Philosophy and Philosophers - an Introduction to Western Philosophy - Chapter 5
back to ...
or ...
CHAPTER FIVE
Empiricism: Locke, Berkeley, Hume
The empiricists in general have tendencies which contrast with those
of the rationalists. Empiricists hold that all the material for knowledge,
our ideas or concepts, and all knowledge of actual matters of fact, as
opposed to logical or conceptual truths, must be derived from, or be
reducible to, aspects of our experience: features of the information
provided by the content of our senses and introspection. Empiricists
deny that it is possible to know by reason alone the nature of what
exists; rather, the nature of what exists can be known only through
experience. We should reject as meaningless ideas or concepts which
cannot be specified as corresponding to any possible experiences. We
should reject knowledge claims concerning matters of fact about the
nature of the world which are not supportable by the evidence of
experience. This leads to a tendency among empiricists to emphasize
that the limit of human knowledge and imagination is bounded by the
limit of our experience. Empiricists reject the rationalist claim that it is
possible to come to know by a priori reason alone the nature of an
intelligible real world inaccessible to experience that stands beyond
appearances. The empiricist may argue that concepts (such as
substance), and the terms that express them, are meaningless or else
must relate to some possible experience, since concepts and terms get
their meaning by reference to some possible experience, but a world
beyond experience cannot be a world that might possibly be
experienced; in either case it is not possible to use meaningful concepts
to talk of a world beyond possible experiences.
The tendency in empiricism is also to deny the existence of natural
necessity: necessity is a property only of logical relations between
concepts, or of logical relations between ideas or thoughts, not
between things or events in the world whose existence, nature and
connections are all contingent; such natural contingent connections can
be discovered not by reason, which can establish only necessary truths
and necessary connections, but only by experience.
114
Empiricism is inclined to argue that there are two exclusive and
together exhaustive types of proposition.
(a) Propositions whose truth, logically speaking, can be known
merely by understanding them, or by deductive reasoning alone,
independently of the evidence of experience: truths of reason.
(b) Propositions whose truth, logically speaking, cannot be known
merely by understanding them, or by deductive reasoning alone,
but which depend on the evidence of experience: truths of fact.
All propositions which tell us anything about the real or actual world
are truths of fact. Propositions stating matters of fact cannot be known
to be true merely by our understanding them, or by our deducing them
from other propositions known to be true by the understanding alone;
if we can know them to be true at all, they must be known through
consulting experience. It should be noted that the distinction is not the
genetic one of how we come to have, acquire, or understand these
different sorts of proposition, but a logical question concerning on
what, once acquired or understood, the truth or falsity of a proposition
depends, and on what knowledge of the truth or falsity of a
proposition depends. If the truth or falsity of a proposition depends
only on the meaning of the terms in it, then it is an a priori proposition
whose truth or falsity can be known a priori by reason alone
independently of empirical evidence. If the truth or falsity of a
proposition does not depend only on the meaning of the terms in it,
then it is an a posteriori proposition whose truth or falsity can only be
known a posteriori by empirical evidence, not by reason alone.
The basic contrast between rationalism and empiricism is an
argument about the extent and nature of what truths it is logically
possible to know a priori by the understanding independently of
experience, by intellectual intuition and pure logical reasoning alone,
and what truths it is logically possible to know a posteriori by the
senses, by experience and observation alone. The rationalist argues
that certain things can be known with certainty to be necessarily true
about the nature of reality, what exists, by a priori reason alone, even
if such truths refer to a reality that lies behind appearances. This the
empiricist denies, arguing that claims to knowledge of truths
concerning the nature of reality or the actual world must seek their
justification, if such justification is possible at all, in experience; a
priori reason alone cannot reveal the real or actual nature or existence
of the world. Reason alone can give knowledge only of what is
necessary (that which must be because its denial is contradictory),
impossible (that which cannot be because its assertion is
contradictory), and possible (that which may or may not be because
its denial is not contradictory), but not what is actual among what is
merely possible or contingent (not impossible and not necessary). If
the premises of a valid deductive argument are true, then the
Locke, Berkeley, Hume 115
conclusion must be true. A deductively valid argument is one in
which to assert the premises and deny the conclusion would be a
contradiction. Conclusions can be validly deduced from premises
independently of the evidence of experience; but if the conclusions
are factual, then such deductions must involve factual premises
which can be known to be true not by reason alone but only by the
evidence of experience; without the evidence of experience any
factual conclusion of a deduction is at best hypothetical and not yet
known to be true.
The spectre raised by empiricism is of two exclusive and together
exhaustive sets of truths: one set is necessary, certain and known a
priori, but says nothing about the actual nature of the world; the other
set is contingent, not certain and known, if at all, a posteriori, but can
say something about the actual nature of the world; this undermines
the search for necessary and certain knowledge about the actual nature
of the world by leaving all truths about the actual nature of the world
both contingent and not certain.
Locke
John Locke (1632–1704) was born in Wrington in Somerset and died
at Oates in Essex. Locke was far from being the caricature of the
philosophical recluse; he was, on the contrary, a man well known in
public affairs, sometimes involving considerable danger; but, despite
his close involvement with controversial political affairs, Locke was
a prudent man. Locke’s father was a lawyer and a staunch Puritan
and Parliamentarian who fought with the Parliamentarian army in
the English Civil War; this began in 1642 against Charles I, who was
beheaded in 1649. Locke attended Westminster School, and in 1652
he went to Christ Church, Oxford. At Oxford he studied the arts
course of logic, grammar, rhetoric, Greek and moral philosophy.
After obtaining his BA he was elected in 1658 to a Senior Studentship
at Christ Church which was tenable for life. He taught Greek and
moral philosophy, but soon became interested in medicine, and
attained the BM (Bachelor of Medicine) degree from the University
of Oxford in 1674.
It was during his time at Oxford that Locke became dissatisfied with
the philosophy of scholasticism and first became acquainted with, and
derived inspiration from, the works of Descartes. Locke was elected a
Fellow of the Royal Society in 1668; there he came to know the chemist
and physicist Robert Boyle (1627–92), whose emphasis on
experimental method and the corpuscular theory of the constitution of
matter impressed and decisively affected Locke: it influenced his
philosophical thought, particularly in its rejection of Aristotelian
modes of physical explanation. Sympathetic to Locke’s views is the
116 Empiricism
motto of the Royal Society, Nullius in verba: “Nothing by mere
authority”. Locke’s thought, both in its purely philosophical as well as
in its political interests, is consistently marked by the advocation of
tolerance and resistance to dogmatism in the face of the limits and
uncertainties of human knowledge. His political thought, as embodied
in the Two treatises on government (1690), became a philosophical
foundation of liberal democracy.
After Cromwell’s Commonwealth, the monarchy was restored in
1660 under Charles II. Through his interest in medicine, Locke had
initially become in 1667 a medical adviser to Lord Ashley, later the Earl
of Shaftesbury. Locke in fact left his college, never to teach there again,
and instead entered into a series of official appointments. Between
1675 and 1679, Locke spent time in France mainly for the sake of his
poor health. His travels in Europe fostered his keen interest in all
aspects of contemporary scientific work. This association and
friendship with Shaftesbury was to bring Locke problems; Shaftesbury
was party to the failed attempt to overthrow and replace Charles II
with Charles’s illegitimate offspring, the Protestant Duke of
Monmouth. Shaftesbury, fearing impeachment for treason, fled to
Holland in 1682, and died the next year; Locke also wisely, because of
his support of Shaftesbury and Monmouth, moved to, and for a time
hid in, Holland under a false name, until returning to England after
the Glorious Revolution of 1688, when the Catholic Stuart King, James
II, fled the English throne, to be replaced by the Protestant William of
Orange, which led to the Hanoverian Succession. From 1691 Locke
lived at Oates in Essex in the house of Sir Francis and Lady Masham
until his death in 1704.
It is important to understand the overall aim of Locke’s philosophy:
it is concerned mainly with determining the nature, scope, and limits
of knowledge and with giving an account of the nature of reality.
Locke’s position stands in contrast to that of many of his philosophical
predecessors and, indeed, some of his philosophical successors. The
heart of the matter lies in the interplay between scepticism and the
scope of human knowledge; and it can be summed up by the aim of
discovering what it is human beings are and are not fitted to know.
Locke accepts that knowledge, properly speaking, is of truths which
are certain and universal. Our inability to refute scepticism in various
areas of human inquiry where we wish to claim to know truths might
lead us to the despairing view that only scepticism can remain in those
areas. Locke emphasizes the limits of human knowledge proper, but in
a way that allows for areas where, although we do not have
knowledge in the strict sense, we are not thereby forced into scepticism
because in many of these areas of inquiry we are still capable of
probable belief; and, indeed, the belief is sometimes so probable that it
is virtually as good as knowledge. What Locke is advocating might be
called degrees of appropriate certainty. This presents us with
Locke 117
something other than a choice between strict knowledge and total
ignorance. In those areas where we cannot strictly speaking know,
Locke argues that we should acknowledge that we have reached our
limitations; but knowledge in the strict sense is usually not required;
the probable belief we may have instead is sufficient for our purposes,
and this, although not a refutation, is the answer to the sceptic. Locke
advocates the view that absolute certainty in many important areas of
human inquiry is not possible for us but nor is it required or even
appropriate; an example is our degree of certainty about the existence
of an external world.
Locke’s strategy in delimiting human knowledge is to examine the
power of the human mind and the objects of thought: ideas. The
philosophy of Locke stands on two main foundations: first, that all
knowledge derives from reasoning about our ideas and, secondly, that
all ideas originate in experience. We cannot in our thinking and
knowledge go beyond the ideas or concepts we actually have—ideas
are the materials of thought and knowledge—and the ideas we have
are bounded by what ideas can be attained through experience.
From this it is not surprising to find that Locke opposes what he
regards as a prevalent notion that we have innate, or inborn, ideas in
the mind independently of experience. It soon becomes clear that what
Locke is most concerned to oppose is the existence in the mind of
innate principles and knowledge; although in denying the existence of
innate ideas—ideas being the building blocks of knowledge—Locke is
also denying innate knowledge of truths. One of the chief motives for
Locke’s denial of innate knowledge is that the identification of a
principle as innate or inborn is sometimes used, especially in moral
matters, as a block to any questioning of the truth of that principle. But
we must, Locke says, think through what we claim to know, and make
knowledge our own. This goes along with Locke’s general suspicion of
authority as a valid ground for accepting something as true.
Apart from certain moral principles alleged to be innate, there were
also said by advocates of innate ideas to be innate basic logical
principles, such as “Whatever is is”. One of the arguments used in
favour of innate principles is that there are some principles that are
universally assented to as true, and this shows them to be innate rather
than acquired. Locke flatly denies that there is such universal assent;
children and idiots just do not assent to abstract principles; but he goes
on to say that even if universal assent were a fact, this would not show
that the universal assent could not be explained in some other way
than by saying that what is assented to is innate. In fact Locke thinks
the argument from immediate universal assent to the conclusion that
particular principles are innate confuses innateness and cases of selfevidence;
the universal assent, on encountering a self-evident
proposition, is fully accounted for by the relation of the terms in the
proposition, meaning we cannot think otherwise if we understand it at
118 Empiricism
all. Locke rejects the idea that there might be innate principles
implicitly in the mind which are not explicitly understood. Moral
rules, which are supposedly innate, are not even self-evident and they
therefore demand reasons to be given for their acceptance. Moreover,
the abstracted ideas or concept terms of abstract principles suggested
as innate can be acquired only after experience of the particular cases
and the gaining of particular ideas.
Locke does not deny the existence of innate capacities—the power to
perceive, believe, recognize truth and falsity, judge, assent to
principles—but none of these capacities actually amounts in itself to
possessing innate ideas, principles or knowledge of truths. If
innateness merely amounted to the capacity to recognize and assent to
truths when presented, then all knowledge, since it involves this,
would be innate—which Locke thinks is absurd.
Locke never questions whether even if there were innate principles
this would make any difference to whether those principles were true;
he never questions the truth of putative innate principles. The reason
for this is Locke’s piety; if there were innate principles they would
have to be true because they could be implanted directly in us only by
God. Locke argues that there are, in fact, no innate ideas and
principles, so the question of their truth or falsity does not arise, and
the positing of them is unnecessary to explain the knowledge we have.
The explanation for all the ideas we have is that they originate in
experience: experience is made up of sensation derived from external
material objects, and reflection derived from awareness of the workings
of our own mind. Examples of ideas of sense-experience are yellow,
elephant, cold, army; examples of ideas of reflective-experience are
thinking, believing, willing, doubting.
Locke is not free from the charge of confusing psychological or
genetic empiricism with philosophical or logical empiricism. Genetic
empiricism is a psychological theory accounting for the way we
actually come to have, or acquire, ideas and knowledge of which
propositions are true and which false; philosophical empiricism is
concerned only with that on which the truth or falsity of propositions
depends and what is logically required in order to justify the claim to
know whether the propositions are true or false. This makes the
distinction between knowledge of truths being psychologically innate
and its being logically a priori. Showing that a certain proposition is,
psychologically speaking, entertained in the mind at a time prior to
any experience would not show whether that proposition were true or
false or have any relevance to justifying logically a claim to know it to
be true or false. Whether a proposition can be known to be true or false
logically independently of experience is not shown by discovering
whether it was in the mind innately or not, but by deciding of what
logical type the proposition is.
Take the following two propositions:
Locke 119
(a) The internal angles of a plane triangle add up to 180 degrees.
(b) There are lions in Africa.
If the truth of (a), which is a necessary truth, were questioned, we
would prove it to be true by showing it is deducible from the axioms of
Euclidean geometry; if (b), which is a contingent truth, were
questioned we could only establish its truth by going to Africa and
looking. The truth of (a), and knowledge of that truth, is, logically
speaking, independent of evidence of experience, whereas the truth of
(b), and knowledge of that truth, is, logically speaking, knowable only
through the evidence of experience. Whether a truth is knowable a
priori or a posteriori is determined by whether the truth can possibly be
established empirically or non-empirically; and this is different from
the truth being actually innate or acquired. We might possess no nonempirical
truths such as (a) innately; but that would not alter the fact
that these propositions are true regardless of any states of affairs in the
world, and they can be known to be true independently of experience
and by pure logical reasoning. We might possess a whole stack of what
turn out to be empirical truths such as (b) innately, and although this
might be psychologically surprising it would not alter the fact that the
truth of these propositions depends on certain states of affairs in the
world obtaining, and they can be known to obtain only through
experience, not by pure logical reasoning alone. A truth such as “Either
it is raining or it is not raining” (“p or not-p”) is an a priori truth
because it is true independently of any states of affairs in the world,
and it logically can be known to be true independently of inspecting
the weather; but it tells us nothing about the weather; it does not help
us to decide if we should take an umbrella. All propositional beliefs,
even if true, which are not logically a priori can be known to be true
only by checking them against the evidence of experience, regardless
of whether we have the beliefs innately or not. Those truths known
independently of experience are said to be necessary in that their
denial implies a contradiction; those truths known only by experience
are said to be contingent, as their denial does not imply a
contradiction. The philosophical concern should be to distinguish
between a priori propositions, which are all those propositions where
the logical justification of knowledge of whether they are true or false
is independent of empirical evidence, and a posteriori propositions,
which are all those propositions where the logical justification of
knowledge of whether they are true or false is dependent on empirical
evidence.
There is considerable uncertainty and controversy over what Locke
means by “idea”. Locke defines an idea as “whatsoever is the object of
the understanding when a man thinks”. Some have taken Locke to
mean by “idea” some kind of mental entity—mental images which are
objects in the mind. The consequence of this (a point raised by
120 Empiricism
Berkeley) is that it immediately leads to scepticism about knowledge
through perception of the external world. If the “veil of perception” or
“picture-original” view is correct, and we only ever perceive ideas in
the mind, then there can be no way of checking if the ideas represent
the external world truthfully, or even if there is an external world at all
corresponding to the ideas. We are locked in a circle of ideas, with the
knower logically blocked off from what is known; our ideas are a
barrier between us and what the ideas are ideas of. Partly because of
this point, which seems too obvious for Locke to have missed, and
which he even seems to point out, other interpreters of his work
suggest an alternative view in which the reification of ideas is resisted.
Locke, it is said, means by “idea” in “idea of X” a mental or perceptual
act, not a thing; “idea” refers to our understanding of X, or our
perception of X, as distinct from X itself; “idea of X” means “X-as-it-isperceived/understood/known/appears”;
and it expresses the
epistemological relation between the knower and the thing known. To
avoid a regress, what must ultimately be caused in the perceptual
process is an act of perceptual awareness itself, not another object of
which to be aware. An “idea of X” involves two entities, knower and
object known, not three by including an entity “idea of X”. The
expression “idea of” points out that our conception or perception of an
object is our conception or perception; it is how it appears to us, as
opposed to how the object is in itself, which may differ from our idea.
This emphasizes the assertion that we inevitably view things under the
constraint or qualification of their being seen from our point of view—
how things appear to us—and that we cannot attain the God’s-eye
view of knowing objects as they are in themselves quite independently
of all reference to its being our perspective. To say I have an idea of X
is just to say I have some understanding of the object X. On this
interpretation, when Locke speaks in a variety of ways about a relation
of resemblance or non-resemblance between ideas and what they are
ideas of, he is not committing himself to this being like the relation
between a picture or image and an original—literal picturing—but
rather the kind of relation that holds between an accurate and
inaccurate description and the object described.
Locke divides ideas into simple and complex. Complex ideas are
compounds of simple ideas. We may experience ideas in complexes, or
even only in complexes, but they must be reducible to simple atomic
unanalyzable ideas. The thinking behind this is that at some point
there are ideas which cannot be broken down into anything simpler
and to have the ideas at all one must derive them directly from
experience. If one has never experienced the simple idea of red, there is
no way that having the idea can be explained by showing how it is
compounded of simpler ideas one has experienced; whereas the idea
of a mermaid, even if one has not encountered mermaids in one’s
experience, is made up of ideas one has encountered in experience.
Locke 121
Locke is not saying that we always experience simple ideas first, and
then build up compounds, merely that all compounds must be
analyzable into simple ideas of which we have had direct experience.
Locke’s position places restrictions on the scope of imagination:
whatever we make up we will only ever be compounding simple ideas
that ultimately originate in experience.
For Locke the meaning of a word derives from its standing for, and
its association with, an idea or complex of ideas; we know the meaning
of a word when we know the idea it stands for. If someone has not
experienced the simple idea X, then he will not understand the
meaning of the word standing for X. We will, in attempting to speak
about that which is, strictly speaking, beyond our experience and is in
no way analogous to anything in our experience, be using meaningless
expressions and talking nonsense because we will be unable to specify
any idea for which the word stands.
If it is the case that we only ever encounter particulars in our
experience from whence we derive particular ideas, the problem arises
as to how we come by abstract general ideas, for which general words
stand as signs—such as “redness”, “man”, “nurse”—which can apply
equally to many particulars. Such general terms are necessary for
communication and knowledge. Pure nominalism holds that all that
any group of particulars under a general name have literally in
common is the sharing of that name; but this leaves unanswered the
problem of universals: namely how we know which particulars come
under that general name in the first place. Locke has more than one
answer. His first answer is that we are blessed with a faculty of
abstraction: by a process of omission the abstract general idea is
formed by leaving out of each idea of particular members of a similar
class all those characteristics in which they differ, thereby including
only that which is common. The general idea will itself be a particular;
but it is not clear what the resultant idea amounts to. Berkeley argues
that Locke’s procedure is impossible: if we take away all particular
features we are left with an impossible idea; we could not represent to
ourselves a red which is no particular shade of red at all; there cannot
be an idea which is merely determinable. Locke’s second answer is that
the meaning of abstract general ideas and words is fixed by “nominal
essences”: we notice similar characteristics between particulars, and
we decide on some set of defining objective particular characteristics
by which we then have the ability to recognize whether any particular
is correctly admitted to a specific general class.
Locke explains the relation between our ideas in the mind of
sensible qualities of external objects and those sensible qualities as
they exist in external objects themselves by making a distinction
between primary qualities and secondary qualities,
(a) Primary qualities: our ideas of primary qualities resemble those
122 Empiricism
qualities as they are in bodies. Primary qualities are size,
extension, shape, movement, solidity.
(b) Secondary qualities: our ideas of secondary qualities do not
resemble those qualities as they are in bodies. Secondary qualities
are hot, cold, sound, colour, taste, odour, etc.
Locke was greatly influenced by the atomic theory of matter
propounded by Boyle; the basic stuff of the natural world consists of
material objects which are made up of an insensible structure or
configuration of atoms or corpuscles which themselves have no
internal structure; these microscopic atoms have only primary
qualities. Locke thinks, however, that the soul is immaterial, although
he does not think it impossible that God could have made thought an
attribute of matter. Macroscopic material objects we perceive appear to
have both primary and secondary qualities, but both qualities at the
macroscopic level depend on configurations of insensible particles
which themselves have only primary qualities. The secondary qualities
we perceive are not in objects as-we-perceive-secondary-qualities to be;
this does not mean the secondary qualities are nothing in objects;
rather, the secondary qualities are in objects some determinate fine
corpuscular structure; our ideas of secondary qualities are a result of
the power of qualities as they exist in objects, as insensible corpuscles,
which produce certain sensations in us. The ideas of secondary
qualities are an effect on us of those qualities in objects as insensible
corpuscles with only primary qualities. The ideas caused in us of
secondary qualities never resemble that which in objects causes us to
have those ideas, but are in objects nothing but a certain configuration
of corpuscles.
Take, for example, the secondary quality red: it is true to say that
object X is red if what is meant is that X has a corpuscular structure
such that under normal conditions it has the power to produce in us
the idea or sensation of red, and thus the object X is seen as red; but it
is false to say that object X is red if what is meant is that red exists in X
in the same way as I have the idea or sensation of red. Locke also
distinguishes a third quality which he simply calls “powers”, which is
the capacity of bodies to cause changes in other bodies such that they
then appear different to us, as when the sun melts wax.
Another way of explaining the distinction between primary and
secondary qualities is through the notion of resemblance and accurate
descriptions. Our ideas of primary qualities can resemble (can be
accurate or correct representations/descriptions of) those qualities as
they are in objects. Our ideas of secondary qualities never resemble
(cannot be accurate or correct representations/descriptions of) those
qualities as they are in objects. This is not to say we cannot be mistaken
about what determinate primary quality an object has; but we can be
right in the sense that the quality exists in the object as the same kind
Locke 123
as that which is perceived. We might misperceive the determinate
shape of X as triangular when it is square, but we are not mistaken that
it really has some determinable shape or other; in this sense our ideas
of primary qualities resemble the qualities as they are in objects. We
will always be wrong about the object having any secondary qualities
if we mean that the secondary qualities ever exist in the object in the
same way as we perceive qualities; secondary qualities do not exist in
bodies in the same way as we perceive them at all. This does not mean
we are incorrect to describe X as red if we mean by this that it has that
determinate corpuscular structure which causes one to have the idea of
red under specific conditions.
God has chosen to connect specific corpuscular configurations in
bodies with the power to produce the specific sensations or ideas we
experience; why a certain corpuscular configuration should produce
just those experiences within us is something Locke regards as
mysterious.
Recent thinking suggests that Locke was not making the distinction
between primary and secondary qualities, but was accepting the
distinction, which he took over from the scientific work of Boyle.
Berkeley objects to Locke’s apparent argument for the distinction that
the primary qualities are invariant and secondary qualities variant
with respect to observers: primary qualities are just as variant with the
changing perspective of the observer as secondary ones. But if Locke
did not try by argument to justify making a distinction between
primary and secondary qualities, then Berkeley’s counter argument is
beside the point. Locke’s chief point in accepting the corpuscular
hypothesis is that it provides an economical unifying explanation of a
great variety of phenomena; and whereas we can conceive of an
explanation of changes in secondary qualities in terms of changes in
primary qualities, the reverse seems inconceivable.
How is Locke entitled to have an idea of, and talk meaningfully
about, the insensible configuration of particles which are too small for
us to experience them, given his empiricism about the origin of all
ideas? Locke’s answer is that our inability to experience such particles
is purely contingent, and did we but have microscopical eyes, we
would see them. Moreover, corpuscular explanations involve
insensible particles which are entities which have properties of the
same kind as, or are analogous to, the properties of macroscopic things
we do experience, namely, primary qualities. We speak intelligibly in
referring to the particles because we have ideas of the kind of
properties they have and therefore understand what we mean by the
words describing them.
Locke’s account of substance, the most fundamental independent
stuff in the world, is subject to different interpretations. On one view
Locke notoriously means by substance “naked substance”, a
something I know not what: a “something”, or substratum in general,
124 Empiricism
beside all the qualities we predicate of objects which “support” all
those qualities. We have ideas of things having various qualities, and
since we suppose that these qualities cannot subsist by themselves, we
suppose there to be a something which they are the qualities of, and
that that something is something beside the qualities themselves. But if
a substratum is imagined to be that which is stripped of all qualities,
one is left not with a special, if mysterious, something, but with an
ineffable nothing. Thus the reason that this substance is not known is
that a propertyless substance is logically or necessarily unknowable.
Other interpretations have suggested that Locke’s remarks
concerning pure substance in general—substratum—are ironic. The
suggestion is that Locke rejects the confused notion of a pure substance
in general and aims to replace such talk with positive talk of
something else, while also wishing to explain how we are led to think
of it as underlying aggregates of sensible qualities. He thinks we are
led to belief in substance through: (a) the grammar of subject-predicate
talk; (b) seeking something to explain the cause of the union of
apparently unrelated aggregates of different sorts of qualities; (c) our
notion that qualities—for example, the yellow, malleable, heavy
qualities of gold—cannot exist separately from something in which the
qualities can exist. Locke’s reinterpretation of substance originates in
substance as the sought-after cause explaining why some particular
substance such as gold should always have the qualities of being
yellow, malleable, and heavy, when there seems to be no connection
between the qualities. The explanation for the connection or union of
these apparently unconnected qualities in all instances of a particular
kind of substance in fact lies in the common real determinate internal
corpuscular structure.
Locke describes the nominal essence of a thing as simply the
qualities or properties we decide to gather under a sortal name, such
as “gold”, for the purpose of classifying particulars into kinds. The
nominal essence gives us a criterion for identification. Although there
are natural constraints on us, the sorting of things into kinds in this
way is created and linguistic.
Locke talks of real essence in two senses. First, the traditional
scholastic sense of real essence as a thing’s substantial form which
makes a thing the kind of thing it is; Locke rejects this as obscure and
having no explanatory use; to explain the properties of gold by saying
that it has those properties because it possesses the substantial form of
gold is just to say gold has the properties of gold; talk of substantial
forms stops us seeking the underlying causes. Talk of underlying
causes refers us to Locke’s second sense of real essence; that is,
Lockean real essence which is the real determinate internal
corpuscular constitution on which the apparent properties depend.
We cannot strictly know the inner atomic structure of things because
our senses are not fitted to perceive them; nevertheless, the notion,
Locke 125
unlike substantial forms, is an intelligible hypothesis which has
genuine explanatory power. Moreover, our lack of knowledge of the
inner atomic constitution of things is, unlike the lack of knowledge of
“pure substance in general”, merely a contingent matter.
Locke defines knowledge as “nothing but the perception of the
connection and agreement, or disagreement and repugnancy, of any of
our ideas”. Propositions are true when the ideas constituting the
propositions are connected in such a way as to make them true; we can
know propositions to be true in so far as we can “perceive” this
connection. Add to this the condition that knowledge must be of truths
that are certain and universal, and what we can be said strictly to
know turns out to be extremely meagre. But we are not left only with
doubt where we cannot have knowledge since we can also have
probable belief. Locke’s overall aim is to commend to us the view that
the lover of truth should not hold a proposition more firmly than the
proof or evidence for it warrants. Locke lists four sorts of agreement
and disagreement of ideas.
(1) Identity or diversity
Here he seems to have in mind logical identity and contradiction.
(2) Relation
Here he is referring to demonstrative logical or mathematical
relations.
(3) Coexistence or necessary connection
Here is meant connection of ideas which reflect the manner of
connection of properties of things occurring together in nature.
(4) Real existence
Here he means what really exists in the world.
Our limited ability actually to perceive the appropriate connection
between ideas in a large range of cases immediately restricts what we
can know, strictly speaking. There is no problem in claiming to know
as true propositions whose ideas can be immediately perceived as
connected or disagreeing, such as “blue is blue” or “blue is not
yellow”; these are intuitive truths. Such truths Locke refers to as
trifling. Locke more dubiously claims some moral truths can be known
intuitively. There is also little difficulty in making a plausible case for
our knowing truths which result from logical deductive reasoning,
such as geometric and mathematical truths, which can be thought of as
made up of intuitive steps or connected chains of intuitive truths
which form the process of demonstrative reasoning. After this
difficulties arise.
Locke himself admits that to have an idea is one thing, but it does
not follow, when not actually receiving the idea, that anything exists
corresponding to that idea. The problem is the lack of any connection
to be perceived between our having an idea and the real existence of
that which the idea is an idea of. A possible exception is the existence
126 Empiricism
of God and that the idea of God entails that God is—which amounts to
a compressed ontological argument. Locke equivocates on what we
can be said to know exists. He thinks that we have intuitive knowledge
of our own existence; he thinks that we can have demonstrative
knowledge of God’s existence; and he thinks that, while we are
actually perceiving objects, we have a belief of such great assurance
and certainty that those objects exist without us that it “deserves the
name knowledge”. He is clear, however, that strictly speaking we
cannot know the truths expressing actual factual connections between
the properties we experience objects to have or know the scientific
hypotheses with which we describe their behaviour (for example the
connection of the idea of “gold” and “soluble in aqua regia” in the
proposition “Gold is soluble in aqua regia”); we cannot know truths in
these cases because we cannot perceive any intrinsic connection
between the constituent ideas reflecting those properties such that it
would make them true; we cannot perceive any necessary connection
between the ideas; all that we perceive is the juxtaposition or
conjunction of the ideas. So in the case of natural science we are not
capable of knowledge, but we can believe with some degree of
probability in the truth of scientific propositions, and the probability of
truth will increase in proportion as it conforms to my past experience
and that of others.
Locke’s view suggests a hierarchy of certainty, here given in
descending order of certainty:
(A) intuition
(B) deductions or demonstrations
(C) sensitive knowledge
(D) natural science.
(A) and (B) strictly constitute areas of knowledge; (C) is knowledge of
the existence of particular objects in the external world as we actually
perceive them, although it is not so certain as (A) and (B); but with (D)
we have only probable belief. Knowledge of our own existence is
included in (A), and that of the existence of God in (B). With these
exceptions, Locke is in danger of leaving us with knowledge almost
entirely of propositions which are hypothetical non-existential (stating
what follows if we accept certain propositions, regardless of whether
those initial propositions are actually true) and verifiable a priori, and
little knowledge of propositions which are categorical existential
(asserting the actual existence and nature of things) and verifiable a
posteriori. There is certainly a problem in claiming to know the general
or universal existential propositions and the existence of objects not
actually perceived which are required for natural science. In short,
knowledge is restricted to necessary certain truths, in which case
knowledge is limited to logical relations and excludes relations of fact
which are neither necessarily nor certainly true.
Locke 127
Locke is Cartesian, or at least rationalist, in giving a necessitarian
account of reality: knowledge of reality would ideally be one of
revealing the natural necessity and connection of things. He differs
from the rationalists in his scepticism over whether natural necessities
can actually be known; but he also differs from empiricists in holding
that there nevertheless are natural necessities—necessities between
matters of fact about the world—which could be known. Thus he does
not fit the traditional empiricist mould for two important reasons:
(a) Locke does not share the empiricist view that all knowledge
which we can know independently of experience by reason
alone is thereby trivial and unable to tell us anything about the
actual nature of reality. Mathematical and geometrical truths are
cases of non-trivial a priori knowledge in which we discover
new truths.
(b) Locke believes, unlike Hume, in natural or metaphysical necessity.
The epistemological problem that we cannot know natural
connections to be necessary and with certainty does not show the
necessary connections are not there. Locke says in addition that
our inability to perceive the connections as necessary is a purely
contingent matter which depends merely on our inability to
perceive the inner microscopic corpuscular structure of material
objects; could we see this structure, we would perceive that the
connection between the qualities objects have is necessary. If we
could see the microscopic structure, we would see that the
sensible qualities or properties of X which depend on that
microscopic structure must occur together necessarily.
Locke does not see the problem Hume uncovers, that no matter how
acute our senses we would only ever perceive one idea A in
conjunction with, or followed by, another idea B, but would never
perceive between them a necessary connection such that B must be in
conjunction with, or must follow, A, and things cannot be otherwise. If
the connection were necessary, then the assertion of (A and not-B)
would be a logical contradiction, but it never is when describing actual
matters of fact. There is no analogous connection between natural
matters of fact for necessary deductive connections or logical relations.
It is never a logical contradiction to suppose that A occurs, but B does
not follow, or that property A is not found with property B, no matter
how many times the conjunction of A and B has been observed.
Necessity based on logical contradiction is the only sort possible. The
universal generalization “All A is B” and the necessary causal
connection “If A occurs, then B must occur”, where A and B describe
matters of fact, cannot be known to hold, or the beliefs rationally
justified, through the evidence of experience or by deductive
reasoning; thus they cannot be known or rationally justified at all; this
is the logical problem of induction and causation.
128 Empiricism
Berkeley
George Berkeley (1685–1753) was born near Kilkenny, Ireland. At the
age of fifteen he entered Trinity College, Dublin, and graduated with
his BA in 1704 at the age of nineteen; he became a fellow of the College
in 1707. The spur to his philosophical writing probably derived from
reading Locke, Newton (1642–1727), and Malebranche (1638–1715).
Berkeley’s New theory of vision appeared in 1709, with a fourth edition
in 1732. His major philosophical works, A treatise concerning the
principles of human knowledge (1710) and the Three dialogues between
Hylas and Philonous (1713), were both published by the time he was
twenty-eight. In 1724 he resigned from his fellowship to become Dean
of Derry. In 1728 Berkeley left, with his wife, for America in an attempt
to found a college in Bermuda to educate the native Indians and the
sons of local planters, but the money for the project failed to
materialize from the government in England. Thus in 1731 Berkeley
returned to England, and eventually to Ireland where he became
Bishop of Cloyne in 1734. In 1752 he moved to Oxford, and died there
suddenly in 1753 at the age of sixty-eight.
It is perhaps more than usually necessary in understanding the
philosophy of Berkeley to place it in its intellectual context; otherwise
Berkeley’s philosophy can seem too obviously false to require serious
examination; his philosophy has been called immaterialism or
idealism, although the two terms are not exactly equivalent.
Berkeley exemplifies one way of stringently applying empiricism:
he conjoins the view that all we can ever know is our immediate ideas
with the view that words and other expressions in our language derive
their meaning only from association with specific ideas; this leads to
the ontological doctrine that only ideas subsisting in minds and minds
themselves can be said to exist because to talk of things existing in any
other way is meaningless as the expressions used in the talk are
necessarily unconnected to any ideas. Expressions not translatable
into, or associated with, some experience are meaningless.
The essential background to the understanding of Berkeley’s
philosophy is formed by a combination of the new scientific
materialism and the representative theory of perception. Scientific
materialism, mainly derived from Newton, proposes a mechanistic
conception of the universe which functions like the works of a giant
clock and a corpuscular hypothesis as to the constitution of matter. The
representative theory of perception, mainly derived from an
interpretation of Locke, is here the thesis that the immediate objects of
perception are always ideas. There are also connected problems arising
from Descartes and Malebranche concerning the relation between the
incorporeal mind and the corporeal body. Berkeley saw the scepticism
that could arise from these beliefs as a scandalous affront to common
sense and a threat to religious belief; but all the forms of scepticism,
Berkeley 129
Berkeley thinks, can be eliminated at one blow by rejecting their
common assumptions.
The scepticism to which the materialistic philosophy gave rise took
three main forms:
(a) the existence of sensible things
(b) the nature of sensible things
(c) the existence and nature of God.
The main additional sceptical problem posed by Cartesianism is:
(d) how matter and spirit can interact.
Materialism gives rise to all the first three forms of scepticism when
combined with the doctrine that we only ever perceive immediately
ideas in our minds by opening an unbridgeable gap between how things
appear to us and how they really are in themselves: a gap between our
ideas and what our ideas are ideas of. Material objects, specifically their
corpuscular structures, are seen as the cause of our ideas; but material
objects do not have, in the same way that we perceive them, all the
qualities that they appear to have. The gap between the ideas that we
immediately perceive and their supposed causes, which we do not
directly but only ever mediately perceive by way of ideas, opens up the
possibility of an insoluble scepticism concerning our knowledge of the
nature and even existence of the objects of the external world. We can
never gain immediate access to the something, whatever it is, that is the
cause of our ideas to check whether the ideas which supposedly
represent the nature of that something are accurate, or even whether the
supposed something exists at all; we can never perceive the something
that is the supposed cause of our ideas immediately but only mediately
in virtue of perceiving immediately intermediate mental objects: ideas.
Materialism also leads to atheism according to Berkeley, since the
posited material substance is to a high degree, or perhaps completely,
independent of God in its operations and existence. Many materialists
supposed that God was ultimately still required as the creator and first
mover of the universe; but if we suppose that the universe has existed
for ever, then God’s existence again becomes dispensable. The existence
of God is still possible, but His existence is not logically required, nor
even obviously important.
An additional, but connected, source of scepticism derives from
Descartes and Malebranche. In the Cartesian view there are two
distinct substances, mind and matter, whose essential attributes are
thought and extension respectively. The problem then arises as to how
their interaction is to be made intelligible: how can the non-extended
mind cause changes in motion of extended bodily parts, such as the
brain, and how can motions of the extended bodily parts cause
changes in non-spatial mental substance which produces thoughts?
This problem led Malebranche to the doctrine of occasionalism: this
130 Empiricism
holds that although mind and body do not interact, God on the
appropriate occasions systematically intervenes to produce the same
result as if they did interact; on the occasion of my willing the
movement of my body God causes the correct bodily movement; on
the occasion of my observing a physical object God causes me to have
the appropriate perception by sharing in His ideas.
Berkeley thinks that materialism is:
(i) Unjustified
The arguments presented for the adoption of materialism are
insufficient,
(ii) Unnecessary
The thesis is extravagant since it posits the existence of material
entities that are not required to give an explanation of the course
of our experiences,
(iii) False and must be false
Matter is not, and indeed cannot be, the cause of our experiences,
(iv) Meaningless
It requires us to give meaning to the term “matter” or “material
substance” which is something we never directly experience,
which is the cause of our ideas; but as the meaning of a term is
the idea for which it stands, and there can be no idea of that
which we cannot experience, then all terms referring to entities
such as material substance, which are beyond experience, must be
meaningless,
(v) Contradictory
It requires that ideas may exist when not perceived by us in an
unthinking corporeal substance or matter.
In several important ways Berkeley is a very strict empiricist. Generally
he holds that the limits of what it is intelligible or meaningful to talk
about must refer to something in the content of our experience. If we
are making some distinction in the world, it must, to be genuine, refer
to some perceivable difference; if a proposition is intelligible, it must
refer to something perceivable. It is surely part of the persuasiveness,
even attractiveness, of Berkeley’s idealism that it asks us to concentrate
only on the actual character of the content of our own minds.
Berkeley’s overall strategy in opposing all the forms ((a), (b), (c), (d),
above) of scepticism derives from closing the gap between our ideas
and what our ideas are ideas of; thus preventing the sceptic from
driving a wedge between the two. Berkeley advocates negatively
immaterialism and positively idealism; he also assumes that if
materialism can be shown to be false, then his form of idealism must
be true in virtue of its being the only alternative to that materialism.
Talk of material objects, in Berkeley’s philosophy, is not a reference to
some material substance which can exist unperceived as the supposed
cause of our ideas but which, since the objects of perception are always
Berkeley 131
ideas, we never actually perceive. To talk of material or sensible objects is
to talk about actual or possible objects of perception, and that is to talk of
ideas or bundles/collections of ideas themselves which must, as ideas,
exist in a mind or spiritual substance. To talk of material objects or
sensible things is not to refer to something other than the ideas we
perceive, it is to talk of those ideas themselves; what we mean by
material objects is just certain ideas or sets of ideas. Any reference to the
nature or character of the world is a reference to, and is only intelligible
as a reference to, actual or possible experiences. What we immediately
perceive in vision is a flat, two-dimensional array of colours and shapes.
In the New theory of vision Berkeley presents arguments to show that
distance is not something immediately perceived but something
constructed from certain orderly relations of the ideas of different senses
in the mind. Thus to say an object is one mile away is just to say that a
certain sequence of ideas—for example, those constituting the experience
of walking forward—would have to go through the mind before we
received such-and-such ideas of touch. This lays the groundwork for the
view that what is perceived (the object of perception), because it is in no
case an immediate perception of something at a distance from us, is
therefore always something in the mind.
The equating of ideas with sensible things, which thereby makes
sensible things mind-dependent, eliminates each of the previously
mentioned forms of scepticism ((a), (b), (c), (d), above) produced by
materialism and Cartesianism in the following way.
(a) The existence of sensible things. This problem is eliminated
because the sceptic cannot drive a wedge between ideas and
things; if the objects of sense are ideas, and we cannot doubt that
we have ideas and thus ideas exist, we cannot doubt the existence
of the objects of sense or sensible things.
(b) The nature of sensible things. This is just the sum of a thing’s
sensible qualities. In addition science no longer purports to reveal
the essential nature of things in the external world whereby it can
establish the necessary connections required for true causal
relations between the sensible properties of things we can
perceive; rather, it aspires only to map the regular correlations
between ideas, that is, between phenomena.
(c) The existence and nature of God. This problem is eliminated by
making God metaphysically indispensable: once material
substance is eliminated, it is necessary to affirm that God exists as
the immediate real cause of those ideas that are not caused by our
imaginations and as the sustainer of those ideas we do not
actually perceive; thus God’s existence is manifest at all times as
the immediate cause of the vast majority of that which we
experience; the supposition that God does not exist is refuted by
almost every experience we have.
132 Empiricism
(d) How matter and spirit can interact. This problem is eliminated by
denying the existence of material substance; then the problem of
interaction between spirit and matter simply does not arise.
Berkeley also rejects occasionalism, arguing that we cause those
ideas which constitute what we can legitimately will, such as
moving our legs.
Berkeley presents various arguments opposing materialism.
(1) Berkeley thinks that the conception of matter as really having only
primary qualities, such as extension, shape, solidity, movement, is
an impossible one; he questions whether it is possible for us to
conceive of a shape which is no colour whatsoever; the conception
of matter required for materialism is impossible, for it involves
matter devoid of all secondary qualities, which are types of
qualities which it could not lack, and from which primary qualities
cannot be separated.
(2) Berkeley argues that it is a logical contradiction to talk of
conceiving of a thing which exists unconceived, for to conceive of
the possibility of something existing unconceived is necessarily to
conceive of that thing. But this argument, although tempting, is
fallacious. It is true that it is not possible for A to be conceived of,
and at the same time both exist and be a thing unconceived; but
that does not mean that at some other time A could not exist as an
unconceived-of-A; thus there is nothing contradictory in A
existing unthought about.
(3) Berkeley turns Locke’s argument concerning the relativity of
perceptions against Locke’s materialism. Berkeley takes Locke to
be arguing for the distinction between primary qualities (shape,
size, motion, solidity) and secondary qualities (colour, taste, heat,
sound, etc.) on the basis that those qualities, not really in objects
as we perceive them to be, are those that vary with the
disposition of the perceiver; such qualities are, as they are
perceived, subjective or in the mind (Locke does not in fact argue
that secondary qualities are therefore merely subjective) and result
from the effect of the insensible particles on us. But Berkeley
points out that if this argument proves that secondary qualities
are ideas in the mind, the same argument proves that primary
qualities are also only ideas in the mind, for these too vary with
the observer. In fact, there is no reason to suppose that in either
case we have shown the qualities to be subjective, for there is no
reason to believe that for a kind of quality to be really in objects,
or be attributed as a real objective property of objects, it must be
invariant with all changes in the observer. Moreover, we would
actually expect the real properties of things to vary with the
observer; for example, size as we get closer to an object.
(4) This argument concerns pain and heat. When we approach a fire
Berkeley 133
closely the heat is felt as a pain in the mind; when we are at a
further distance from the fire the heat is felt merely as warmth.
We are not tempted to say that the heat felt as pain is in the fire;
so we should also say the same for the lesser degrees of heat felt
as warmth, that heat is an idea in the mind.
(5) In this Berkeley runs together the notion of matter with what
Locke has to say about substance in general. He attributes to
Locke an account of substance which he thinks unintelligible, and
then takes this to be Locke’s account of material substance or
matter, so that is also unintelligible. Locke’s discussion of
substance in general seems to suggest that it is characterized by
being the “support” of all qualities; the qualities cannot subsist
alone, so substance is that in which the qualities subsist. But if
substance is the support of all qualities whatsoever, then any
attempt to give it a positive characterization is impossible, since
to do so would be to attribute qualities to it; thus substance
becomes an unknowable qualityless “something”. While this
argument is perfectly flawless as an attack on a qualityless
substratum, it is wide of the mark as an attack on matter because
no materialist would suggest that matter is qualityless.
The general form of Berkeley’s positive argument for idealism is as
follows. Sensible things (ordinary objects) are those things perceived by
the senses, and those things perceived by the senses are ideas. It follows
that sensible things are ideas or collections of ideas. In addition, ideas
can exist only if perceived by minds. With this additional premise it
follows that sensible things cannot exist unperceived.
Repeatedly Berkeley asks how the supposed “material substance”
should be characterized: what qualities or properties does it have?
Indeed any concept, apart from that of mind, if it is to be given a
meaning at all, must be translated into talk about some possible or
actual experiences. Whatever is suggested as the nature of “material
substance”, he points out that, if we can make what we are talking of
intelligible at all, the quality referred to is something that we
experience; but what we experience immediately is ideas, and hence
the existence of the quality is as an idea in the mind; and if we refer to
something that we do not experience, then he does not understand
what we mean when we refer to it.
Berkeley makes a distinction between immediate and mediate
perception; respectively between the immediate sensations of the
various senses, which involve no inference and about which we cannot
be mistaken, and that which is suggested by these perceptions. The
proper objects of perception are strictly speaking only those things we
perceive immediately, and all else that we claim to perceive is a
construct or inference from immediate perceptions.
Thus Berkeley identifies the normal everyday objects or sensible
134 Empiricism
things that we talk about with ideas or bundles of ideas; but in making
things into ideas he thinks he can show that he has not made them any
less real. Berkeley’s idealism is opposed only to the philosopher’s
conception of material substance as that in which sensible qualities
that we perceive through the mediation of ideas subsist when we do
not perceive them. Berkeley concludes that the very meaning of saying
that sensible objects exist is that they are perceived—although at times
he suggests that an object’s existence consists in its being perceivable.
Berkeley moves from the commonsense belief that sensible things are
simply what we perceive, to idealism which holds that the existence or
being of sensible things consists in their being perceived or at least
perceivable. In the end Berkeley holds to the view that to be or exist as
a sensible object is to be actually perceived, and not to the
phenomenalist view that to be is to be perceived or perceivable—to be
perceivable is to exist as a mere permanent possibility of sensation.
Thus, in Berkeley, with respect to sensible things, esse est percipi: to be is
to be perceived. This is not the only meaning that can be given to
existence, however: minds or spiritual substance, which have ideas,
also exist. To exist is thus also to perceive: esse est percipere: to be is to
perceive or be a perceiver. So in full we can say esse est aut percipi aut
percipere: to be (exist) is either to be perceived or to perceive. Spirits are
not, like sensible things, constructed as phenomena out of perceived
collections of ideas; they are that substance in which ideas inhere.
This position might seem to suggest implausibly that when sensible
things are not perceived by us they cease to exist: that they would
come and go out of existence. This would be true if only our own or
only human minds did the perceiving. But Berkeley’s view is only that
to exist is to be perceived by some mind or other. This is part of God’s
place in Berkeley’s world, although strictly speaking God does not
perceive ideas since He lacks senses, He nevertheless sustains in
existence by His mind those ideas of sensible objects not actually
perceived by us. Ideas that do not subsist in our finite minds subsist in
the infinite, omnipresent, omnipotent mind of God. God is essential to
Berkeley’s system; and if the system is true God is indispensable to all
of us. God is required for two main reasons. First, God is required to
give the continuity of a sustained existence to sensible things
unperceived by us. Second, God is required as the cause of those ideas
we experience which are not caused by us. The only entities capable of
being real efficient causes are minds, which alone are active as they are
capable of willing; ideas themselves are inert and incapable of being
real efficient causes. Berkeley agrees with Locke that causality can only
be understood through the experience of willing, but goes further in
saying that the only intelligible cases of causation are those that
involve willing. We are to a limited extent capable of creating ideas
through the faculty of imagination, but most of the ideas we have are
not caused by us; they must therefore be caused by some other mind;
Berkeley 135
nothing but the infinite mind of God could account for the richness,
stability and orderliness of the ideas we perceive. God directly causes
us, without the unnecessary mediation of any material substance, to
have those ideas which we call ideas or perceptions of sensible things,
which are those ideas not caused by ourselves.
Berkeley maintains the distinction between perception of reality and
the imagination, and denies the suggestion that he has turned the
world into mere fancy. Initially the distinction is made by pointing to
those ideas that come before our mind that are not products of our will
and imagination; these ideas are ideas of reality and have some other
cause, and that cause is God. In short, the real is characterized by being
those ideas caused by God. However, dreams also are involuntary
although caused by us. Also the problem remains of how we identify
which ideas are God-caused. There is, argues Berkeley, a greater
strength (force and vivacity), order and coherence among ideas we
refer to as being of reality.
There remains too the problem of distinguishing veridical
perceptions from illusions. A stick appearing bent in water is a genuine
perception, since it is not caused by us; it is an illusion, not in isolation,
but in virtue of its relation to the sequence of other ideas we have, such
as whether it is followed or not followed by the experience of a
straight stick if we feel the stick in the water or the sight of a straight
stick if we take it out of the water.
Berkeley seems to say there is an “archetype” (original) idea in the
mind of God which God wills us to perceive. We perceive ideas as well
as imagining ideas. God imagines and wills ideas only; if this were not
the case, we would have to posit an infinity of Gods as the cause of
each other’s perceptions. God wills that we perceive “ectypes” (copies)
of aspects of the archetype ideas in His mind. The notion of two or
more people perceiving the same thing, although their ideas may be
qualitatively different, seems to depend on there being a common
archetype.
We can summarize Berkeley’s ontology in the following way:
136 Empiricism
Reference here to “Minds”, of course, includes the mind of God.
Berkeley’s idealism claims not to question the truth of the
judgements of common sense; rather it claims to affirm them and to
make clear what affirming those truths really means. Berkeley’s world
will appear exactly the same as the world containing matter; it makes
no difference to the course or order of our experiences. Nevertheless
Berkeley’s world is different even if it looks the same.
This brings us to Berkeley’s views on the meaning of words or
terms. The meaning of terms is the ideas for which they stand; if
there is no identifiable idea corresponding to, or associated with, a
term, then it is meaningless; if the term has a meaning at all, it must
refer to some feature of experience: to a particular idea or collection
of ideas.
This leads us to examine Berkeley’s objection to abstract ideas as the
meaning of general terms. Locke had suggested, according to Berkeley,
that it was possible to form abstract ideas from particular ideas and
that this explained the meaning of general terms and their ability to
apply to any particular of a class of particulars similar in some respect;
thus we form the abstract idea of triangularity, which is what the term
“triangularity” stands for, and so it applies indifferently to every
triangle. A general term such as “man” applies to all things of the same
kind, namely men. The abstract idea applies indifferently to all
particulars of a certain class by virtue of including only that which all
the particulars have in common and nothing in which they differ.
Berkeley thinks that Locke’s notion of our forming abstract ideas is
both impossible and unnecessary. It is impossible because the process
of abstraction involves separating qualities that cannot be separated,
and running together qualities that are incompatible. In the case of
triangularity we have to separate off just the property of being a
triangle from that triangle being, for example, any particular or
determinate size or colour; it is also an idea of a triangle which is no
particular kind of triangle, so it must, to be general, be an idea which is
at once all and none of the differents kinds of triangle. Berkeley thinks
that we cannot form such an idea. Abstract ideas are unnecessary
because terms can be general without their meaning deriving from
their standing for abstract ideas: terms become general through their
being used to stand for a class of particulars which are similar in some
relevant respect.
The connection of this with Berkeley’s objection to materialism is
that he sees the route to positing material substance as dependent on
the possibility of abstraction. If we can form abstract ideas, it is
possible to argue that we can speak meaningfully, through the
formation of an appropriate abstract idea, about something that exists
which is not, and could not be, an actual object or content of
experience; we can thereby give meaning to terms such as “matter” or
“material substance” and so refer to something other than what we can
Berkeley 137
actually experience—which is particular ideas—and then posit its
existence independently of its being perceived. If Berkeley has shown
that abstract ideas are impossible, and if abstract ideas are required to
give meaning to terms such as “matter” or “material substance” which
refer to that which can exist unperceived, then he has shown that all
talk of matter or material substance in this sense is meaningless or
unintelligible.
Berkeley is, however, strangely inconsistent in his empiricism,
since he sees fit to talk, and claims to talk meaningfully, about mental
substance and God, of both of which we can never actually have
ideas, so talk of them should strictly be meaningless. Ideas can only
be like other ideas; ideas are passive or inert whereas minds are
active; ideas are thereby debarred from representing spiritual
substance. Berkeley tries to get round this by claiming that although
we cannot, strictly speaking, have ideas of spirit, we can have a
notion of it. He intends by this to contrast spirit with matter: whereas
the latter has been shown to be impossible or contradictory, mind is
at least possible and intelligible, and we can therefore form some
notion of its operations.
The only sense that Berkeley gives to causation is that of active
willing. Ideas themselves are inert and passive, incapable of willing,
and therefore incapable of causal influence. The supposed material
substance in which qualities are said to inhere is also lifeless and
passive, and would therefore be incapable of causal influence. Only
spirits are active; it follows from this that the cause of all ideas must be
some spirit or mind. Some ideas are caused by our own finite minds,
as when we imagine ideas; but the vast richness of our other
experiences must be caused by the infinite mind of God.
When it comes to his analysis of natural science, in particular
physics, Berkeley’s views find powerful echoes in modern
instrumentalism. Berkeley argues against essentialism in physics:
essentialism suggests that beyond the phenomena or appearances that
we observe, the phenomena are caused by and united in an ultimate
reality whose essential nature (such as atomic structure, extension, or
substantial form) finally explains all phenomena and the necessary
connection between phenomena observed to be constantly conjoined.
This necessary connection takes the form of logical deducibility. The
positing of some kind of essential nature is required to give a
foundation to unifying causal laws which are the characteristic aim of
science. A causal law of the form “If A then B”, or “All As are Bs” does
not merely describe the accidental juxtaposition of A and B in our
experience, but aims to identify a necessary connection between A and
B such that we say if A happens, then B must follow, and if something
is A, it must also be B; in short A and B are connected in a way that
could not be otherwise. That there exist such necessary connections
between ideas we experience is denied by Berkeley; no such necessary
138 Empiricism
connection is perceived between phenomena. There are no essential
natures in things beyond experience; indeed, it is senseless to posit an
essential nature in a reality of things beyond phenomena which would
account for the necessary connection of our ideas; all that we ever
experience is a succession of ideas among which we perceive patterns,
associations and regularities. The search for such unattainable
necessary connections only breeds scepticism about the achievements
of science when science fails to show that it can establish how the
world must be. Scientific theories do not present us with the truth
about reality—metaphysics and theology do that—rather their value
lies in their usefulness as general rules by which we can predict
phenomena: what ideas will follow what, and what ideas are
invariably found together. By limiting the aspirations of science
Berkeley hopes to secure science from scepticism, and at the same time
make room for the indispensability of theology.
Ideas are seen by Berkeley as natural signs; the experience of idea X
is a sign that idea Y is about to follow; and it is our job to learn what
these regularities are and to come to know the rules which correctly
map the patterns of ideas; but we must not suppose that we have
thereby discovered necessary connections between the ideas that could
not be otherwise. That the ideas follow each other in regular order is
entirely dependent on the will of God who chooses to present to us
ideas in definite regular patterns, the rules of which we can learn. In
learning the order of natural signs in science we learn the “language of
God”: the signs He systematically presents us with. The experience of
getting closer to a fire will be followed by the experience of pain; but
the two experiences are not necessarily connected; the relation
between the two experiences is contingent; there is nothing about the
experiences themselves, or about any further thing which is the cause
of the experiences, which means that the juxtaposition of the
experiences could not be otherwise. Yet we can trust in God that He
will invariably maintain a regular order which it is possible for us to
learn. In this way science is seen merely as a more systematic attempt
to chart our experience than our everyday understanding, but not
different in the kind of knowledge it produces.
It is, however, not true to say that Berkeley gives a regularity theory
of causation. Although the mapping of regularities between noncausally
associated ideas is the aim of science, real efficient causal
influences take place between spiritual substances and ideas.
On Berkeley’s view, the use in science of various terms such as
“force”, “gravity”, “attraction”, “cause”, “effect”, and “insensible
particles” is harmless provided we do not think that these terms name
real entities in the world which explain the causal necessary
connection of phenomena or events we experience; such terms should
be seen as merely useful suppositions or hypothetical posits which
may aid us in making predictions. They do not describe facts about the
Berkeley 139
world; but we can use them to help us predict phenomena; the
phenomena can be understood as occurring as if they were facts about
reality. From the point of view of facilitating the discovery of the
general rules describing the order of phenomena, the truth of what one
supposes as a mechanism is to be valued purely for its convenience,
and its truth is irrelevant, for its truth as a mere useful supposition
does not arise at all. More positively we can say that anti-essentialism
encourages us always to seek further explanations because it does not
assume there will be, and we might find, some point at which
explanations are exhausted and complete.
Nevertheless, the sense in which we can be said to learn the language
of God gives some residual meaning to scientific theories or laws being
true; not every invariable correlation will constitute a law of science;
the use of the terms language and signs suggests a structure that,
although not necessary, does have an order of meaning and syntax
analogous to that of a language.
Many problems reside in Berkeley’s system. It is difficult to see how
his proof that God exists can be valid if based on the premise that ideas
that are not perceived by our minds must, if they are ideas of real
things, continue to exist, and can do so only in the mind of God. No
possible empirical evidence could verify the proposition that the ideas
constituting object A exist unperceived by us. We are precluded from
establishing by experience the ontological continuity of ideas
constituting sensible objects when we do not experience them by the
fact that any attempt to gather appropriate empirical evidence would
be self-defeating: we cannot get a sly glance at things unperceived.
This is rather like trying to determine whether the fridge light goes off
when one closes the door, except that in the case of ideas constituting
sensible objects it is a logical impossibility, not an empirical difficulty
involving empirically determining if things exist unperceived. If the
only guarantee we could have for knowing real things exist
unperceived is following a proof that God exists, then a proof of the
existence of God cannot, without being circular, use as a known
premise that real things exist unperceived when not perceived by us.
The basis for idealism is that all that we ever perceive is ideas or
sensations—light, colours, sounds, smells, tastes and the like—which
can only be conceived of as existing in the mind. It is this that must be
denied in an effective refutation of idealism. We must say that we can
be immediately aware of physical objects in perception; what we
perceive is appearances or aspects of objects themselves, not other
entities called ideas that mediate between us and objects perceived.
If Berkeley were to stick strictly to his empiricism in using as
evidence only the immediate content of our own minds, then it is
difficult to see how he could avoid extreme solipsism: there is nothing
he can be sure of except the nature and existence of the ideas of which
he is immediately or currently conscious.
140 Empiricism
Hume
David Hume (1711–76) was born in Edinburgh, into a family of the
minor gentry near the Scottish Border; the family home was the estate
of Ninewells, close to the village of Chirnside near Berwick. David
Hume’s father died in 1713, leaving his mother to bring up David and
two siblings, of whom David was the youngest. Their religious
education was Calvinist in character with regular attendance at kirk.
Hume entered Edinburgh University in 1723 when not quite twelve.
Here he received instruction in Latin, Greek, mathematics, physics and
philosophy, and became acquainted with the work of John Locke and
Isaac Newton; but he left the university around 1726 without taking
his degree. By this time he had arrived at the atheism that was to last
for the rest of his life. He returned to Ninewells where, following the
family tradition, it was proposed that he turn to law as a profession;
but Hume had no appetite for the law and instead spent time studying
great classical literature. In 1734 Hume entered the offices of the West
India company in Bristol, but his stay here was very short-lived, and
he went to France where he could live more cheaply, first in Rheims,
and then at the small town of La Flèche in Anjou; here he wrote A
treatise of human nature. He returned to London in 1737 and after some
difficulty eventually found a publisher; the Treatise appeared in 1739
and 1740, by which time he had returned to Ninewells. The book did
not receive the high level of attention he had hoped, although he
exaggerated when he said that “It fell dead-born from the Press.” In
1745 Hume’s application for the professorship of philosophy at
Edinburgh University was rejected. From 1747 onwards Hume earned
his living chiefly as a diplomatic secretary, which involved travel
abroad. During this time he continued to publish short essays on
various topics, and began work on the Enquiries concerning human
understanding and concerning the principles of morals, in which he sought
to rectify the presentational and stylistic deficiencies which he thought
had led to the modest acclaim awarded to the Treatise; the Enquiries
was published in 1748. In 1752 Hume became librarian to the Faculty
of Advocates in Edinburgh, having been turned down in 1751 for the
Chair of Logic at Glasgow despite the support of the vacating
professor, Hume’s friend, the economist Adam Smith (1723–90); it was
as a librarian that Hume began his History of England. In 1761 he
became a personal secretary at the Embassy in Paris and was extremely
popular in Paris society. In 1766 Hume returned to England with the
philosopher Jean-Jacques Rousseau (1712–78); however, Rousseau’s
chronic paranoia and unreasonableness soon caused them to fall out.
Hume retired from work in 1769 and lived in Edinburgh. In 1775 he
was struck by a fatal wasting disease of the bowels and he died the
following year.
Hume’s affable disposition while terminally ill was typical of his
Hume 141
general character; he also remained unshaken in his rejection of any
kind of survival in an afterlife. Although lean in his youth, in later
years he had a rotund physique, and he took pleasure in food and
good conversation. Despite having a formidably sharp intellect, he
seems to have had a generally amiable, sociable, cheerful personality.
There is a tension which runs through Hume’s philosophy between
scepticism and naturalism. The sceptical side involves the employment
of various arguments showing that we lack any rational justification
for beliefs usually regarded as fundamental to our view of the world.
There are three beliefs of particular importance that come in for this
treatment:
(a) existence of causation and the rationality of induction
(b) existence of the external world: bodies continue to exist
independently of us in the external world
(c) existence of a permanent self.
In each case Hume sets out to show that we have no rational
justification for the belief, but also how the belief is a fundamental
product of the faculty of imagination in human nature. Hume’s
purpose in revealing the lack of rational warrant is to show the limits
of what human reason can account for. The naturalist strand in
Hume’s philosophy now enters for he does not draw the conclusion
that because we lack rational justification for these beliefs we ought
therefore to reject the beliefs. It is a fact that we do irresistibly,
invariably and universally hold these beliefs, which are the foundation
of thought and necessary for our survival; if our holding of these
beliefs cannot be accounted for through our possessing sufficient
rational grounds for the beliefs, then it is still to be explained why
nevertheless we hold these beliefs, think the way we do, and remain
unshaken by sceptical arguments directed against them. In short, one
possible explanation for why we hold these beliefs is that we have
rational grounds for doing so, but where we do not have rational
grounds there must be some other explanation for why we have these
entrenched beliefs. The explanation is to be found in the science of
human nature. This science reveals that the way we come to form these
beliefs is the same kind of way as other animals form beliefs; it is
therefore quite proper to say that animals reason.
Sceptical arguments or reasoning can operate only against other
arguments or reasons; but given that the explanation of our holding
certain fundamental beliefs or thinking in certain ways is not to be
found in our having reasons at all, the sceptical arguments or reasons
against these fundamental beliefs or fundamental ways of thinking
find no purchase; rational arguments are simply irrelevant. There is no
question that we ought to think differently because we lack rational
grounds in these cases, as the sceptic suggests, since nature,
specifically human nature, ensures that we cannot help thinking in
142 Empiricism
these ways; these ways of thinking are fundamental facts about human
nature which are explained by non-rational laws describing how we go
on or function; the beliefs thus produced are not thereby irrational;
they would be irrational only if we supposed that the explanation of
our having the beliefs is based on insufficient rational justification and
that rational justification is required. We are psychologically
constituted in such a way that, given a certain course of experiences,
we will inevitably come to hold certain kinds of beliefs.
In our philosophical search for the ultimate foundations for our
beliefs we come to see that certain of our most basic or fundamental
beliefs are rationally groundless or unjustified; but we also come to
understand that they are not the kind of beliefs that can be rationally
grounded or justified; therefore the lack of rational justification is not
to be thought of as a deficiency in these beliefs. They are not the kind
of beliefs which we can be rationally justified or unjustified in holding;
so showing there is no rational justification for the beliefs does not
show them to be irrational or confused; rather, they are non-rational,
but beliefs that we must have resulting from the way our natures
fundamentally are. This position can be further defended by pointing
out that if we enter into the process of giving reasons at all and
suppose that it can ever be successfully brought to an end, there must
be some beliefs for which reasons neither can be given nor are
required; justification has to end somewhere.
An analogy may help. If we take the notion of love we can clearly
understand that cases may arise where L giving reasons to M why M
should love L rather than N is simply out of place; it is not that L’s
reasons are bad reasons; reasons of any sort are simply irrelevant and
make no difference; it may just be a fact that M loves N and not L, and
that is an end to it. One might as well argue with a tree that it is
unusually early to come into leaf, or with an avalanche that it is wrong
to fall on villages.
This naturalism has serious consequences for anything like
Descartes’ project for an absolute, non-species specific, objective
conception or understanding of the world based on pure reason, not
on concepts dependent on our contingent biological or psychological
constitution. For it turns out that some of our most basic conceptions
and beliefs are not transcendent and eternal, but depend on
contingent facts about human nature. Descartes supposes that the
fundamental conceptions involved in a truly scientific view of the
world are are either intuitively obvious or rationally justifiable, and
thereby are true universally for any intelligence whatsoever. Hume
argues that these conceptions are dependent on human nature being
what it is and functioning in certain ways, and without a nature
which reacts in certain ways to experience such conceptions or ideas
would not arise at all, since they cannot be derived from or justified
by universal and valid deductive reasoning or experience. Our
Hume 143
fundamental concepts and beliefs, which we apply to, and regard as
real features of, the world, are species-dependent, not nonperspectival
and absolute. That we have an idea of, and belief in,
causality and induction, a belief in external physical bodies and in a
relatively permanent self, depends on our reacting to certain
experiences in certain ways; such ideas are neither a product of the
pure necessity of reasoning nor derivable from passive observation of
the world; our having these ideas depends on experience combining
with the way human nature functions.
The tension between scepticism and naturalism arises from the
uncertainty as to whether any particular case of a belief lacking
rational justification should lead us to reject the belief or lead us to
conclude that it is vain and unnecessary to ask for justification. The
answer would seem to involve assessing how fundamental the belief is
to human nature; that is, to what extent it is universal, irresistible and
permanent.
Hume maintains the view common to other philosophers of his
period that we are only ever immediately acquainted with the contents
of our mind: perceptions. He divides perceptions in the mind into
impressions and ideas. These are to be distinguished not by their origin,
but by their degree of force and liveliness; impressions are lively
perceptions or experiences and ideas less lively. Impressions are the
primary or first appearance in the mind of any mental content, ideas
are secondary and derivative weaker copies of impressions. Roughly
the distinction is between actually experiencing X, and thinking about
X. Fundamental to Hume’s philosophy is that ideas, which are,
generally speaking, the materials of thought, are faint copies of
impressions and that we cannot have a simple idea of which we have
not had a simple impression. Every simple idea has a corresponding
simple impression that resembles it, and every simple impression a
corresponding idea; that is, every simple perception appears both as
impression and idea. This account has the odd consequence that to
think about X (say, a pain) is mildly to experience X (a pain), which is
surely false. We can have complex ideas of which we have no
corresponding complex impressions, but only if they are made up of
simple ideas copied from simple impressions we have had. The reason
for this view is that Hume wishes to identify the correct impressions
from which we derive ideas. There are two possible sources of ideas:
impressions of sensation and impressions of reflection. Impressions of
sensation are basically sense-experiences; impressions of reflection are
often new impressions which derive from the natural way we react to
certain impressions of sensation. If we have an idea which is derived
from an impression of reflection in this way, then the existence and
nature of the resultant idea partly depend on the workings and nature
of our mind, and the idea is not something derived wholly passively
from experience of the world. The question is whether we are then
144 Empiricism
justified in regarding the resultant idea as corresponding to a real
feature of the world, or whether the idea does not correspond to a real
feature of the world since it is simply a product of the way we react
naturally to certain impressions of sensation, which in themselves do
not contain that idea. For example we find that the idea of evil and evil
acts is not derived from anything observed purely in acts, but results
from the impression of reflection, abhorrence, we naturally feel, because
of the reaction of human nature, at seeing certain acts; that we then
regard evil as really in the world, and certain acts as abhorrent, results
from the idea of evil being projected onto certain acts in the world,
although it is not derived from something observed passively as really
being in the world. If we did not react in certain natural ways to
produce these impressions of reflection we would not, from observing
the world, find any passive impressions of sensation from which the
idea of evil could derive.
The meaning of a term is to be found in associating the term with
the correct idea. If we cannot find any impression of either sensation or
reflection as the origin of an idea which is presupposed in the
corresponding term having meaning, then we must conclude that we
are deluded when we say we have the idea, and the term which
publicly articulates the supposed idea is in fact meaningless. But we
must look carefully; if we cannot find an impression of sensation
(perceptions of red, chairs, mountains, as well as sensations such as
hot, cold, pain) we may well find an impression of reflection (feelings,
passions, emotions, basic appetites, such as anger, sadness, hunger)
from which an idea we have derives; but this has the important
consequence that the true meaning and implications of the term
corresponding to the idea may be quite different from what we
thought them to be. We will have to conclude that if an idea derives
from an impression of reflection or inner sentiment only, then it is not
an objective feature of the world, but one that depends on our natural
propensity to react to experiences in certain ways according to our
human nature.
Hume distinguishes between memory and imagination on the basis
of the distinction between impressions and ideas. Memory: the order/
sequence and combination of ideas is the same as the original order/
sequence and combination of the impressions. Imagination: the order/
sequence and combination of ideas can be different from the original
order/sequence of the impressions.
The imagination is of fundamental importance for Hume’s account of
why we have the beliefs we do have. The order/sequence and
combination with which ideas feature in our imagination is not
random but has rules governing that order; there are forces of
attraction, which, although not intrinsic to the ideas by themselves,
govern the way simple atomic ideas and complex ideas are associated
as a result of fundamental propensities of human nature.
Hume 145
Hume argues that all perceptions are really distinct from each other;
they can exist at different times; they can thus be conceived existing
separately without any contradiction; therefore any connection, if it
exists at all, between perceptions is contingent and not necessary. It is
the human mind that, according to certain natural propensities,
associates perceptions which have logically distinct existences and
between which no necessary connections are ever discovered by
reason or observation; but it is from this feeling of being determined to
associate ideas in certain ways, which is an impression of reflection,
that the idea originates of the perceptions themselves being necessarily
connected.
Hume sees part of his function as explaining why we hold certain
fundamental beliefs; this he does through discovering and calling
upon the laws governing the order of perceptions in our minds. The
basis of Hume’s explanation is the “principle of association of ideas”;
this explains why we in fact think as we do, although we may have no
rational justification for doing so. Ideas become associated in our
minds in specific ways and this controls the order or sequence of
thoughts through our minds. There are three main factors that
determine which ideas are associated in the human mind:
(i) resemblance: qualitative similarity
(ii) contiguity: proximity in space and/or time
(iii) cause and effect: the thought of one idea leads to the thought of a
causally connected idea.
The mind naturally moves smoothly from one idea to another in
accordance with these principles of association. If we have an
impression of A, or entertain an idea A, we naturally move to the idea
B related to it in the highest degree by some or all of the above
principles. Ideas are mental atoms among which Hume attempts to
describe the rules governing their behaviour.
The objects of human understanding and inquiry fall into two
exclusive and exhaustive classes. The distinction is sometimes called
“Hume’s fork”: this contends that all meaningful propositions can be
divided into one of two types:
(I) relations of ideas
(II) matters of fact and real existence.
All propositions of type (I) concern the abstract relation of ideas, and can
be known to be true a priori because their denial would involve a
contradiction and they are thus necessary. Examples are truths in
mathematics and logic. They are intuitively or deductively certain. The
examination of the meaning of the constituent ideas of the propositions
alone is sufficient to establish their truth or falsity. All propositions of
type (II) concern connections between matters of fact and the actual
existence of things, and can be known to be true, if at all, only a posteriori
146 Empiricism
by experience, and not through examining the meaning of the constituent
ideas alone because their denial does not involve a contradiction and
they are thus contingent. Examples are the propositions of natural
science and common-sense statements of fact. The price of our knowing
propositions of type (I), however, is that they are trivial truths that can
tell us nothing about what is actual and contingent, but only what is
possible (not contradictory), impossible (contradictory) or necessary
(denial is contradictory). Thus we cannot know any truths about the
actual contingent or real world a priori by pure logical reasoning alone; if
we can know truths about the world at all, we must rely on the evidence
of experience. Propositions that do not concern either relations of ideas or
empirical matters of fact are meaningless.
Closely connected with this is the way Hume shows that we lack
reasons for our fundamental beliefs by showing that the only two
possible sources of rational justification do not provide reasons for
those fundamental beliefs.
(I’) Reason
Justification by intuitive, demonstrative, deductive or logical a
priori reasoning.
(II’) Senses
Justification by the evidence of observation or a posteriori
experience.
These are exhaustive of the sources of rational justification. Hume
purports to show that rational justification from either source,
demonstrative reasoning or experience, is lacking for our fundamental
beliefs in causation in the world and inductive inference, in the
existence of physical objects in the external world, and in a persistent
self; thus they cannot be rationally justified at all; nevertheless, the
mechanics of the mind are such that we hold irresistibly these beliefs so
necessary for our survival. Hume’s positive contribution is to give an
account of why, in fact, given that rational justification cannot account
for it, we do hold these basic beliefs.
Why, in particular, do we form beliefs about matters of fact that we
have not observed on the basis of what we have observed?
Characteristically this takes the form of an inductive inference of the form:
But is there any rational justification for this inference? Take, for
example, the propositions “All unsupported bodies fall”, “The sun will
rise tomorrow”, or the propositions “All A is B” and “If A occurs, then
B must occur”. These are characteristic of the propositions of natural
Hume 147
science and common sense. Is there any rational justification for our
assertion of these propositions? As it stands the above inductive
inference, which might be used to support such propositions, is clearly
deductively invalid: it is possible for the premise to be true but the
conclusion false. In all such instances we move from cases we have
observed to cases we have not observed on the supposed basis of there
being a causal relation. That is, A is the cause of B, which supposes that
A occurring is necessarily connected with B occurring.
If the inference from A we have observed to B we have not
observed, and the belief that they are necessarily connected, is to be
rationally justifiable, it must be because of (I’) reason or (II’) the senses.
Hume thinks both fail to provide such rational justification.
Hume is clear that the causal connection between A and B, which
describe events in the world, is not explained and rationally justified
by (I’): its being logical or deductive. The relation between event A and
event B is not like the relation in a deductive argument between
premises and conclusions. If the connection were deductive, and hence
logically necessary, then the assertion of A and the denial of B would
involve a contradiction. But in the case of connections of events or
matters of fact this seems never to be the case; the assertion of a
matter-of-fact connection and its denial seem equally conceivable. The
logical relation which holds between a plane figure being three-sided
and its being triangular, or its internal angles being equal to 180
degrees, is the kind of relation that would, if it applied, make a
connection necessary and enable us to justify rationally the inference
to cases we have not observed from cases we have observed; but such
a relation does not hold between events in the world corresponding to
our ideas of them A and B. A and B can exist at separate times,
therefore A and B are separable in thought; the existence of A can be
supposed without supposing the existence of B, and the assertion that
A is always found with B is therefore a contingent, not a necessary,
truth. In short, if it is ever the case that A and B can exist at different
times, we can conceive of A and B as separate, and any connection
between them cannot be necessary.
It might seem as though the causal relation is deductive, and thus
we can know a priori the connection between A and its causal
consequences B because we know the kind of thing A is: say a billiard
ball. But the question arises as to how we know what kind of thing A
is. Hume argues that in the case where A is something entirely new to
us, we can know that B, or anything else, will follow only by experience.
Logically speaking, apart from what would be logically contradictory
anything could happen. If it seems as though we can deduce B from A,
this is because we have already observed the behaviour of A-like
things and included in the definition identifying A as A (what is an A)
the relation to B. We cannot from examining A-in-itself or alone prior
to any experience of A, before a characterizing definition of it that may
148 Empiricism
include B as a causal consequence, deduce what will follow. In
identifying A as an A—as something of a certain sort—we already have
to include certain potential causal consequences; we cannot separate
what we mean by an A—what A is—from all its causal consequences.
To show that we could by pure reason alone deduce B from A, and yet
by this produce new non-trivial knowledge, we would have to define
A independently of its causal consequences, but this is impossible if
what we mean by an A—and hence use to identify something as an A
in the first place—must include the range of A’s causal effects. That
certain causal consequences are connected with A is not something
that can be known a priori.
Alternatively it might be the case that we make the inference from A
to B, and are rationally justified in doing so in accordance with (II’),
because we observe in experience a necessary connection between A
and B when observing the conjunction of an instance of A and B, or B
following A; but in fact we observe no such necessary connection
between A and B, but simply observe A and B occurring together. The
hammer is thrown, hits the window, and then the window breaks;
there is no necessary connection observed as part of this, but rather a
sequence of logically distinct events. We observe no necessary
connection between observed matters of fact themselves, but only
events conjoined with or following one another. Nevertheless, we still
believe some events and ideas to be necessarily connected, and it
remains to be explained why we do so.
Partly the formation of beliefs about the unobserved on the basis of
the observed is founded on the principle that “Every event has a
cause”. But Hume shows that this principle lacks rational justification
by showing that it cannot be justified either by (I’) or by (II’); this
applies Hume’s fork. First, it is not a necessary logical truth, since its
negation does not imply a contradiction; the assertion of an uncaused
event is conceivable. He notes that the assertion of an uncaused event
does not involve the contradictory assertion that the uncaused event is
caused by “nothing”, rather it asserts that the event has no cause at all.
Second, it is not a truth that can be known empirically, since it can be
neither established nor refuted by experience; logically there is no
hope of examining all cases. It cannot be confirmed because we cannot
examine all cases to show every case has a cause; it cannot be refuted
because in any given case we cannot examine and exclude everything
that might be a cause.
If the causal relation between A and B is not deductive, then the
move from the observed to the unobserved on the basis of observed As
and Bs is an inductive inference, and if the assertion of general
propositions such as “All A is B”, or “If A occurs, then B must occur”,
is to be rationally justified, then they depend on some kind of
“uniformity of nature principle”: that conjoined events that we have
observed will hold in cases we have not observed. Thus, events that
Hume 149
we have observed to be constantly conjoined in the past will continue
to be so in the future. In short, the uniformity principle asserts that the
laws of nature will hold in cases we have not observed, and the future
will resemble the past. The acceptance of the uniformity principle
would make the inference, from cases we have observed to those we
have not, rationally justified—the inference would be valid—by acting
as a required premise in the inference from instances we have
observed to those we have not observed.
But what is the rational justification for the truth of the uniformity
principle itself? Again Hume’s fork is applied, this time in testing the
uniformity principle. First, if the uniformity principle were a logical
truth, then its negation would be a contradiction, its denial
inconceivable; this is clearly not the case; it is certainly conceivable that
any law which has operated in all cases until now should cease to
operate in the future and should fail to operate in cases we have not
observed. The uniformity principle cannot be known a priori. Secondly,
the uniformity principle, if it is itself merely a further matter of fact,
cannot be justified by experience for any such attempt will be
irredeemably circular. We might try to justify the uniformity principle
by experience a posteriori by saying: the uniformity principle itself has
always operated or held in the past, and so it will continue to operate
or hold in the future, therefore the uniformity principle is justified by
experience. In applying this to the uniformity principle itself such a
justifying inference is circular, since it is exactly the kind of inference
which depends for its validity on accepting the uniformity principle:
that past observations are evidence that the future will operate in the
same way.
The startling conclusion that Hume draws from his analysis of our
belief in unobserved matters of fact is that such beliefs lack all rational
justification, and thus having a rational justification is not responsible
for our making the inference from observed A to unobserved B
following the observation of A and B conjoined in the past. We do not
make the inference from A which we observe to B which we do not
observe because we are rationally justified in doing so. Nevertheless
we do make the inference, so there must be some other explanation for
why we make it.
Hume gives on the one hand an account of causation and what is
involved in the idea that events are causally connected, on which all
150 Empiricism
moves to unobserved matters of fact depend; and on the other hand an
account of the conditions under which we hold the belief that events
are causally connected.
C’ (a) spatial and temporal contiguity
(b) temporal priority: cause comes before effect
(c) necessary connection between cause and effect
C’’ (a) observed spatial and temporal contiguity
(b) observed temporal priority: cause comes before effect
(c) observed repeated constant conjunction.
The reason for these accounts is that Hume wishes to argue that C’
describes the necessary and sufficient conditions for events being
causally linked and what is involved in the idea that they are, but we
come to hold the belief that they are causally connected in just those
conditions or circumstances described in C’’, and those conditions or
circumstances do not rationally justify the belief as true, nor is there
any other way of doing so.
When we believe A and B in sense C’ to be causally connected:
(1) We make the inference from observed As and Bs to unobserved
As and Bs.
(2) We believe or expect, not merely think, that B will occur
following a fresh observation of A.
(3) We believe the connection between A and B to be a necessary
connection: that it could not be otherwise.
We do not have any rational justification for the inference involved in
(1), for the relation between A and B is neither logical nor justified by
experience. We have no rational justification for the belief (2), since it
cannot be based on either logic or experience. We do not have any
rational justification for the belief (3), since the relation is not logically
necessary, nor is a necessary connection between instances of events A
and B something we observe in experience of A and B; we observe A
conjoined with B, but we do not, as a feature of our experience of them,
observe any necessary connection.
So (1) our making the inference from A to B, and (2) our belief that B
will follow an observation of A, and (3) our belief that A and B are
necessarily connected are not explained by our being rationally justified
in the inference or the beliefs. Hume’s conclusion is that these are not a
matter of rational justification at all. It still remains to give an
explanation of these matters.
The explanation Hume gives returns us to features of human nature,
the principles of the association of ideas and how we react to certain
experiences. The explanation in all the cases (1), (2) and (3) derives
from habits or customs of the imagination: mental habituation. This
tendency to mental habituation is a propensity of human nature. The
basis of the explanation is that repeated observation of the constant
Hume 151
conjunction of A with B as in C’’ sets up a habit of association in the
mind of A and B, and it is this that leads us to (1) make the inference
from A to B in cases we have not observed, (2) believe that B will occur
having had a fresh impression of A, (3) believe that A and B are
necessarily connected.
Hume gives an account in C’’ of the circumstances or conditions in
which we in fact come to judge that A and B are causally connected,
rather than where we are rationally justified in so doing. The
explanation of our belief in causal connections then derives from the
product of those circumstances and our natural reactions in those
circumstances. Following the repeated observation of the conjunction
of A and B in our experience in accordance with the conditions C’’
there is set up in our minds the habit or custom of associating A and B;
and these are just the circumstances in which we say A and B are
causally connected. Taking points (1), (2) and (3) above in turn, Hume
gives the following accounts.
(1) Making the inference from A to B
It is just a fact about our fundamental psychological constitutions
that in circumstances C’’, following the observed repetition of A
and B in conjunction, we do make the inference from A to B. The
repetition of A and B constantly conjoined in our experience sets
up a habit or custom such that on the observation of A we
compulsively move to the idea that B. Thus we infer the idea of B
from the idea of A in cases we have not observed, but the move is
not a rational move at all, since it is neither deductive nor justified
by experience.
(2) Believing that B will follow A
To understand our expectation or belief that B will follow A on
observing A in conditions C’’, we must understand what a belief is
for Hume. He explains a belief as being just the degree of liveliness
or force of an idea, and not a difference in, or addition to, the
content of an idea; the difference between merely conceiving or
thinking X and believing X is a matter of the force and liveliness
with which the idea of X strikes us. In the case of believing B will
follow A, Hume’s explanation is that there is a transference of force
by a kind of inertia from the fresh impression of A to the idea that
B, which enlivens B, where the habit of associating A and B exists,
and this turns the mere thought of B into a belief or expectation—a
lively or vivid idea—that B. It should be pointed out that
sometimes Hume presents a somewhat different theory of belief,
whereby it is a difference of attitude towards an idea, or the
manner in which an idea is conceived or entertained, which
constitutes believing an idea, and which makes believing
something feel different from an imagined fiction: it is an idea
being more strongly or vividly conceived or entertained that
152 Empiricism
constitutes a belief in an idea, rather than a difference in the
vivacity of the idea itself. It is not clear if these two theories can be
reconciled: in the first theory, belief is a matter of how an object of
thought strikes us, in the second theory it is a matter of how we
take hold of the object of thought.
(3) Believing that A and B are necessarily connected
The inference of B from A is not based on the necessary connection
of A and B; rather, the idea of the necessary connection of A and B,
essential to the belief in a causal relation C’, depends on our in fact
compulsively making the move from the impression or idea of A to
the idea of B following the repeated observation of the conjunction
of A and B as in conditions C’’. We have no impression of a
necessary connection between A and B derived from observing the
conjunction of A and B themselves: we just see A happen, then see
B happen. But if the idea of necessary connection, and hence our
belief in causation between events, is not to be a delusion and
meaningless, there must be some impression from which it derives.
The idea of necessary connection derives from a new impression of
reflection, which in this case is the feeling of determination resulting
from the mental habit of our passing from the idea of A to the idea
B, following previous repeated observation of the constant
conjunction of A and B. The idea of necessary connection does not
correspond to anything in the impressions of A and B themselves,
nor does it arise from the perceived repetition of their conjunction
alone, which would in itself produce no new impression; it
corresponds to a new impression of reflection which is a generated
feeling of determination, as we habitually pass in the mind,
because of an associative propensity of human nature, from the
idea of A to that of B, on having repeatedly had experience of the
conjunction of A and B. The idea of necessary connection, and that
of causality which depends on it, would not have arisen at all,
because there would have been no corresponding impression from
which it could arise, except for the propensity of human nature to
produce a suitable new impression of reflection; no impressions of
A and B would alone be sufficient to give rise to the idea of
necessary connection. There is no circularity involved in this
account: the idea of necessary connection derives from the feeling
of determination, whether there is actually any determination or
not, because we in fact move and have a propensity to move from
A to B, which establishes a habit in our minds, following exposure
to the repeated observation of constant conjunction of A and B.
That the idea of necessary connection derives from an impression
of reflection or feeling in this way has a very important
consequence: that the necessary connection, and therefore causal
connection, that we suppose to exist between events themselves
and our ideas of those events is, in fact, in the mind, not an
Hume 153
objective feature of the world; it is something we project onto the
world owing to habit, not something observed in events in the
world, and it is falsely regarded as an objective feature of the
world or real relation connecting events we observe.
In sum, the belief in causal connections, which includes necessary
connection, depends on our natural movement from one idea to
another, not the other way around.
Hume gives an analogous account of the remaining fundamental
beliefs (mentioned at the beginning of the Hume section): (b) the
existence of the external world: bodies continue to exist independently
of us in the external world and (c) the existence of a permanent self.
The strategy is the same: we have no rational justification for these
beliefs through reason or the senses, but nature through the
imagination has ensured that we have these beliefs, and human nature
gives an account of this non-rational mechanism. We believe that there
are bodies existing continuously and independently of us, and that we
are the same self over time.
Hume begins by saying that it is vain to ask if bodies (external
material objects) exist or not, since we cannot help believing that they
do; the question of interest, therefore, is what accounts for having that
belief. The belief in the external world is constituted by a belief in
objects that exist continuously (when not perceived, for example) and
exist independently of perceivers. Reason cannot justify this belief: not
only is it not the case that most people use rational arguments to come
to this belief, but also it is not possible to give a demonstrative proof
that the external world exists such that a denial would be a logical
contradiction. The senses cannot justify the belief: all that we have, if
we examine our sense-experiences or perceptions carefully, is
impressions which are perishing (non-continuous or interrupted) and
dependent (internal or mental) for their existence and nature on
perceivers. All that we are aware of is perceptions which are perishing
and dependent; we do not perceive any objects distinct from
impressions. So what features of our perceptions lead us to believe, or
produce the belief, that our impressions of sense are of external
material objects which do exist continuously and independently of us?
It is not the force or involuntariness of certain impressions that
accounts for the belief, for these are features of impressions, such as
pains, that we do not suppose exist independently in the external
world. The features of our sense-experience from which the belief
derives are the constancy and coherence of certain series of perceptions
which lead the imagination, operating according to certain propensities
of human nature, to overlook the fleeting and internal nature of
impressions. The series of perceptions can be constant in that there are
resembling collections of perceptions in a series even though there may
be gaps between them, as when I look at the table in my room, go out
154 Empiricism
and come back and look again. The series of perceptions can be
coherent in that although the collections of perceptions in a series
change, they do so in a predictable way, as when I come back to my
room and find the fire has burnt down as expected. First, we resolve
the conflict between the gaps in our perceptions and their constancy by
regarding the gaps as only apparent, with the object of our perceptions
really continuing to exist in the gaps. Second, we explain the coherence
of our perceptions by the supposition that the objects perceived exist
constantly and independently in the gaps when not perceived.
Our belief in continuous and independent objects is one in
something that preserves identity through time; this would strictly
involve perceptions which are invariable and uninterrupted. We have
bundles of perceptions which, although perishable and interrupted,
also exactly resemble each other and thus they exhibit constancy.
Because these bundles exactly resemble each other the human mind
overlooks the gaps and lazily treats them as if they were the same
uninterrupted perception. Thus we come to form the belief in, or lively
idea of, continuous and independently existing objects corresponding
to these bundles of perceptions; the belief or lively idea which fills the
gaps itself derives its liveliness from the resembling impressions either
side of the gaps in our perceptions. In short, we naturally and
habitually confuse a series of interrupted but resembling perceptions
with the alike single continuous perception that would be invariable
and uninterrupted, and thus believe that sensible objects exist as
continuous and independent objects.
The belief in the self, or a personal identity that persists over time,
receives similar treatment. Its existence is indemonstrable by reason.
Through experience when we look into ourselves we do not perceive
anything corresponding to the permanent self, or spiritual substance,
in which perceptions inhere, but only particular fleeting perceptions
themselves. The human mind is really a bundle of distinct perceptions
between which we perceive no real or necessary connection. The
explanation for the belief in the self which we nevertheless have arises
from the natural association of ideas which is a product of the
perceptions themselves with unavoidable wedding or associative
propensities of human nature, giving rise to an impression of reflection
which is a feeling that the ideas are connected; but this association of
ideas and the consequent feeling of connectedness between the ideas
depend on us and our nature, and the connection is not a real
connection between the ideas themselves. It is from this feeling of
connectedness, which is an impression of reflection, that the idea of the
mind being unified in a single self, which is a continuous and
unchanging thing, arises and is ascribed to what are really separate
and variable perceptions; this leads us to mistake what is really a
collection of logically distinct perceptions for something that is
connected in a unity and has identity.
Hume 155
Generally we cannot know if the connections we feel exist between
perceptions are real, for we never perceive necessary connections
existing between them, but merely perceive one following another. In
fact, we know perceptions to be distinct existences or atomic; they are
able to exist independently of each other without logical contradiction.
The idea of connection between them is just a copy of a parent
impression of reflection—the sentiment or feeling of determination in
the mind as we naturally associate ideas—but we can have no
knowledge of whether the connection actually holds.
Nature has taken care that we hold our most fundamental beliefs.
We irresistibly believe in causation and inductive inference, and
believe in the existence of independent continuous external bodies and
a persistent self, even though we have no rational justification for the
beliefs from reason or experience. Thus nature ensures that the
arguments of the sceptic find no purchase against processes that are
not a matter of rational justification at all but are a matter of deep
instincts in human nature.
156 Empiricism
continued to ...