2. We now need to consider that `initial symptom', the registering of
which is the basic cognitive process. Let us call this perceiving
the appearance per-se.
a. Think again of the new-born infant's experience - or of the
kaleidoscopic pattern of colour or, for that matter, of a
melody simply as heard. New-born infants do not recognise
melodies, or even recognise that they are melodies - but they
hear the same patterns of sound as we do. We want to say `that
particular structure of sensible qualities which just exists as
that pattern' or to say `that particular structure of
sense-experiences which just occurs in that pattern' and we
do, in effect, say (or assume) both though, as we shall see,
moving from the latter to the former is the most significant
`step' we take in the perceptual process.
b. For the present, however, let us allow that the term
`appearance' somehow covers both and consider what we are
distinguishing from what when we use the term `appearance' in
everyday speech. We say things like `He is really a very nice
chap, though his appearance is against him' - meaning that his
look, sound, smell, etc, would not, for most people, be conjointly a symptom of a `very nice chap', so that people are
surprised to discover that he is a nice chap. This is O.K. -
but how do they discover this except from (further) appearances
(of him)?
Compare -
i) `It appears to be getting fine again' - which expresses a
belief that it is highly probable that the clouds will
drift off and the sun will shine - with
ii) `You should not judge by appearances' - which cautions
people against `judging the sausage by its skin'.
The first seems to imply that appearances are what we judge by;
the second that we oughtn't to!
Consider also: `Life is real, life is earnest and things are
not what they seem'. We frequently use `seem' and `appear'
interchangeably - but with a certain ambivalence in both cases
about the relationship between appearing (or seeming) and being
so.
Let us be quite clear that there is a difference; to say that
X seems, or appears, to be Y is not to say that X is Y. But
neither is it to say that X is not Y.
The simple fact is that we must judge by appearances since
there is nothing else to judge by. Any situation, as a
phenomenon, is for each of us simply the totality of `its'
appearances - i.e. that complex of present and predicted
appearances. That set of appearances is both what counts as
that object and the only possible basis of awareness of that
object. So the distinction we make between what it is and how
it appears can only be either:
i) between our own phenomenal worlds and the `causal-to-
phenomenal' objective reality - or
ii) (within the phenomenal) between our first inferences and
our subsequent, more cautious or `educated' inferences -
an acknowledgement that we do make mistakes, do
misidentify, do misreason - that any belief could be
wrong. Here consider again 2.b. above. Had we said `He is really a nice chap
though first appearances are against him.' there would be no problem.
c. But now try to see ( and this is a difficult point) how, in
terms of i above, ii above involves a misuse of `appear'. It
is really invoking again that `unnecessary middle-man', the
phenomenal world. For if there is error, it is about objective
reality. Within my phenomenal world what something is, just
is, how it appears to me.
The error, therefore (when error occurs) is not in the
appearance, it is in the inferences we make from the
appearance; it is not the `presently presented' which is wrong
or inappropriate, it is the further predicted experience, the
behaviourally manifested expectations which are the belief
about the state of affairs. And the only way that we ever can
discover these to be wrong is by perceiving further appearances
which are different from those we had expected to perceive.
Some years ago one Australian police-force thought it would be
a good way to cut back speeding on major highways to place
life-size, plywood, cut-outs of policemen on motorbikes just
projecting out from behind bushes. For a while it was quite
effective; approaching motorists judged from appearances that
they were being watched. Of course, as they came alongside,
they realised, from further different appearances, that they
had been fooled (and no doubt sped up again). But there was
nothing wrong with the original presented appearance; a
photograph would have shown exactly the same pattern of
colour-discriminations as the motorist registered and, under
the circumstances, the inference was a perfectly natural
(though as it happens, wrong) one.
Suppose somebody (who does know the difference between rabbits
and hares) takes a hare to be a rabbit - a simple mistake. This
can make sense only on the assumption that (by his own rules)
it really is a hare. And this can only be that it does present
`hare appearances' which he has (at the `reasoning level')
mistaken for `rabbit appearances'. If it really had,
consistently and continuously, presented rabbit appearances, then, for him
it would be a rabbit. Always some of the appearances (which count for us as being that situation
) are
presented and some (most by far!) are merely guessed. Error
occurs when we guess too much from too little.
d. But, let's go back to the plywood policeman. In relating the
incident, the motorist might say `I wasn't thinking about my
speed and (what I took to be) the sudden appearance of a police patrolman gave me
quite a start'.
This, when one thinks about it, has to be a slightly different
use of `appearance'. An appearance (as we have been using the
term) could not be sudden - any more than a picture or a melody
could be sudden. The only things that can be sudden are events
- changes, movements, actions. So, here we are not talking of
the appearance as an entity; we are talking of the appearing
(to somebody) as an event. The event is not the coming into existence of the policeman, or even of the `police-appearance';
they were there all the time. The event is the appearing to
the motorist of that pattern of sense-stuff which is for him a
symptom of a `look out - police' situation.
Now, reconsider the distinction suggested in a. above between
`a structure of sensible qualities which just exists' and `a
structure of sensory experience which just occurs'. It should
now be possible, and useful, to think of this distinction as
that between appearances and appearings. We can then see the
relationship between the `parallel aspects' of the basic
cognitive activity:
i) the appearing - experience per-se - how a diversity of
sense-stuff (a structure) manifests in any sensory mode or
modes - and
ii) the appearance - assumed to be presented by, to be the
presented aspect of, whatever part of objective reality
we, thereby, assume ourselves to be perceiving.
It should be seen that, for grown-up people, the same set of
circumstances/events counts as both of these but that,
nevertheless, i) does not imply any objective reality whereas ii)
does imply an objective reality. Thus, to accept our appearing
as an appearance is to accept an external world independent of
ourselves.
d. It is only such an acceptance that gives any rationale to the
notion of seeming (and which explains why `seems' and `appears'
are used so interchangeably). To say `this seems to be X'; is,
in effect, to say `This is an X-type appearing - and I
therefore must assume that it is the appearance presented by an
X situation though, as with all inferences (beliefs), I could
be wrong'.
But the vital point is that the appearing, the `unclassified
classifiable', the newborn-infant-type experience, is what it
is or, perhaps, we should say how it is: and is (at its own level) perceived as
such, whether
subsequent judgements are right or wrong. Errors about the
what it is cannot alter the how it presents.
3. We are now well placed to consider what might be called `stages of
perception'.
a. We have already established, pretty conclusively, that sensing
as such, is not a cognitive performance, though it is essential
to any cognitive performance and its occurrence can only be
within a cognitive performance. If this still presents any
problem, consider a very simple analogy: To speak of the
colour of my shirt is not to speak of my shirt. But that
(instance of that) colour cannot exist without that shirt - nor
can that shirt exist without being some colour.
But, we have also made the point that, whilst sophisticated
perception `goes beyond the given', there must be awareness of
the given (as such) for this to be a sign for whatever it is we infer.
And such awareness can arise only from perception. We must
therefore distinguish between:
i) perceiving the given - i.e. being aware of how we are sensorily
experiencing, of the structuring of the
sense-stuff - and
ii) perceiving that some state of affairs, independent of
ourselves, pertains - i.e. recognising the `stage 1'
perceiving as presented by a symptom (for us) of some
customary complex, some familiar state of affairs in the
external world. It should be noted that this recognition
can be at any level of vagueness/precision. I could, for
instance, take my presently perceived `appearings' as a
symptom of a sitting-reading situation or as a
revising-philosophy-notes-for-internet-publication
situation. To accept the latter is to imply the former
but not vice-versa.
Think again of the kaleidoscope. `Stage I' is the perceptual
awareness of that particular multi-colour-structure experience
(per-se) we are having; Stage `II' is seeing this as a pattern
of colours presented to me by little pieces of glass contained
in a tube into which I am looking. To see it as such is to
predict certain specified changes in experience under certain
specified circumstances - e.g. that if I turn the tube the
colour pattern will change (the pieces of glass will shift) and
that if I remove the tube from my eye the whole experience will
cease. This is all that `going beyond the given' amounts to -
taking for granted certain dispositional characteristics of the
external world on the strength of occurrently presented
characteristics. Or, perhaps we should say `on the strength of
occurrent experience' since the very assumption of a presentation is already `going beyond the given'.
b. Here, however, a rather uncomfortable point needs to be made.
When we looked at the sensing/perceiving relationship we
recognised a total interdependence; there could, therefore, be
no question of any chronological ordering - for sensing to be
occurring, perceiving must be occurring. When we look,
however, at `Stage I' and `Stage II' perceiving there is
plainly a logical sequence; the one rises out of the other.
And, though we are never conscious of any time-lapse between,
say, `registering a particular look' and observing that a blue
pen is on the desk, there must be a chronological sequence.
There is an obvious sense in which `Stage I' must be achieved
in order for `Stage II' to commence.
We might expect, therefore, that there would be some situations
within our experience where we just happen to stop at Stage I -
perceive the `how' without considering the `what'. Do we ever
do this?
The problem here is that, being somewhat sight-dominated, we
tend generally to think of visual examples and it is hard to
imagine `registering a look' without some inference, however
vague, about what it is the look of. Here it is interesting to
compare a child's painting of, say, trees on a hillside, with
that of a skilled naturalist painter. The child knows that
grass and leaves are green, tree trunks are brown, sky is blue
and so on, and produces the painting to that formula without
much reference to the `actual look' as experienced. The skill
of the painter (not easy to acquire) is in unthinking, putting
out of mind, any preconceptions about what that bit of the
external world `should' present to visual experience - and
reproducing the `look' as it actually is experienced. It is
doubtful, however, whether even the most dedicated painter ever
totally loses his awareness that this is (the look of) a wooded
hillside or whatever.
But when we turn to other senses, `stopping at Stage I' seems
rather more likely. Surely people do occasionally register
sounds and smells and tastes and feels just as auditory,
tactile or whatever experience without any consideration at all
of what they are the sounds, smells or feels of.
Remember, there is an obvious sense in which the appearing
per-se, the `unclassified classifiable', is what is `captured
by' a camera or a tape-recorder. And, whereas it is just as
hard to look at a photograph without assumptions of what it is
a photograph of as it is to look at a view without assumptions
of what it is a view of, it is not so odd to hear recorded
sounds just as a pattern of auditory experience.
c. We may seem here to be wandering a bit from analytical rigour
into experience-based musings. But, having asserted
analytically that there are two conceptually distinct stages
of perception - `of the given' and `beyond the given', and
having allowed a continuous process but insisted upon some
chronological sequence, it is beholden upon us to square with
experience the point of departure, as it were, from Stage
I. The point has been made (1. d. above) that our inferences
operate in a chain-reaction way; each situation recognised
(from a symptom of it) becomes itself a symptom of a more
complex (i.e. more precise, sub-class) situation (albeit with
diminishing probability asserted as the precision increases).
Let's think of this `chain process' in the observation of a
thunderstorm. Consider:
i) Jones has a sensory experience of the kind we describe as `seeing a flash of
light' -
ii) He identifies it as (made by) an electrical flash in the
sky -
iii) He anticipates hearing thunder - i.e. identifies it as a
thunder-storm situation.
iv) He thinks it highly likely that rain will find the leaking
spot in his roof and that his dog will be a nuisance.
v) He is prepared for a pretty gloomy evening - etc
Here it is worth noting that what we have been calling
sophistication just is this auto-responsive symptom-complex
chain-reasoning. Indeed, psychologists have great fun
discussing the nature of intelligence but, philosophically, to
say that Jones is more intelligent than Brown is simply to say
that Jones has a capacity for making more extensive successful
inferences from the `data' (from his own sensory experience
per-se) than Brown has.
The important point here, however, is that Stages ii, iii, iv
and v - and any further stages - are different only in degree
of complexity. It is the move from i to ii which is crucial -
from perceiving a change pattern within experience per-se to
assumptions, however modest, about the external causes of
that change pattern. It is significant that at Stage ii we
know precisely what would count as Jones being in error. The
flash might, for instance, have been a car headlight
momentarily reflected by a low cloud. But this would in no way
alter Stage I as such; as experience, experiencing a flash
just is experiencing a flash. We might, then, say that the
point of departure is the point at which error about the
world becomes possible because some (however limited) knowledge
about the world is being assumed.
4. We should now look more closely at Stage ii, the formation of
beliefs about objective reality.
a. Firstly - we do not perceive tomorrow's events; we predict them - and
we do not perceive yesterday's events; we recall them.
Nevertheless we do perceive (some part of) today's events - and
that perceiving of them involves (along with the `Stage 1'
apprehension of the `given') both recalling and predicting. My
perceiving that this is a pen is my predicting, from my present
sensory experience, certain further experiences of a `what-counts-for-me-as-a-pen' kin
d - and my making this prediction is
my recognising that present sensory experience as (the symptoms
of) the customary complex `penness' - i.e. recalling that, in
my experience, situations which present these (among other)
visual appearances are pen-situations.
Identification rests upon conditioning - and conditioning is
possible only because there is recollection.
It can be said, then, that, whereas Stage 1 could be seen
simply as the apprehension of the appearance (or, more
properly, the appearing), Stage 11 is the prediction of
further appearances (seen as the appearances of certain states
of affairs), from that present appearance, in the light of
past, recalled, sequences of appearances.
Here consider a situation familiar to most of us, approaching a
traffic light when it is changing. In Stage 1 terms we could
describe this as visually perceiving greenly/roundly (in a
structured, multicolour background) and then
visually perceiving yellowly/roundly. In Stage 11 terms we
could describe it as seeing a green circle replaced by a yellow
circle, taking this to be a changing traffic light and,
thereby, from prior experience, expecting a further change to a
red circle. We might say that, in Stage 11 we treat Stage 1
as evidence of a familiar state of affairs.
b. The traffic-light example, simple though it is, draws attention
to a complication which is overlooked in characterising Stage 1
as simply the apprehension of the appearance per-se. We are
in fact dealing with two quite distinct movements beyond this
mere apprehension -
The relationship between these two `movements beyond the given'
will be of crucial importance when we go on to consider the
perception of physical objects in space, and should be clearly
grasped at this point.
Try to imagine a being which is conscious (and intelligent) but
not self-conscious - which cannot therefore, think in external
world terms at all. This would in no way preclude that being
from being conditioned into expectancies (of further experience
per-se) on having any identifiable experience. Within the
example, on perceiving greenly/roundly it finds itself
expecting to perceive yellowly/roundly and then redly/roundly -
all within the experience per-se framework.
And, for every stage of experience per-se, every actual or
anticipated appearing, there is (or would be if he did think in
external-world terms) a corresponding actually presented or
anticipated aspect of the state of affairs. Thus we might say
that, for us who do think in external-world terms, there are
two parallel perceptions - of present and anticipated
appearings, and of present and anticipated experiences of the
world - with, at every stage, the first being all that can
count for us as the second.
In philosophy, schematic diagrams can be dangerous but, on this
point, it just might be useful to look at the `perceiving a
thunderstorm starting' situation (considered in 3. e. above) in
these terms:
experience per se
appearings which counts for
us as
the experienced
(presented appearances)
|
a sudden visual change'' '' seeing a flash of
lightning in the sky
creating expectation
of '' ''creating expectaion
of
an auditory change of
a particular, predicted
kind
'' ''
hearing thunder
jointly creating
expectation of '' ''
jointlycreating
expectaion of
a tactile change of
a particular predicted
kind'' ''getting drenched by
rain
| | | | | | | | | | | | | | | | |
and so on
This should hammer home the, never-to-be-overlooked, point
that, no matter how extensive or complex the situation
perceived to be so may be, all that can count for us as that
situation being so is an appropriate set of actual
sensory experience. We are not saying that the experience is
the external world situation, that would be plainly silly. We
are saying that the experience is the only possible indication
of what that external-world situation is. We cannot, therefore
attribute to that external-world situation anything (other than
objectivity itself) which is not represented by actual present
or anticipated experience per-se.
c. Here again, the limitations of language cause some problems. We
are talking of perceiving, whether we perceive states of affairs
or merely perceive appearances and we have only the one set of
words for the modes of perceiving - `see', `hear', `feel',
`taste', etc. This gives rise to some strange-sounding
questions such as `Should I here (in a given context) speak of
tasting a pineapple, or of tasting the flavour of a pineapple -
or, perhaps, of tasting that flavour (which I take to be
pineapple)?', `Do I hear a bell, or do I hear a sound which I
judge to be made by a bell?', `Are there circumstances in which
I can see the ``look'' presented by a doorway without seeing a
doorway?'. It would not be profitable to try to answer these
question; it is sufficient to realise when, and why it might
seem necessary to ask them.
d. So the terms used indicate only the sensory mode of the
perception, whether it be of the state of affairs or merely of
the presented appearance. Normal usage, however, always
assumes the state of affairs - that what we perceive,
irrespective of what (how much of it) is actually presented,
is whatever counts for us as that state of affairs. As pointed
out several times, to perceive that the train is coming is to
perceive that all the circumstances which count for us as that
situation do pertain, whether the mode of perceiving be seeing a train-look,
hearing a train-sound, smelling a train-stench or just feeling the ground
tremble slightly; these are all the symptoms (for us) of the totality.
We should here also note two additional points. Most of our
perceptions are in two or more sensory modes simultaneously;
unlike our blind man and deaf man, most of us both see and hear
the train coming. And perceptions are not discrete, one at a
time, things; all of our senses are active and we are
simultaneously aware of a broad spectrum of states of affairs
at different levels of precision/vagueness. There can no more
be a rule for what is one perception than for what is one event
perceived (or for what is one fact - refer back
to Paper 6). Insofar as anyone isolates a particular belief,
based upon current sensory experience, he, ipso-facto, isolates a
particular perception.
Here it is worth considering the popular expression `seeing is
believing'. As that exrpession is generally used, it signifies
that we cannot but believe that something is so if we `see it
with our own eyes'. It is, however, quite literally true:
seeing is (one kind of) believing since it is one kind of
perceiving - and we cannot be said to have perceived that X is
so if we do not, thereby, believe that X is so. Here note also
that believing is expecting, predicting - and what is predicted
(at the experiential level) in perceiving is, as outlined
above, further sensory experience, taken as experience of the
external world.
d. We must be careful, however, not to make the logical gaffe of
assuming that, because perceiving is believing, all believing
is perceiving. Jones can believe that he was born in Wales
without anything we would regard as occurrent perception being
involved. And he can believe that it will be hot next
Christmas without any obvious occurrent perception being
involved.
It would, however, be quite wrong to assume that perceiving is
limited to what is going on at the time of the perception in
question. We can look out of the window after a night's sleep
and perceive that there was a storm during the night and there
will be problems getting the car started. What is presented
to us is a very wet road surface, plants blown over and
rubbish strewn around - and water dripping out of the engine
area of the car. There is always a predictive element in
perception of states of affairs and, in theory, no limit to how
far that prediction can extend in the chain-reaction way
considered in 1. d). So long as the belief that the state of
affairs pertains arises from the (Stage I) awareness of a
presently presented appearance, then that state of affairs is
being perceived. We might say that we are presented with
evidence and, thereby, perceive that which it is evidence of.
It is, therefore, a quite legitimate use of `perceive' to say
`I perceive clearly that, unless there is a change of
Government in Burma, there will be a bloodbath there
within the next ten years' - provided only that this belief
does stem from some present sense-perceptual experience - such
as hearing a radio-broadcast or seeing reports in a newspaper.
Note, however, that this in no way destroys the distinction
which has been made between knowing-by-perceiving and
knowing-by-remembering. There is a quite clear difference
between perceiving (from the debris) that there was a storm the
night before (which we were not at the time aware of) and
remembering that there was a storm the night before (which we
were aware of) whilst we are in fact not looking at anything at
all to do with that storm.
It is perhaps less clear that the distinction can be maintained
between perceiving and predicting. Unlike the remembered, the
predicted has not yet happened (been experienced) and it is
hard, therefore, to see on what grounds any prediction could be
made without reference to the presently perceived situation.
In principle it seems possible to be, say, sitting eating
breakfast and, out of the blue as it were, predict that there
will be a storm in the afternoon - but it might seem improbable
that such a prediction should be made or, if made, that it
should prove to be correct. We could, however, whilst eating
breakfast recall that yesterday was very humid and also that,
at this time of year, humid days are generally followed by
stormy days and, putting these ideas together, make the
prediction. This is plainly different from the situation
where, during breakfast, we observe the weather outside and,
being somewhat expert in meteorological affairs, predict from
that (i.e. perceive that) there will be a storm in the
afternoon.
e. The point has been made that perceiving is believing. But the
question was left open whether it is necessarily knowing -
whether it counts as perceiving only if we get it right.
Normal usage does suggest that it does; we use such
expressions as `misperceive' and `seem to perceive' where we
believe that an error has been, or might have been, made. So,
let's allow that `Jones perceives that there is a tree in the
yard' does imply that there is a tree in the yard. It is not
the case, of course, that `Jones knows there is a tree in the
yard' implies `Jones perceives that there is a tree in the
yard'. He may have observed it, or been told about it,
yesterday and remember that it is there.
So, we can allow that seeing, hearing, smelling, etc. are kinds
of knowing - and that what counts as Jones seeing a tree in the
yard is -
i) Jones believing that there is a tree in the yard -
ii) His so believing as a direct result of present, visual,
sense-experience -
iii) There actually being a tree in the yard.
But the situation can arise, and often does, where requirements
i and ii are met but iii is not. This is what we call
perceptual error; as a result of the perceptual process Jones
believes to be the case something which is not the case. The
thing that must be noted clearly is that Jones simply is not
seeing a tree; it is not the case that he is seeing a seeming
tree. There are no seeming trees in the world.
5. Finally we should look at the two kinds of perceptual error,
related, as might be expected, to the two stages of perception.
a. We allow that perceiving is knowing - but both perceiving and
mis-perceiving are believing. Without belief (being certain)
there can be neither knowledge nor error. And errors are
always about states of affairs; to make an error is to believe
that a state of affairs pertains when it does not pertain.
Now, when Jones seems to be perceiving that e.g. a blue pen is
on his desk, he is believing two things -
i) that a particular appearance is being presented to him -
ii) that the `presenter' of this appearance is the state of
affairs he classifies as a blue pen on his desk.
And in either of these beliefs, he could be mistaken. In one
case he could actually have the sensory-type experience (as of
a pen on a desk) but without there being a pen-on-desk
situation (or any other external-world situation) presenting
it. This we call hallucination.
In the other case he could actually have the sensory experience
presented by a state of affairs - but not by the pen-on-desk
state of affairs he takes it to be. This we call illusion -
taking one state of affairs for another, different state of
affairs.
All perceptual error is of one or the other of these kinds. In
theory it could be of both kinds at once - Jones could have
what seemed to be a perceptual experience but was in fact an
hallucination - and misinterpret the `data', take what is in
fact a dog-type appearing to be a cat-type appearance - i.e.
believe himself to be looking at a cat. However, whilst this
is theoretically possible, nothing could establish that it had
occurred - other than Jones subsequent recollection (in
imaging) of the appearing itself and realisation of his own
misidentification.
The more important point to note is that, although the two
kinds of perceptual error are `related to' the two `stages of
perception', they both occur at Stage II. Hallucination is not
an error about the sensory-type-experience per-se. That just
is whatever it is. The error is in assuming it to be caused by
external reality and, as soon as external reality is invoked at
all, we have already gone beyond the given - made predictions
about experiences not presently occurring.
b. This raises the intriguing, yet somewhat sterile, question of
whether, and if so how, error could occur at Stage I - the
apprehension of the appearing as such. Since we are dealing
with perception, we are invoking belief - and any belief must
be propositional and could be wrong. What we are perceiving is
not simply the sense-stuff; it is that the sense-stuff
variations present a particular structure. Since they may or
may not present that particular structure [here think again of
the kaleidoscope just as experience per-se], it is possible for
error to occur - though nothing at all could make this error
manifest. We could not both believe that structure to be of a
certain nature and observe that we were wrong! This is why
such enquiries are somewhat sterile - but they do at least put
paid to any quest for infallibility at some level of
awareness. As soon as it is taken seriously that there can be
no cognition less than perception and that perception implies
belief, it should be clear that the notion of infallible
awareness is necessarily nonsense.
c. To return to hallucination - plainly it can occur in any
combination of the sense-modes. Within whatever sense-modes
are involved, the experience is as if appearances were being
presented by external reality when in fact they are not. As we
shall see later, imaging is the memory-activity of recalling
sensory-experience per-se and we might, therefore, characterise
hallucination as image experience presenting as occurrent
sensory experience - so long as this does not confuse us about
the nature of image-experience.
When we say that it could occur in any combination of
sense-modes, we should perhaps add `short of all sense-modes
simultaneously'. Dreaming is not hallucinating - though
something very similar is obviously going on. In dreaming we
are totally detached, sensorily, from external reality. It
seems to be characteristic of hallucination that there is at
least some sensory `anchor' to external reality which renders
the seeming-sensory-experience in other modes plainly
discrepant. When Macbeth `saw' his dagger floating in the air
before him, he found that he could not grasp it - his tactile
sense was still functioning normally. Even if he had also had
an experience as of grasping it, he would still have been
aware that he was in the castle banquet hall. Drunkards
suffering delirium tremens may believe that they see and feel
creepy-crawlies of various kinds but they are still able to
tell this to real people in the real world.
It should be noted, of course, that to hallucinate is not
necessarily to make an error. A sufferer from an illness might
be warned that certain drugs prescribed could be hallucinatory
- but `not to worry'. When he or she subsequently `saw' red
insects crawling over his or her arms, the inclination to
attempt to brush them off would probably be very strong - but
there would not be a mistaken belief that there were really red
insects there - any more than there is a belief in the audience
that the conjuror is really making a rabbit materialise from
thin air.
It should also be noted that hallucination becomes apparent as
hallucination only when, either some part of an experience is
discrepant with some other part, as with Macbeth's dagger, or
what we seem to be experiencing does not square with what other
people are, or claim to be, experiencing. A solitary man, a
Robinson Crusoe, could have any number of hallucinations
without ever realising it. He might think that strange things
were happening if, for instance, trees and animals appeared
and disappeared in unexplained ways. But what are, for each
of us, at the time, strange things happen throughout our
lives - that is how our conditioning is constantly adjusted.
When you consider these last two examples, you will realise
that the very notion of hallucination is intelligible only
because we all do make a distinction between our phenomenal
worlds and objective reality. Without such a distinction, the
experience itself would be reality, irrespective of how odd
it might seem to us.
d. Turn now to illusion - the situation where the appearance is
indeed presented (quite properly) by external reality but, for
any number of reasons, is mis-identified. This can be as
simple and common a situation as believing that we see Jones
across the street when in fact it is Brown or as complex as
seeming to see the conjuror (illusionist!) sawing somebody in
half and joining them together again - where no amount of
re-looking at the tele-film record stops us from seeming to see
this strange sequence of events. In both cases exactly the
same thing is happening - we are mis-judging the state of
affairs because we do not have enough, or detailed enough,
experience of that customary complex to identify it correctly
at the appropriate sub-complex level from the symptom in
question - the presented appearance.
The example should, however, draw attention to an important
distinction between
i) appearing to be - the man across the street appeared (to
us) to be Jones - and
ii) appearing as if - we do not in fact believe that the man
on the stage has sawn the lady in two because we do
recognise this as a very special sub-class complex - a
conjuring trick - and, by that recognition, what we see is
just what we would expect to see. Our small grandchildren, however, who have no experience of this
sub-class, are quite disturbed; they do believe that a
lady has been sawn in half - at least until she leaps up
and takes a bow.
When this difference (between appearing to be and appearing as
if) is recognised, it should be realised that illusion is not
some special class of perceptual error; it is merely a
possible source of perceptual error at the inference level.
Misunderstandings have arisen because people use the term
`illusion' in two rather different ways - e.g.
i) They are under the illusion that he was angry (when he was
not) because his face was flushed by fever -
ii) It is just an optical illusion that the moon looks (as if
it were) bigger on the horiizon than high in the sky.
The first of these is an explanation of an error that somebody
has made; the second is merely noting an apparent oddity of
presented appearance - which, nevertheless is `perfectly
natural' (i.e. what we would expect) within its own special
context.
People sometimes give as examples of illusion such things as
round plates `looking eliptical' when they are viewed from the
side or sticks `looking bent' when they are half-submerged in
water. But it would be quite extraordinary if a round plate
did not `present an elliptical appearance' when viewed from the
side - and only very young children would expect the stick to
be bent when it is taken from the water. Contrast this with
well known optical illusions like 
where, even when we
know the lines are equal, they still look different (here
compare the induced hallucination) and tactile illusions like
drawing crossed fingers down our own noses and `seeming to
feel' two noses side by side.
e. It should now be plain that illusion is simply a discrepancy
between the state of affairs suggested (to somebody) by a
presented appearance and the state of affairs actually
presenting that appearance. If that which is suggested is
accepted, error occurs. If, from prior experience, it is known
that such seemingly discrepant appearances are presented in a
particular sub-class of situations (like mirages which make
distant ships look quite close - or even `present them' upside
down - in particular weather conditions) then, although the
feeling of oddity remains, no error is made about the state of
affairs in external reality.
So - to summarise - neither hallucination nor illusion
necessarily involves perceptual error - but all perceptual
error (wrong belief about what is presently being presented to
us) arises either from hallucination or from illusion.