1. Hume’s Problems with Induction
The psychology of induction
Hume tried his best to do away
with the science of induction by psychologizing our understanding of it. Of
course, there is a psychology of induction, since humans have a psyche and
induce. But Hume attempted to reduce induction to psychological mechanisms, i.e.
to substitute a psychology of inductive thought for the logic of inductive
thought. He proposed a description that effectively eliminated the possibility
of evaluation and prescription. He sought to permanently undercut all attempts
to validate induction.
With this goal in mind, Hume
proposed a psychological theory of generalization. Generalization
was to him a mere quasi-mechanical or instinctive reaction of expectation due to
repeated imprints in the mind; it was, effectively, an acquired habit.
Essentially, Hume was arguing that the repeated experience of cases of X that
are Y drives us to conclude that all X are Y (i.e. to expect that yet unseen
cases will conform to past experience), even though in principle things might
well (and often do) turn out otherwise.
But according to inductive
logic, Hume’s theory is just a hypothesis that has to, itself (like all
hypotheses), be confirmed repeatedly and never infirmed. Hume cannot regard it
as somehow exempt from or transcending inductive logic. It is subsumed by it
like any other theory. In fact, there is no psychological drive such as Hume
projects – and his theory is itself proof of that, since he himself is aware
that things might (and often do) turn out differently than expected.
It is important to notice that,
in practice, while we do frequently generalize, we often do so tentatively fully
aware that we might have reason to change our minds later on. Moreover, we often
abstain from generalizing, because we do not want to proceed hastily or because
we are already aware of contrary evidence. Also, we often particularize after
having generalized, due to coming across new evidence to the contrary.
It follows from such simple
considerations that Hume’s claim to a psychological law is empirically
inaccurate. It is a false observation, an overly hasty generalization from
limited or selective introspection. Not only does it not explain the phenomenon
of generalization, nor replace the need for a logical and epistemological
treatment of the issue, it is an erroneous psychological claim, incorrect
Another attempt at reductive
psychologizing was Hume’s attempt to write-off causation as mere association
of ideas. Basically, this suggests, Hume had personal difficulty
distinguishing the fact of causation from our way to knowledge of causation;
because he confused the two issues, he tried to conflate them.
notion of association of ideas was of course his belief that what we perceive
(when we seem to perceive the world) are not things in the world out there but
images of such things produced in the mind through sensations. Due to this
erroneous (because internally inconsistent, self-contradictory) analysis of the
experiential process, he seems (in some people’s eyes) to have some
credibility in affirming causation as mere association of ideas.
For Hume effectively adopted
John Locke’s theory of human knowledge as his starting point. This theory
admittedly seems like common sense: we have senses and they obviously somehow
produce images and memories in us. However, this is the basis of the worldview
that has come to be called Naïve Realism (or uncritical materialism). It seems
reasonable, but upon reflection it is found to be wobbly.
If the senses truly produce
images in our minds of the world beyond them, it follows that we have no direct
knowledge of the world out there at all, but only knowledge of the said images
(this term here intends all phenomenal modalities, i.e. not only sights, but
also sounds, smells, tastes, and various touch sensations). In that case, how do
we know of the bodily senses at all, and on what basis could we at all affirm a
world beyond them? It is a seemingly inextricable dilemma.
At first glance, to affirm that
our cognitive relation to the world out there is mediated by ideas seems
innocuous. It seems obvious enough that our ideas, or most of them, somehow
‘represent’ or ‘correspond to’ the world. But upon reflection, such a
view of how our knowledge is constituted and justified is logically untenable.
How can we claim our ideas representative or correspondent to reality if we have
no immediate contact with it by which to make this judgment? How indeed can we
even claim our ideas not to represent or correspond to reality? We are
seemingly doomed to utter ignorance.
To his credit, Hume (unlike
became aware of the insuperable difficulty that the common sense theory of
knowledge raised. Less to his credit, Hume derived a deep skepticism from this
puzzle, because he effectively assumed there was no other approach. That is,
rather than considering Locke’s particular theoretical approach to have caused
the dilemma, he viewed the problem as a definitive cause for doubting all human
knowledge as such.
That such a radical doubt in
turn cast doubt on his own faculty of knowledge and conclusions apparently did
not cross Hume’s mind (or not sufficiently). For, though henceforth
fundamentally a skeptic, he continued seeking and claiming knowledge. But he did
not try very hard to find a solution to the inherent problem. He never
discovered the solution made possible by a phenomenological approach.
This approach is encapsulated by
the aforementioned principle of induction, which starts the enterprise of
knowledge with regard to appearances rather than to sense perceptions.
‘Appearances’ refers to the contents of consciousness irrespective of their
source, so this term does not have presuppositions like ‘sense perceptions’.
It is not a verbal issue, but one of ordering of knowledge, note well. In a
phenomenological perspective, Locke’s theory regarding sensations and ideas is
just that – one attempted explanation of certain appearances. Seen in this
light, the difficulties it presents seem far less threatening.
Now, all this is said here only
to explain why Hume was more or less bound to opt for a reduction of causation
to ‘association of ideas’. Since his viewpoint effectively divorced
ideas from their objects, he could not talk about the objects themselves without
some nagging discomfort, and he was pretty well cornered into rather discussing
But it must be stressed that for
us, who are free of the dilemma posed by Locke’s theory thanks to a more
phenomenological approach, the scenery looks very different. We can logically
distinguish ideas from the objects they intend – be these objects physical,
mental or spiritual. Although ideas might conceivably always appear in certain
sequences, this is not for us sufficient reason to declare the objects they
intend to be causally related.
Here again, we must apply
deductive and inductive standards to judge the issue.
For a start, it is worth
pointing out that the concept of association of ideas is inherently one of
causation. Leaving aside Hume’s view of causation as mere constant conjunction
as against connection, to say that ideas are associated in some way is to claim
a connection of some sort between them. If we think in terms of one idea
‘giving rise to’ another, or we use any other such expression, we are
thinking causation. The implication may be tacit, but it is clearly there.
That the causal sequence
concerns the specific kind of thing we call ideas, rather than the kind of thing
we call objects, is irrelevant to the relation itself, which is conceived as
technically the same irrespective of the kind of thing related. Causation is a
certain kind of relation between terms or theses, which has nothing to do with
their actual contents.
To say that the idea of X causes
the idea of Y is as much a claim to causation as to say that X causes Y. The
formal proof is that we can call “the idea of X” a special case of X, and
“the idea of Y” a special case of Y. In formal logic, X and Y are symbols
for any two terms; they are not reserved for objects as against ideas. For this
reason, the principles developed with regard to X and Y are universal.
If we formally admit a causative
relation between ideas (or impressions, sensations, concepts, beliefs, thoughts,
or any such mental phenomena), there is no reason for us not to admit a
causative relation between other kinds of things (i.e. between non-ideas, viz.
the objects of most ideas). To accept the one and refuse the other, as Hume
does, can only be arbitrary, for there is nothing to formally distinguish the
two. The variables differ, but the underlying relation between them is the same.
In short, our use of the word
association in one case and causation in the other is a mere verbal
embellishment. Hume’s main argument is thus based on a superficial verbal
distinction. And here again, his attempt to substitute psychology for logic is
implausible. The truths of logic are independent of any psychological thesis.
Secondly, Hume is incoherent
when he formulates a concept of association of ideas that is meant to exclude
a concept of causation between the objects the ideas refer to. Such an exclusive
contrast between the two concepts commits the stolen concept fallacy. For to
invalidate the association of ideas, i.e. to point out that ideas may be
erroneously associated, we need to have a more objective knowledge to compare
to. It is logically impossible to claim associations of ideas to be occasionally
or inherently wrong, without claiming separate knowledge of the true causation
between the objects concerned.
In the very act of
downplaying or denying causation between objects by positing association of
ideas, Hume is relying on his and our past experience that sometimes
associations do not match causations. If we had no such past experiences, we
could not comprehend Hume’s argument, or be convinced by it. Hume’s
discourse tacitly implies his and our ability to grasp causation independently
of association, i.e. that we all have access to some objective reality.
Hume is here committing the same
silly error Kant would later commit when claiming that things as they really are
(“in themselves”) are radically different from things as they appear. How
could he know it? No one can consistently postulate a conflict between
reality and appearance without having access to both. If someone accuses humans
of total delusion, he forfeits all logical right to discuss the presumed
‘real’ world, for all such discussion (even hypothetically) would be
self-contradictory, since it is itself a claim to some knowledge.
The critic cannot claim to be an
exception to the general rule he posits. We cannot project a scenario that
excludes us – but some people keep trying to! We admittedly all have some
illusions sometimes; none of us are infallible – but this is a far cry from
It should be noticed that we are
well able to distinguish the two classes, i.e. ideas and objects. Hume does so
in practice, though he denies our ability to do so theoretically. Indeed, how
could his discourse be at all meaningful to him and us, if we could not all make
the distinction? If apparent objects were truly no more than ideas, it is
doubtful we could even imagine such a distinction; certainly, it would be
logically self-contradictory in the way that Kant’s dichotomy later was.
Thirdly, let us consider the
facts of the case in more detail. Note that we ordinarily pass no time wondering
whether our ideas are repeatedly conjoined, but only concern ourselves with
their objects. Moreover, we might ask whether any two ideas are ever in our
actual experience constantly conjoined; the answer seems evident to me – it is
no. On the other hand, many objects do seem to us constantly conjoined.
Moreover, if we introspect
sufficiently, we easily notice that ideas may become associated in our minds for
reasons that have nothing to do with the objects they intend. Such association
is not based on constant conjunctions, but on a single coincidence. The strength
of mental association is not due to statistical frequency. For instance, a
certain musical tune reminds me of a certain woman, just because it happened to
be playing in the restaurant where we sat the day I met her. I may well have
heard the same tune a hundred times before, without any association occurring.
This means that in our common
everyday experience, without reference to Hume, the conjunction of ideas and the
conjunction of the objects they intend are two quite different issues. Even
if we observed our ideas and found them constantly conjoined, we would not
necessarily conclude that the objects they intend are causally related; we are
not (most of us) that stupid. As well, we are well able to believe two objects
to be causally related even while our ideas relative to these objects do
not readily arise together.
It is also worth pointing out
that, intuitively, we have the volitional power (often if not always) to arouse
or suppress ideas, whereas we do not seem to have similar power relative to
apparent objects. We can ignore objects, or forget them, but that does not wipe
them out: if we look for them again they reappear or someone else might still
see them. But in the case of ideas, or more precisely many memories and derived
imaginations, we experience a greater power of manipulation. On this basis, we
expect the associations between ideas to be more tenuous: they depend more on
All such simple observations and
arguments again take us to the conclusion that Hume indulged in an excessively
hurried generalization, from very little introspection and reflection. He was
either lazy or dishonest, focusing on the data that supported his pet theory and
ignoring the data and reasoning that contradicted it. The matter is open to
objective judgment – it is not my word against his: everyone can carefully
consider the data and judge independently.
The philosophical sciences of
logic, phenomenology, epistemology and ontology provide the blueprint and
guidelines for induction. There is of course additionally the need to consider
the psychology of induction, since after all induction is an activity of the
human psyche. Through such a complementary study, we can better comprehend how
induction actually occurs. But psychology and logic are two very different
Briefly put, I would describe
the psychology of induction as follows. The human soul has powers of cognition,
volition and valuation. All three of these functions come into play in every
inductive act. The end is cognitive; the means is volitional (combined with
non-volitional elements, provided by the nervous system, mainly the brain); the
motivation comes from the valuing of knowledge, or the things or events that
knowledge can serve as a means to.
The relation between the said
philosophical sciences (including logic) and the psychology of induction (in an
individual at a given time) is that the sciences (to the extent that they are
known to the person concerned and kept in mind) influence the inductive
activity of the person. They do not determine it, note well, but they influence
it. This relationship thus leaves room for the cognitive, volitional and
value-oriented factors of induction.
If the person has a low degree
of knowledge or understanding of the scientific underpinnings of induction, he
or she will naturally often make errors. However, even without formal training
and reflection on the issues of induction, most people do subconsciously
frequently think logically and thus a lot of the time have some measure of
success in their inductions. Humans, after all, have considerable natural
intelligence; else they would not have survived till now. The said sciences are,
after all, very recent productions of the human mind.
The root of Hume’s problem
with induction is perhaps his misconception as to what ideas
are. I suggest that in his mind’s eye, ideas are clouds of ‘mental stuff’
produced by sensation. These perhaps very often look like the objects that
generated our sensations, but we cannot be sure of that since we have no access
to such objects other than through ideas. Thus, what we actually perceive and
know are only ideas. Thus, ideas are veils that separate us from reality, rather
than conduits to reality.
This view is, as already pointed
out, self-defeating, since it accuses also itself of ignorance and error.
However, the point I want to stress here is how ideas are reified in
Hume’s discourse. Because he effectively visualized ideas as atoms of mental
substance, his view of human knowledge as a whole was completely distorted.
In fact, an idea is something
very abstract, an intention
towards some object, a relation of pointing in a certain direction,
directing our attention hither, rather than a substantial entity. An idea is an
idea of an object. It has no existence apart from an object of some sort
(although, of course, the object concerned need not be real, but may be
It is certainly true that the
physical processes of sensation play a central role in our noetic relation to a
domain beyond our apparent physical body. But it does not follow that
what we perceive when we sense this ‘external world’ are sensations or even
of the world.
The only coherent theory is that what we perceive is the
The images we form in our minds of such primary perceptions are
only ex post facto memories of what we perceived.
The abstract concepts we form thereafter are not mere
manipulations of concrete memories, but relations we intend to the objects
The fact that we perceive
external objects, and not impressions or ideas of those objects, is certainly
marvelous, so much so that we still cannot understand how that might happen. But
our difficulty and failure to explain this marvel of nature is not a reason
enough to deny its occurrence. That we perceive the world is obvious enough; how
such a thing is possible is a distinct question, which we may never answer.
Science does not normally deny the very existence of what it cannot thus far
Note well, we can claim
knowledge that we directly perceive the external world itself, without
claiming to know yet just how we manage to do so. We know we can, because
this is the only consistent theory we can posit, as already explained. But
exactly what role the senses and brain play (other than memory production,
storage and reactivation) in this evident direct perception is still an open
question. The fact that a partial question remains does not invalidate the truth
of the partial answer already obtained. There are many issues in the special
sciences that remain unsolved to date – and we do not for that reason throw
out the knowledge we already have.
It does not follow from such
non-skeptical, objectivist theory of knowledge that perception or conception can
never be erroneous. Errors in human knowledge are essentially conceptual, and it
is the task of logic to minimize them. Perception sometimes seems wrong, after
the fact, due to our noticing later percepts that seem to contradict the
earlier. In such cases, we realize that in fact we drew some conceptual
inference from the initial percepts, which the later percepts make clear was
unjustified, and we correct our previous assumption. This is just an application
of the laws of thought and the principle of induction to sorting out conflicting
Once we comprehend human
knowledge in this truly enlightened manner, it becomes clear why Hume was so
confused and self-contradictory in his views of induction, and other logical and
philosophical issues. If one starts with false premises, one is very likely to
end up with false conclusions. He should have been more careful.
Philosophers like Hume have
always found the idea that we might indeed be perceiving and conceiving the
world out there, and not merely our impressions and ideas of it, difficult to
comprehend or explain. This is understandable, because this seeming ability of
ours (viz. external consciousness) is something truly surprising and, well, miraculous
– no better word for it comes to mind.
But then these same philosophers
take for granted that our inner perceptions and conceptions are valid and not in
need of explanation. They apparently do not realize that this ability (viz.
internal consciousness) is also miraculous – indeed, just as miraculous. For
the difference between the two, after all, is just one of distance. And who is
to say how big the soul (the subject of consciousness) is or where it is in fact
located? Why do they assume that it is more ‘inside’ than ‘outside’ the
In both cases, there is
something marvelous, inexplicable – namely consciousness, a line of relation
between an object and a subject. How can one existent (a soul, a spiritual
entity) experience another (a mental or material phenomenon)? In the case of
self-intuition, the subject and object are exceptionally one and the same. But
even this is a marvelous event, that something can experience itself.
The mere fact of consciousness
is the biggest mystery. In comparison to it, the issues of how far consciousness
can go, and how in some instances it is aroused and made possible by sensation
and yet the body does not block or distort our view – these are relatively
Of course, a theory of the exact role of the senses remains highly desirable.
Obviously, each sense organ (whether in humans or other animals) somehow gives
the overall organism ‘access to’ a range of data of a specific sort, and no
other: e.g. human eyes open the window to a range of light waves (the visible
spectrum) but not to all frequencies (not to radio waves, ultraviolet rays or
microwaves, for instances) and not to other modalities (such as sound or
chemical signals). The different sense organs have evolved over millions of
years (at different rates and in different directions in different organisms).
Without these sense organs, we would not (so it seems) be able to sense external
reality. So their role is not only that of memory production, but they are
somehow essential to the actual contact between the organism as Subject and
material objects it perceives. Even so, to repeat, it cannot consistently be
affirmed that what the Subject perceives are internal products of sensation. Nor
is the explanation that sense organs serve to filter out some of external
reality sufficient. The sense organs must have a more significant role in the
Subject-Object interface. But what?
I am stereotyping things a bit, because in truth Locke was somewhat
aware of the problem, and so was Berkeley after him (and before Hume).
Perhaps the philosopher most to blame should be Descartes. But I cannot here
get into the fine details of history.
This has come much later in the history of philosophy. Even Immanuel
Kant, Hume’s intellectual successor, never grasped phenomenology, but
instead produced a complicated system of philosophy that increased the
appearance-reality chasm. Note that when I use this term, I do not
necessarily mean the Hegelian or Husserlian attempts at phenomenology,
though these two later philosophers certainly played important roles.
The reader is referred to my work on phenomenology for more on this
Hume obviously in fact believed in the existence of the external
world, since he invested so much of his time writing and publishing books
for others to read!
Whereas Locke used the word idea very generally (including all mental
phenomena, even emotions), Hume distinguishes primary impressions from
derivative ideas, i.e. simple empirical sensations from the more complex
mental constructs made with them. However, I here use the term idea much
like Locke, because Hume’s finer distinction does not affect the issue at
The word “intention” is very well chosen here, note well. It is
not the idea, or the name for it, that intends the object – it is us, we
the subject, who do. The word does not refer primarily to an act of
consciousness, in the sense that Husserl defined consciousness with
reference to some mysterious “intentionality”. Consciousness is not
essentially an action, but rather a receptive event. No, intending refers to
an act of volition. The subject (I, you) programs such an intention
into every notion or symbol he produces. The subject wills his attention
(awareness, consciousness) in the direction of the object concerned when he
again comes across that idea or word. When we communicate, we pass such
guides to mental action to each other.
A verbal problem to always keep in mind is the equivocation of the
word “sensations”: used in a general sense it refers to all sensory
material, whereas more specifically it makes us think of touch sensations.
Likewise, the word “images” tends to evoke visual images, but in the
present context it is meant to refer to any resemblance, i.e. equally to
auditory and other sensory phenomena. Such equivocations may seem anodyne,
but they mislead many people.
More precisely, memories are physical items (produced
by sensations of visual and auditory phenomena) stored in the brain, which,
when (voluntarily or involuntarily) reactivated, project mental
images or sounds that we inwardly perceive and recognize as previously
directly perceived (in the physical world, when that is the case). In the
case of smells, tastes or tactile phenomena, I suspect we cannot in this way
‘recall’ past or present perceptions, but only ‘recognize’ them as
familiar, so the term memory has a slightly different meaning. Note well
that we do not commonly confuse our perceptions of material things and
events with our memories of such perceptions; it is only armchair
philosophers like Locke and Hume who equate these two experiences, quite
For example, just what is a “force” like gravity in physics? Or
just what is “energy”? Isaac Newton admitted his ignorance, saying “hypotheses
non fingo” (meaning, I have no explanation); and even after modern
developments in physics, like the Relativity and Quantum theories, we still
do not know just ‘what’ these abstractions refer to concretely or
‘why’ these processes occur. Despite this partial (and even crucial)
ignorance, we do not consider physics less of a science. For what is
science? It is not omniscience, but merely a guarantee that our current
opinions are the best possible in the current context of experience –
because the most rigorously induced.
I leave open whether we can experience other souls. Some
people suggest it is possible, i.e. claim a sort of other-intuition. Some
people claim even to have experienced God.