet with in poems and romances" (Book I, Part I, Section 3). It will turn out that the imagination plays a central role in Hume's explanation of the how the human understanding operates.
The most important function of the understanding is to form new ideas that take us beyond our present and past circumstances. These complex ideas are built up out of simple ideas, and the main task of the science of the understanding is to discover the principles by which they are composed. In Section 4, Hume enumerates three natural principles which govern the production of complex ideas:
Hume admitted in Part III, Section 6 that these are only "general" principles. They are neither necessary nor sufficient for the production of ideas. In the case of resemblance, you might be fascinated by the person you and not think about your ex-lover. Similar experiments apply to contiguity and cause and effect. Resemblance is not sufficient for the production of new ideas. Nor is it necessary, due to the freedom of the imagination. Our minds are capable of jumping seemingly at random from one idea to another. For this reason, Hume referred to the principles together as a "gentle force" (Book I, Part I, Section 4) that moves us generally but not universally. It is as important, in Hume's science as the understanding, as the force of attraction is in Newton's science of physical bodies.
Resemblance, contiguity, cause and effect, are all what Hume called "natural relations." Such a relation is "that quality by which two ideas are connected together in the imagination and the one naturally explains the other" (Book I, Part 1, Section 5). There need be no explict comparison being made between the two here. When an explicit comparison between two ideas is made, no matter how they arise in the mind, the relation is "philosophical." Hume claims that "it is only in philosophy that we extend [the word 'relation'] to mean any particular subject of comparison, without a connecting principle" (Book I, Part I, Section 5).
As with the principles of association, Hume had an exact enumeration of the all the kinds of comparisons we make to relate philosophically two ideas.
In Section 6, Hume discusses our ideas of substances and their modes. The fact that Hume treated of ideas of relations before the ideas of substances is a break with the philosophical tradition. Earlier philosophers of the modern period regarded real relations to hold between substances and to depend on substances. Thus substance, or the idea of substance, must be treated first, in their view. But for Hume, the primary object of investigation in the science of the human understanding is the comparisons the mind makes between ideas.
Hume describes what the ideas of substance and mode have in common as being:
nothing but a collection of simple ideas that are united by the imagination and have a particular name assigned them by which we are able to recall, either to ourselves or others, that collection. (Book I, Part I, Section 6)A substance has the additional feature that either the ideas in the collection are either referred to an unknown something as the principle uniting them (as Locke had postulated and Hume called a "fiction") or else they "are at least supposed to be closely and inseparably connected by the relations of causality and contiguity" (Book I, Part I, Section 6). This has the consequence that when we find a new idea that is related to the others, we add it to the collection, "as much as if its idea had from the beginning made a part of the compound one" (Book I, Part I, Section 6). So if we discover that gold dissolves in a certain acid ("aqua regia"), we suppose that quality to belong to gold, due to the presumed principle of union.
An idea of a mode, on the other hand, is related to the idea of substance in one of two ways. One is that the substance and mode are not united by contiguity and causality, and hence distributed among many things, such as the idea of a dance. The second is that the mode is associated with the substance without reference to a uniting principle, as with the idea of beauty.
The emphasis placed by earlier philosophers upon substance is misguided, in Hume's view. If we examine the alleged idea of substance, we find that there is no corresponding impression. So, given the copy principle, we have no idea of it either. All we are left with in our idea of substance is purely relational: a collection of simple ideas.
Ideas of substances, modes, and relations are the only complex ideas of sensation that we have. Hume followed Berkeley in denying that the human mind can create abstract general ideas. Recall that ideas are copies of impressions. Impressions are always precise in their qualities, for example length. So an idea of a line will also have a precise length. The length of the line is not distinguishable from the line itself, and so it cannot be separated from the idea of the line. Like Berkeley, Hume held that particular ideas may be used in a general way "by being annexed to a general term, that is, to a term which, from a customary conjunction, has a relation to may other particular ideas and readily recalls them in the imagination" (Book I, Part I, Section 7).
Our text contains the last section of Part II of Book I. The rest of Part 2 deals with the ideas of space and time. In Section 6, Hume treats the notion of "external" existence, which has a superficial connection to the idea of space. The notion of external existence is really metaphorical: to exist externally is to exist "outside" the mind, or independently of sense-perception. Hume follows Berkeley in denying that we can conceive of external existence.
Let us fix our attention out of ourselves as much as possible, let us chase our imagination to the heavens or to the utmost limits of the universe, we never really advance a step beyond ourselves, nor can we conceive any kind of existence but those perceptions that have appeared in that narrow compass. This is the universe of the imagination, nor have we any idea but what is there produced. (Book I, Part 2, Section 6)Hume does not at this point endorse phenomenalism. The issue of the existence of bodies is treated later, in Part IV, Section 2. Here, he is only making a point about the alleged concept of external existence. Hume does, however, offer a small morsel of what is to come. He tells us that although the idea of an externally existing thing is not "specifically" different (i.e., different in kind or species) from perception, we nonetheless consider perceptions "in different relations, connections and durations" (Book I, Part II, Section 6). This hints at phenomenalism, as if "external objects" are nothing but perceptions connected in a certain way, but Hume self-consciously begs off the topic until Part IV. In Parts I through III, his only concern is to describe what goes on within the mind, not its consequences for metaphysics.
More generally, as did Berkeley, Hume claims that we cannot separate existence from any idea. If we form an idea, we consider it as existing. "To reflect on something simply and to reflect on it as existent are nothing different from each other. That idea, when conjoined with the idea of any object makes no addition to it" (Section 6). This stance, if correct, would defeat the ontological argument for the existence of God. Descartes had claimed that the idea of God is unique, in that it is the only one that has existence necessarily attached to it. But for Hume, all ideas have existence necessarily attached to them. If you could prove that God exists from the bare idea of God, you could prove that anything exists from the bare idea of it.
Part III of Book I of the Treatise is entitled "Of Knowledge and Probability." In this part, Hume develops an account of the nature of knowledge and of probability. In Part IV, he goes on to ask what we know and what is probable for us.
Hume discusses knowledge in Section 1, and there he makes only a passing reference to what knowledge is, when he states that only four of the seven philosophical relations "can be the objects of knowledge and certainty." It is not until Section 11 that he gives a full account of what knowledge is. Here he gives a definition of knowledge which he attributes to some unnamed philosophers, namely, "that evidence, which arises from the comparison of ideas," or "the assurance arising from the comparison of ideas."
This is similar to how Locke defined it: "the perception of the connection of and agreement, or disagreement or repugnancy of any of our ideas" (An Essay concerning Human Understanding, Book IV, Chapter 1). Locke used the example of our knowledge that white is not black: "what do we else but perceive, that these two ideas do not agree?" Similarly, our knowledge that the three angles of a triangle are equal to two right angles is based on the perception "that equality to two right ones does necessarily agree to, and is inseparable from the three angles of a triangle."
In Section 1, Hume applies this Lockean conception of knowledge to the seven philosophical relations. He finds that only four of them can be established solely on the basis of the perception of the related ideas.
Relations of proportions in quantity or number may be intuited if the differences are great enough. One might observe that one figure is larger than another, or that one number is greater than another. In many cases, however, the proportions can only be determined by demonstration, as occurs in arithmetic and algrebra. Geometry is a special case, because our powers of observation are not acute enough to detect the small differences about which we reason, and so it does not result in the same "perfect exactness and certainty" as arithmetic and geometry. This deficiency is discussed at length in Part II.
The remaining three philosophical relations cannot be the objects of knowledge:
Suppose you observe a block of marble at one time, and at a later time you observe a block of marble in the same place, and which looks exactly like the first block. Are the two the same or not? Two cases are possible. One is that the original block continued to exist in its location while you were not observing it, in which case there is a relation of identity. A second is that there was a switch, and now a different block is in the same place. Because there is no detectable difference in the ideas, the relation of identity would hold or not hold without any "change in the ideas."
Since our beliefs about these three relations fall short of knowledge, it might be thought that they therefore amount only to probability. The title of Section 2, "Of probability; and of the idea of cause and effect" reinforces this conclusion. But remarkable, probability is not mentioned once in the body of the section. It is only in Section 11 that Hume gives a definition of probability, i.e., "reasoning from conjecture" and "that evidence which is . . . attended with uncertainty." (Section 11 is not reproduced in our text. Click here for an on-line version.)
Beliefs about relations of identity and situation in time and place fall under the definition of probability, because the mere inspection of the ideas does not allow us to "discover the real existence or the relations of the objects" of those ideas (Section 2). Thus we are uncertain about whether these relations hold. Causality, on the other hand, promises to bridge the gap between ideas and existence. In our example of the piece of marble, we can secure the identity of the object if we can assure ourselves there is a cause that insures that it remains the same.
In Section 11, Hume will refer to some products of our causal reasoning as "proofs." When we reason according to the relation of cause and effect, and our reasoning leaves us "entirely free from doubt and certainty," then we have proved our conclusion. Proof does not have the authority of demonstration, which positively connects ideas in our perception, but it lacks the uncertainty of probability. Before Section 11, however, Hume treated all causal reasoning as probable reasoning.
So the relation of cause and effect is central to our understanding of the world. It allows us to go beyond our ideas to discover real existences, and it may allow us to remove any uncertainty about the real existences that are discovered by this use. For this reason, it is the subject of Sections 3 through 10 of Part III. The issue that Hume raises, and which will concern us at length here, is about the origin of the belief that some object A causes a change in an object B. Beliefs of this kind are the basis of causal reasoning. How do we come by such beliefs, and in particular, are they the products of reason?
Following his general principle that ideas are copies of impressions, Hume seeks the impression that gives rise to the idea of cause and effect. His quest begins in Section 2 of Part III and is not completed until Section 14. The path he takes is quite round-about.
The impression will not be one of a particular quality, because no quality explains all cause/effect relations. Any kind of object, with any qualities at all, may be a cause or an effect. A quality such as Locke's impulse, for example, is not present in cases in which the cause is the human mind. If the idea of causation is not derived from the idea of a quality, it must be derived from an idea of a relation. So the next task is to discover the characteristics of our idea of the relation of causality.
Hume observes that if we have an idea of A as the cause of B, then we think of A as being related to B in three ways:
As Hume noted, something more than contiguity and succession are needed to have an idea of causation. He discovered that the further ingredient is that there is a necessary connection between A and B. "That relation is of much greater importance than any of the other two above mentioned" (Book I, Part III, Section 2). As Hume puts the matter later, in Section 14, the task at this point is to answer the question: "What is our idea of necessity when we say that two objects are necessarily connected together?" As with all ideas, the idea of necessity must be traced to original impressions, but it is hard to see what such an impression might be. For the same reason that applied to the idea of causation of which it is a component, the idea of a necessary connection is not derived from the idea of a quality in the objects. Further, the relations of contiguity and succession are not sufficient to establish a necessary connection. It will cost Hume a great deal of labor to come up with a plausible candidate.
Rather than attacking the problem directly, Hume suggests that it be approached through two different questions, which might provide a hint as to what the required impression might be.
Philosophers have always claimed that we have intuitive knowledge of the general principle that "whatever begins to exist, must have a cause of existence." Hume sets out to undermine the assurance the philosophers have about the truth of this principle. First, it is not one of the four kinds of relations which are known with intuitive certainty. Second, Hume presents an argument to the conclusion that there is no other relation that is intuitively known that could serve as the foundation for knowledge of the general principle that there is a cause for every beginning of existence.
Hume claims that any demonstration of the claim that every beginning of existence must have a cause requires a demonstration of the related claim that it is impossible that something can begin to exist "without some productive principle." This second proposition "is utterly incapable of demonstrative proof," and so the latter principle is also incapable of being proved.
The argument that there is no proof is devastatingly simple. The idea of an effect and its cause (productive principle) are distinct. But what is distinct is separable in thought, so we can conceive of the effect without the cause. "It will be easy for us to conceive any object to be non-existent this moment, and existent the next, without conjoining to it the distinct idea of a cause or productive principle." Because the cause and effect can be separated in thought, it is at least possible that they are separated in reality, in which case "reasoning from mere ideas" can never demonstrate their connection. (This argument is repeated in the first paragraph of Section 6.)
Hume goes on to show how all previous attempts by philosophers to demonstrate the principle are "fallacious and sophistical." Locke thought that it is intuitvely known "that bare nothing can no more produce any real being, than it can be equal to two right angles" (An Essay concerning Human Understanding, Book IV, Chapter X, Section 3). The problem with this claim is that the denial of the principle is simply that there is no cause for the existence of some being, not the claim that "nothing" is the cause. As Hume puts it, if we exclude every thing from being a cause, "we really do exclude them." Similar problems are raised for the arguments of Hobbes and Clarke. An even worse argument is that a beginning of existence must have a cause because it is an effect. But to suppose that a beginning of existence is an effect is question-begging: it is just to suppose that it has a cause.
Having shown to his satisfaction that there is no knowledge of the principle that every beginning of existence has a cause, Hume concludes that the principle can only be probable, being based on observation and experience. The obvious question to ask at this point is how observation and experience give rise to our belief in the principle. But Hume chooses to "sink" this question in another question, i.e., "Why do we conclude that such particular causes must necessarily have such particular effects, and why do we form an inference from one to the other?" He surmises that the answer to the second question will be an answer to the first as well.
So the task now is to explain how we make the particular causal inferences that we do. In all causal reasoning, we begin with a perception of something A and end with an idea of B either as effect or as cause. As noted earlier, the having of an idea is the same thing as taking it to exist, so the end-product is an idea of B existing. Finally, this idea of B existing is not merely imagined, but rather has the status of a belief. How does this take place?
In Section 7, Hume describes the process of causal inferences. We observe that they take place only after a constant conjunction or repetition of perceptions like A and ideas like B. We recall this constant conjunction and, "without any further ceremony, we call the one cause and the other effect, and infer the existence of the one from that of the other." This relation of constant conjunction turns up "where we least expected it, and were entirely employed upon another subject."
As an illustration of the process, consider a case where a person has never witnessed a certain effect, say what happens when a piece of gold is submerged in aqua regia (a mixture of nitric and hydrocholoric acids). The gold might turn the mixture a certain color, float to the top, dissolve, etc., for all one can tell by considering the gold in itself. But after repeated trials, it is observed the the gold dissolves every time. Then the belief that gold will dissolve comes upon us when we observe its being placed in aqua regia. Granting that this is how the causal inference is made, we are still left with a puzzle. What is it about repeated trials that moves us to form the belief?
Here Hume says that an answer may be forthcoming if we consider which faculty of the mind produces the belief: reason or the imagination. So the first issue is, can reason exploit the fact of constant conjunction to yield beliefs? The answer is no. Our reasoning would have to be based on the principle "that instances of which we have had no experience must resemble those of which we have had experience, and that the course of nature continues always uniformly the same" (Book I, Part III, Section 6). This principle itself cannot be proved, because we are able to conceive of a situation in which there is no resemblance and nature is not uniform. This is a reprise of the reasoning used above to show that there is no knowledge of the principle that every beginning of existence must have a cause.
So the belief in B would have to be produced by probable reasoning. Somehow the fact that ideas like B generally accompany perceptions like A makes it probable that B will accompany A in the present case. The reasoning might go like this.
The question that now arises concerns the status of the new premise 3, which we will call the "uniformity principle" or "UP." For the belief in B's existence to be produced by the argument, there must be a belief in UP. That is, the belief in UP plays a causal role in the production of conclusion 4. So we have to ask about the cause of the belief in UP. It is not intuitively or demonstratively certain, as it we can easily imagine its volation. So if it is to be the product of reason, the belief must be caused by probable reasoning.
But now an insurmountable problem arises. Probable reasoning has just been postulated as the cause of the belief in UP, yet we as we observed in the second argument, the belief in UP is at least a partial cause of probable reasoning. So if probable reasoning is the cause of belief in UP, causal reasoning must be a cause of itself, which is absurd. As Hume puts the matter, "The same principle cannot be both the cause and effect of another, and this is perhaps the only proposition concerning that relation which is either intuitively or demonstrably certain" (Book I, Part III, Section 6). So probable reasoning cannot be said to be the cause of our belief in UP. "We suppose, but are never able to prove that there must be a resemblance between those objects of which we have had experience and those which lie beyond the reach of our discovery" (Book I, Part III, Section 6).
Without UP to assist it, probable reasoning does not establish the probability of the existence of B. Instead, "When the mind therefore passes from the idea or impression of one object to the idea or belief of another, it is not determined by reason, but by certain principles which associate together the ideas of these objects and unite them in the imagination" (Book I, Part III, Section 6). The thesis that belief in causal relations is not established by reason is perhaps the most famous in all of Hume's writings.
The reader may have noticed a recurring pattern in Hume's argumentation. Several times he sets up two possibilities: a given belief is based either on reason or on experience. It is shown that it cannot be based on reason, due to the conceivability of the opposite. Then it is shown that being based on experience has further consequences, that are ultimately quite devastating. This pattern of argumentation has come to be known as "Hume's fork," which is a specfic case of argument by dilemma.
We are still left with the question of which principles are responsible for uniting the perception of A with the belief in the existence of B. Perhaps a clue can be found by examining what we mean when we say that someone has a belief. Everyone can tell the difference between his belief in the existence of B and his mere idea of the existence of B. Yet it is hard to say what the difference is. According to Hume, what distinguishes the two cannot be found in the content of the idea of B, but only in the manner in which B is conceived. Our mere ideas are not nearly so lively and vivacious as those ideas which amount to belief. So Hume defines belief as "a lively idea related to or associated with a present impression" (Book I, Part III, Section 7).
Now we may link this account of belief to the problem at hand, i.e., how it is that the mind produces a belief in the existence of B on the basis of the perception of A and the constant conjunction of perceptions like A and ideas like B. It must be that the constant conjunction enlivens the idea of B, elevating it to the status of belief. The belief is said to be based on "custom," which carries us immediately from the perception of A to the idea of B. It is not based on anything like a "power" in A or B taken by themselves.
What we call probable reasoning, non-demonstrative reasoning from cause to effect or from effect to cause, turns out to be based entirely upon custom, which in turn is what produces a lively idea given a present impression. Hume states this thesis dramatically.
Thus, all probable reasoning is nothing but a species of sensation. It is not solely in poetry and music we must follow our taste and sentiment, but likewise in philosophy. When I am convinced of any principle, it is only an idea which strikes more strongly upon me. When I give the preference to one set of arguments above another, I do nothing but decide from my feeling concerning the superiority of their influence. Objects have no discoverable connection together, nor is it from any other principle but custom operating upon the imagination that we can draw any inference from the appearance of one to the existence of another. (Book I, Part III, Section 8)This thesis is supposed to be the direct outcome of the application of the experimental method to the workings of the human mind. It should be noted that the use of the experimental method is itself a kind of causal reasoning, so Hume's conclusion applies to the reasoning that established it as well.
Recall that the goal in undertaking this circuitous investigation of causal inference was to try to discover the impression that gives rise to the idea of necessary connection, which is a component of the idea of cause and effect. Now Hume has all that he needs to accomplish his aim. When we reason causally we make an immediate transition from a perception to a lively idea. When we feel the force of that transition, we have the impression that gives rise to the idea of a necessary connection. We think that B is tied necessarily to A because we are determined to believe in the existence of B when we have the perception of A. "It is this impression, then, or determination, which affords me the idea of necessity" (Book I, Part III, Section 14).
This treatment of necessity is quite shocking, as Hume recognized. "I am sensible, that of all the paradoxes, which I have had, or shall hereafter have occasion to advance in the course of this treatise, the present one is the most violent." It defies the opinions of nearly all earlier philosophers regarding "one of the most sublime questions of philosophy, namely that concerning the power and efficacy of causes, where all sciences seem so much interested" (Book I, Part III, Section 14).
The heart of the problem is that previous philosophers had claimed knowledge of these matters through innate ideas. Hume took it that the innate ideas doctrine was throughly discredited, which meant that the notions of power, etc. must be based on experience.
The principle of innate ideas, which alone can save us from this dilemma, has been already refuted, and is now almost universally rejected in the learned world. Our present business, then, must be to find some natural production, where the operation and efficacy of a cause can be clearly conceived and comprehended by the mind, without any danger of obscurity or mistake.
What is so shocking about Hume's account is that it discovers that causality is a subjective relation, rather than the objective relation philosophers had thought it to be. "Upon the whole, necessity is something that exists in the mind, not in objects" (Book I, Part III, Section 14). Hume accounted for the natural tendency to view causality as objective by appeal once again to the imagination. It is due to the mind's "great propensity to spread itself on external objects" (Book I, Part III, Section 14), projecting what it finds in its impressions and ideas on the object itself.
It is important to note that Hume compares the subjectivity of causation with what he finds to be the subjectivity of mathematics. The necessity that 2 x 2 = 4 "lies only in the act of the understanding by which we consider and compare these ideas" (Book I, Part III, Section 14). Most philosophers of the period found mathematics to be the hallmark of objectivity. Thus Hume's subjectivism is quite profound and extensive. It was left to Thomas Reid and Immanuel Kant to take up the challenge of refuting it.
The Definition of Cause and Effect
Now that the origin of the idea of causality has been discovered in a round-about way, Hume is finally in a position to give definitions of this idea. There are two definitions, since cause and effect is both an philosophical and a natural relation. That is, it is a relation that we find when comparing ideas and a relation which is responsible for the transitions that take place in the mind, respectively. The definitions of 'cause' are as follows:
An object precedent and continguous to another and where all the objects resembling the former are placed in like relations of precedence and contiguity to those objects that resemble the latter.
An object precedent and contiguous to another and so united with it that the idea of the one determines the mind to form the idea of the other and the impression of the one to form a more lively idea of the other. (Book I, Part III, Section 14)Having given these definitions, Hume is able to back up his claim in Section 3 that there are no demonstrative or intuitive arguments to prove that every beginning of existence must have a cause. The first definition makes no mention of "absolute" or "metaphysical" necessity, and so we can conceive of a beginning of existence without there being any resemblance, precedence and contiguity as in all experienced beginnings of existence.
Hume thought that the second definition of "cause" makes it more clear that the principle in question cannot be proved to be true. The "determination of the mind" is something for which we have no evidence except observation and experience, and in fact is itself "perfectly extraordinary and inexplicable."
With this belated definition of cause in hand, we may now backtrack to see what Hume had to say about how we reason about causes, where this reasoning yields only probability (or proof) and not knowledge. Here we get a look at the positive side of Hume's doctrine of causality, viewed as a philosophical relation. Hume grants that some "kinds of probability are received by philosophers, and allowed to be reasonable foundations of belief and opinion" (Section 13). He sets out to describe the reasonable, or "philosophical," kinds of probability (Sections 11 and 12), and then the unreasonable, or "unphilosophical," kinds (Section 13).
Here we will take a brief look at the treatment of the primary kind of philosophical probability, the probability of causes (Section 12). The paradigm of causal reasoning is as follows:
Unfortunately, things usually do not go so smoothly. We often find instances of contrary cases which introduce a layer of uncertainty, diminishing the probability that As cause Bs. Uneducated people believe that contrary cases signify an breach of causal influence, as if causes sometimes fail to operate uniformly. But philosophers recognize that they signify instead that the causal processes involved are more complicated than they appear. There are hidden causes at work when there is contrareity, such as a piece of sand stopping the workings of a watch.
The psychological effect of contrariety on our imagination is that it interrupts the habit of mind to expect things to continue the way they have been. A number of different possibilities come into view. "The first impulse [to transfer the past to the future full and perfect] therefore is here broke into pieces and diffuses itself over all those images of which each partakes an equal share of that force and vivacity that is derived from the impulse" (Book I, Part III, Section 12). So the probability of any one cause is the share of vivacity that cause has in comparison to the vivacity of the others.
Part III of the Treatise presents an account of the workings of the human mind, which turn out to be mostly acts of the imagination, under the influence of habit. In Part IV, Hume turns his attention to skepticism about the powers of the human mind. Here, he advances a number of skeptical arguments, which he seems to have thought are irrefutable. And yet he draws back from the abyss of skepticism in most cases.
Skepticism, as Hume saw it, is not a merely theoretical position. Instead, it is the absence of belief, induced by reflection on the shortcomings of the human cognitive faculties. A loose connection between As and Bs, due to observed cases of Bs without As or As without Bs, only results in a fairly weak belief. By contrast, one would be a skeptic if one considers one's faculties as incompetent to draw a connection between As and Bs.
In Section 1 of Part IV, Hume presents a skeptical argument that threatens to induce skepticism with respect to all the conclusions of human reasoning. The first step is to reduce all demonstrative knowledge to probability. (Hume in fact claims that "all knowledge resolves itself into probability," but his argument applies only to knowledge that is the result of reasoning, rather than intuition.)
Demonstrative knowledge, it will be recalled, is the result of the comparison of ideas with one another, in a way that requires no information about any really existing things. Mathematical knowledge is of this kind. But even though no information outside the confines of the mind is required to compare the ideas, the mere appearance of agreement or disagreement is not enough for knowledge. We need further the assurance that we have made the comparison properly. But this is not ascertained merely by comparison of ideas, as it concerns the proper operation of a causal mechanism. So probability is an essential factor in all so-called demonstrative "knowledge." But if probability contaminates knowledge and reduces it to its own level, i.e., all "knowledge" is nothing but a species of probability.
Recall that probable reasoning involves an uncertainty due to incomplete experience, contrary causes, and imperfect resemblance. We arrive at a probable conclusion due to this "original uncertainty inherent in the object." Hume claims that reason obliges us to consider a second uncertainty, namely, that our faculties may not have operated rightly in the formation of the first probability. This requires us, Hume contends, to adjust the first probability downward, even if experience shows that we are quite good in forming this kind of judgment. Hume seems to have held the view that when two uncertainties are compounded, no matter what they are, the result is even greater uncertainty.
But this is not the end. The initial evaluation of one's competence must itself be evaluated, which further reduces the original probability. This process would go on to infinity, with the original probability diminishing at each step, until at last the original probability would be reduced to nothing. Since probability is just the vivacity of an idea, it is the same as the degree of belief, and so belief is reduced to nothing.
This reasoning is perfectly general, and so understanding the argument threatens to induce total skepticism about the conclusions drawn by reasoning. But in fact, this does not happen. Something stops the reduction of the vivacity of belief before much damage could be done. This something is the strain that this endless self-evaluation would produce. The more the mind struggles to comprehend what it is doing in the process of self-evaluation, the less an effect on belief the process has. So belief is never in any real danger of diminution by this process of self-examination. Nature has seen to it that we do not out-smart ourselves in this way. "It is happy . . . that nature breaks the force of all sceptical arguments in time, and keeps them from having any considerable influence on the understanding" (Section 1).
Berkeley had put forward a number of arguments intended to show that extended solid bodies are collections of ideas, dependent on minds that perceive them. In Section 2 of Part IV, Hume considers skepticism with regard to the existence of mind-independent bodies, though without ever endorsing it fully. He begins with the affirmation that nobody can doubt that bodies exist, and he proceeds to develop an account of the causes of our belief in the existence of bodies. Yet by the time he completes the account, he finds that it has problems that promote skepticism.
One would expect a discussion of the cause of our belief in the existence of bodies to begin with some account of what bodies are supposed to be. But Hume did not begin in this way, perhaps because the outcome depends on his account of the idea of substance, which is found in two later sections, 3 and 5. Instead, Hume ask the question why we believe that bodies exist independently of the mind (exist "distinctly" from the mind). Then he adds a new wrinkle by asking the parallel question why we believe that bodies continue to exist even when not present to the senses.
Hume's view was that although we may come to believe these two things in different ways, it is a fact that bodies exist independently of minds if and only if they exist when not present to the senses. If a body exists independently of minds, then its existence is not affected by whether we perceive it at any given time. Since we do not perceive all bodies all the time, they exist when we do not perceive them. And if they exist when we do not perceive them, the only way this could happen is if they exist independently of their being perceived. It will turn out on Hume's account that our belief in their independent existence depends on our belief in their continued existence.
Hume's strategy in deciding this question is to run through the three sources of belief (lively ideas) in the human mind. They are:
That the senses can produce belief in neither continued nor distinct existence is pretty obvious, though Hume considered a number of arguments in support of this claim. The senses yield impressions, which are fleeting perceptions dependent on the mind, and hence are not continued or distinct existences themselves. We might wish to say that they are images of something distinct from perception, but Hume asks how a single impression could give rise to the belief in a second object, a body distinct from them.
Most philosophers have held that it is reason that produces belief in continued and distinct bodies through a process of inference from our sense impressions. But here Hume notes that the reasons cited by the philosophers are subtle and are unknown to the great masses of people, who manage to produce their beliefs without them. In fact, the "vulgar" do not even distinguish between perceptions and objects distinct from them, according to Hume. Since, as we have seen, this is quite unreasonable, reason is not the source of their belief. (One might ask here whether the belief could be produced by bad reasoning, but Hume seems to have thought of reason as naturally leading to the truth: "Our reason must be considered as a kind of cause, of which truth is the natual effect" (Section 1).)
We are left with the imagination, which Hume had already shown is capable of inducing belief in relations of cause and effect. Imagination gives rise to belief in continued existence in a way that can be explained similarly to how belief in causes and effects is explained. Beliefs exist when ideas are made vivid by a process of the imagination.
There are two features of our perceptions which are the bases for the enlivening work of the imagination. The first is the constancy of perceptions. Many times, we perceive an object, turn our attention away, then find perceptions that greatly resemble the originals. But things often change a great deal when we are not observing them, so constancy is not the only factor at work. The other is coherence, the fact that these changes occur in an orderly way. We make predictions about how our perceptions will change, and they largely turn out to be right.
What coherence provides to the imagination is the basis of a kind of causal reasoning (which, it must be understood, is a function of the imagination). Suppose you are handed a letter and told that it was delivered to you from abroad. Your mind would not rest easy with the supposition that it just popped into existence at the present time. Rather, you are comfortable when this event fits your expectations. In this case, the most coherent account of the matter is that the letter was written some time ago and traversed a great distance, intact, before being placed into your hands.
It is a feature of the imagination that it endeavors to preserve the unity of perceptions, and the postulation of continued existence suffices for this purpose. Hume compares this process to the continued motion of a ship (specifically, a galley) after the crew has stopped rowing. It is a kind of smooth transition, from being powered to coasting at increasingly slower speeds, similarly to the transition of the mind from the idea of an effect to an idea of its cause.
Coherence alone cannot account for the continued existence of bodies. If bodies were not constant as well, if they did not preserve most of their features most of the time, there would be no basis for coherence. If letters, for example, are observed sometimes to vanish and then reappear with no explanation, there would be no way to explain the existence of the letter just delivered to me.
But constancy creates a problem for the common view of things ("the vulgar system"), which according to Hume is that there is no difference between objects and perceptions. Given this identification, how can an object continue to exist when the perception ceases to be? This "kind of contradiction" makes the mind uneasy. In its struggle to reconcile the conflict, the imagination will "disguise, as much as possible, the interruption, or rather remove it entirely by supposing that these interrupted perceptions are connected by a real existence of which we are insensible" (Book I, Part IV, Section 2). The uniformity of actual perceptions and the regularity of memory reinforce this tendency of the imagination to create a vivid idea of the unperceived object, which is a belief in its continued existence.
Unfortunately, this attempt to resolve the conflict through the creation of a fiction of perceptions which continue to exist when not perceived is fallacious. This can be seen by considering belief in independent existence, which is required to explain how objects continue to exist when not perceived. It is an outright contradiction to suppose that a perceived object exists independently of the mind, as Berkeley had pointed out. Here, Hume's discussion in Part II, Section 6, of the idea of external existence finally bears its skeptical consequences.
Philosophers understand this conflict and try to avoid it by distinguishing between the perception and the object. It is the object that continues to exist, that is independent of the mind, rather than the perception. This is the "philosophical system" of Descartes and Locke. Hume thought that this system is of no real value. First, it is impossible to establish that the system is true. As has already been seen, reason is of no use in establishing mind-independent objects, as its data are restricted to perception, and it can only project patterns of perception into the future. And how could the imagination come up with such an idea, since it only makes a transition from a perception to another idea?
The second problem with this distinction beween perception and object is that there would be no reason to make it at all if one did not begin with the vulgar system. According to the vulgar system, our perceptions continue to exist even when not perceived. But "a little reflection destroys this conclusion" (Section 2). Rather than draw the proper conclusion, that there just is no continued existence, philosophers instead try to patch up the vulgar system by postulating a new object which can be said to continue to exist.
Hume regarded the philosophical system as a "monstrous offspring" of two views that cannot both be true. . Our imagination intimates that perceptions continue to exist, while reflection tells us that perceptions are interrupted in their existence. The philosophical system tries to have it both ways. There would be no need to make this compromise if we were fully convinced of either system. We could stick with the vulgar system and deny that our perceptions are interrupted, or we could satisfy reflection by simply denying that there is continued existence.
Because of these difficulties, Hume finds himself, when thinking of them, inclined toward skepticism regarding the beliefs produced by the imagination. They spring entirely from two features of perception, coherence and constancy, that Hume regarded as "trivial." He does not see how any "solid and rational system" can be devised. This skepticism recurs and can never be totally eliminated. The only way to avoid it is by being careless and inattentive, i.e., simply by ignoring the difficulties. Nature will restore our beliefs, which after all are too important for her to trust to philosophical theorizing.
Section 3 of Part IV examines the "ancient philosophy," with its emphasis on substance and the qualities that are supposed to depend on it. Not surprisingly, Hume argues that there is no impression of which the idea of substance is a copy. In Section 4, he turns his attention to the "modern philosophy."
One way of understanding the difference is that the ancient philosophy appeals to principles of the imagination that are "changeable, weak and irregular," while the modern philosophy professes to appeal to "permanent, irresistible, and universal" principles of the imagination. (Book I, Part IV, Section 4). Hume's own principle of causation is of the latter type. So he asks the question whether the principles of the "modern philosophy" are of the right type, or whether instead they share the problems of the principles of the ancients. Not surprisingly, Hume finds the modern philosophy lacking.
The principles of the modern philosophy involve the distinction between primary and secondary qualities, which is really a principle of ancient atomism that can be found in Descartes and as is made explicit in Locke. As Hume puts it, there are two principles, one negative and one positive.
Hume offered a different argument against the distinction--one that he thought "very decisive" (Book I, Part IV, Section 4). Consider extension, which Descartes had taken to be the principal attribute of bodies. Our idea of extension, Locke had claimed, comes to us from vision and sight (An Essay concerning Human Understanding, Book II, Chapter 5). But what is it, Hume asks, that vision presents to the mind as extended? All we find is an array of colors. But colors, according to the modern philosophy, resemble nothing in the object. And we cannot represent real extension through a series of perceptions of this kind.
This leaves touch as the only source of a resembling idea of extension. Again according to Locke, touch gives us an idea of solidity, and we can compose an idea of an extended solid object by repetition of feelings of solidity (An Essay concerning Human Understanding, Book II, Chapter 4). But here we run up against a problem. We only feel solidity as the resistance of one object to the motion of our body. If you fall down and hit the floor, which does not give way, you understand what solidity is.
Unfortnuately, this way of sensing solidity presupposes that there are solid objects which do not give way on impact. We must already know that there are solid objects in order to arrive at a notion of solidity. But the fact that objects are solid is what was to be shown, so the argument is circular and hence fallacious. Hume advanced other arguments to undermine the claim that our idea of solidity resembles a quality in objects, but they will not be discussed here.
Sections 5 and 6 both concern what we might call the self. There has always been much dissention in philosophy over the nature of the self. Some hold that the self is purely physical, some that it is a composite of mind and body, and some that it is purely mental. The latter view can be found in Descartes (though he sometimes spoke of the self as a composite). Locke had claimed that there is a mental substance which supports all the perceptions ("ideas" for Locke) of the mind.
Hume found the alleged notion of a support of perceptions to be unintelligible. Once again, he challenges the reader to find the impression from which the idea is derived. Which among our many impressions is the one that represents the self? Does it come from sensation or reflection? Is it pleasant or painful? Does it endure, or does it come and go? Hume was confident that no answer could be given to these questions.
With the results of Section 3 in mind, we can then say that there is no idea of either material or mental substance. All we have to work with are perceptions which look and feel different to us. Still, they come to us in very uniform patterns, so much so that we regard the objects of these perceptions (if there is indeed any difference between perception and object) as standing in causal relations. I approach too near the fire, and I feel pain. This happens again and again, without exception. So the imagination regards the fire as the cause of pain. There is no need to explain any "power" in the fire to cause a perception in the mind. Hume's system neatly takes care of what we now call the problem of mind-body interaction.
If there is no mental substance, can we say that there is an idea of an enduring self? Again, Hume looks for the impression and comes up empty. All we find is individual perceptions that we associate with one another. We know of no basis for associating our perceptions except for the activity of the imagination. We create a fiction of identity when there is enough similarity among perceptions to yield a smooth transition between distinct perceptions. We arrive at this idea in the same way when we think of animals and plant as identical things persisting over time. But in all these cases, just as with the case of "power," any legitimate idea we have of identity is purely subjective.
We began our discussion of Book I of the Treatise by noting that its goal was to produce a scientific account of the cognitive operations of the human mind. The result is a very austere system, with one kind of object, perceptions, and a few ways in which these are produced. Sensation and reflection provide us with the original perceptions, and all else is, at bottom, the work of the imagination. The underlying principle of the imagination is that of smooth flow or easy transition from one perception to the next. From these meager starting-points, Hume professed to explain all our most important beliefs.
The common people have no notion of how their minds work, and they rest content in the beliefs imposed by the imagination, be they fictitious or real. But the philosopher is in a different position. Once he understands that imagination is responsible for our belief, and that even the basis of our causal reasoning, is the product of the imagination, he must wonder whether any of our beliefs are true. Hume wrestled with this problem in the concluding Section 7 of Part 4.
As Section 7 is not reprinted in our text, not much will be said about it here. What can be said is that the skeptical sentiment Hume reported in Section 2 grows upon him until he despairs of ever attaining the truth. But this skeptical sentiment is broken by nature, when he dines, converses, plays backgammon, etc. He could remain in the non-skeptical state of the common person, but he wants to be able to do philosophy, especially to establish a scientific account of morality.
The solution to this problem is continue to philosophize, but in a careless way. This means that he will not pay attention to skeptical problems as he proceeds. Instead, he will try to produce a system that is "satisfactory" and that will stand the test of time, even if it is subject to skeptical doubts. This is enough to satisfy his curiousity and his ambition. And, so long as it does no harm, there is no reason not to indulge one's self in such a pleasant enterprise.
[ Course Home Page | G. J. Mattey's Home Page | UCD Philosophy Department Home Page ]
This page and all subordinate pages copyright © G. J. Mattey, 2003, 2006, 2007. All rights reserved.