If you were asked what the shape on the left was the most likely answer was a number three.
But place exactly the same figure in a different context - among letters - then it can be read as a B, even though it does not have the upright bar expected with the letter B. This is what occurs if its is part of Titbits.
Sometime in 1968 the Titbits banner heading was changed to that on the right, presumably because someone liked their titbits to be complete. This example indicates that the world is not just simply what it appears to be. What is 'seen' is constructed and represents an accommodation between what is known and the information received through the visual sense. A few examples from classic psychology experiments are shown below to illustrate that we take into account the context when we 'see' information and that it is the context that 'determines' what is 'seen'.
In the first example there is a character rather like the B in the original Titbits title. But when one reads across, rather than down, instead of being 'seen' as a letter it is read as the number 13 because of the context it is in. Similarly the example with what is normally read as H in the word THE contains a character that could just as easily be read as an A if the surrounding letters were C and T as in the word CAT.
The following sections take this type of explanation of visual perception a stage further to demonstrate how and why semi-subliminal elements of ads are often overlooked and yet are subject to the same rules of perception as other images. They may thus possibly influence those who view them.
The visual system is extremely complex and it is impossible to do it justice on this site. Interested readers are recommended to view the article in the edition of the Geographic magazine containing the original of this small scale reproduction. Alternatively any psychology textbook or textbook on perception will provide stimulating reading. A number of these are listed on the Psychology Bibliography. The sections below can only give a brief indication of how we can be both fooled by visual information and how even ambiguous information can be interpreted meaningfully. There is also an indication of how such ambiguity, when presented in adverts, need not prevent automatic processes from extracting some information, even if this never reaches conscious awareness.
It is generally believed that visual illusions are simply curiosities but by studying illusions it is possible to come to an understanding of how normal vision works - and how it can be fooled by unexpected cues, such as those provided in many ads containing semi-subliminal material. And how knowledge of such rules can be put to use in constructing semi-subliminal and manipulative adverts.
Normally we do not expect illusions in dealing with the everyday world. We have learned that the world is 3 dimensional. We thus interpret all visual signals as if they came from a 3 dimensional object. It is this 'over learning' which allows us to look at pictures, photographs and paintings and 'see' them as if they were 3 dimensional objects, even though we know they are clearly only 2 dimensional representations.
In certain circumstances, the fact that we are presented with incomplete information, as is always the case with two dimensional pictures, leads to inappropriate conclusions. For example, in the figure on the right we can either perceive faces or a vase. Viewing the image normally means we can never 'see' both at the same time.
In the Kanizsa triangle illusion, presented on the Imagination page, it seems as though there is a brighter triangle superimposed on another triangle in the background. Despite appearances, there is, in fact, no difference in brightness levels between the central area of this image and the surrounding area. Similarly with the length of the lines in the illusion shown below. They seem to be different lengths but, as measurement will demonstrate, they are in fact identical in length. You can prove this to yourself by drawing the same shapes on any piece of paper.
This illusion occurs because we know that element in the real world that are nearer to us are larger than those that are further away. Thus, on the assumption that one line is closer than the other, the visual system 'decrees' that it must be smaller than what lies further away. We thus perceive what we would expect to occur naturally in the real world rather than the inappropriate 2 dimensional representation that does not reflect real world reality. Rules regarding one object obscuring another that is further away also play a part in determining how we view two dimensional images - and judge illusions such as the Kanizsa triangle.
Two dimensional representations, whether presented on this site or in advertising, can only be interpreted on the basis of cues concerning colouring and shading. The rest of our understanding draws upon our knowledge of the real world, lit from above, with objects in front obscuring those behind, those closer appearing larger, etc.
Here we have an image of Labrador dog. Once its position is identified (the rollover indicates the position of its head) subsequent viewing leads to easy 'recognition'. On initial viewing, however, it may take some time to identify the dog because all the visual system 'picks up' is a montage of black and white. Our understanding of the real world leads to the organization of the visual input into meaningful images: a dog, a pathway, the base of a tree, etc. This also applies even if the image is further degraded as in the illustration on the right. Most individuals identify this as a man riding a horse.
In these illustrations there was no intention to mislead viewers. However, if this type of imagery were included in advertising the intentions of the advertisers ought to be considered suspect. Advertising agencies know that 'over learning' the rules that apply to reality leads viewers to accept as normal that which is clearly abnormal when considered retrospectively.
The following illusions indicate some additional processes which occur normally but which on occasion can lead us to make false judgements. Some of the illusions also illustrate how we can make sense of incomplete and ambiguous images, the sort that appear regularly in manipulative and semi-subliminal advertising. As you will discover, application of standard visual rules can lead one to identify with some degree of precision what many would consider simply variations in texture without any meaning.
First, close your right eye and stare at the red cross. Then slowly move closer to the screen. The red dot will eventually disappear from your view when it is focused on the 'blind spot' on your retina. This is the area where the optic nerve passes through the retina. As a result there are no photoreceptors to respond to sensory input and whatever is in front of that spot cannot be seen. However, we do not have a blind spot in our visual field because the brain improvises 'information' from the surrounding area and 'fills in' the 'blind spot'. We see a complete image. In this case the lines surrounding the red spot fill the space. In real life information from the scene we are viewing 'fills in the gap' in our vision.
Next, stare at the cross in the middle of the coloured patches for a while and then shift your gaze to the cross in the grey area you will see patches of colour appear there. This is generally because the neurones that are sensitive to one colour become fatigued after use and become 'overridden' by their complementary colours. You thus see what is not there: clockwise from the upper left (red) you perceive green, red, blue and yellow instead of the original red, green, yellow and blue.
Look now at the two dark lines in the illustration on the right. The one on the right seems distinctly longer than that on the left. Our knowledge of perspective 'tricks' the brain into 'seeing' the line on the left as short, because it is 'closer'. Even when we know we are looking at a two dimensional image and not a three dimensional scene our visual systems are still bamboozled.
The small triangles in the centres of the two larger triangles are identical in colour. However, because our perceptual system takes all of a scene or context into account when determining colour, we see one as being slightly darker than the other because of the surrounding colours. Consider this when you look at the Kanizsa triangle on the Imagination page.
It was noted above that we learn that reality has certain characteristics and these are taken into account when we judge what we are seeing. Here is an excellent example. The two illustrations alongside are identical, except that that one has been rotated 90 degrees relative to the first. This simple rotation leads us to judge them differently. The differential judgements arise because we take into account the angle at which light is presumed to 'hit' an object i.e. from the sun up above. So we do not just see a 2 dimensional object, we see it as if it were in the real 3 dimensional world.
The first illustration seems nicely balanced but when turned 90 degrees one can perceive an X of concave and convex circles respectively.
The illustration on the left gives the impression that the long lines are converging or diverging. This is an illusion created by our judgement of the cross hatched lines. The long lines are in fact parallel to each other and it is the action of brain cells that gauge orientation as they respond to a mixture of alignment signals that that mislead us.
Where the white lines cross you should be able to see small, dark, spots. These do not actually exist. Our perception of them is actually created by brain cells that respond selectively to black and white. To produce a crisp, clear, image the retinal cells that surround the area responding to input are either inhibited or excited. Where there is a pattern such as in this illusion, instead of clarifying our perception of the world this process 'produces' the impression of spots where none in fact exist.
When one is looking at ambiguous aspects of semi-subliminal advertising the same processes are at work trying to detect 'edges' around incomplete figures. Instead of perceiving incomplete figures we 'see' meaningful figures. For example, the image of the Devil on the left has no outline or edge, yet nevertheless we see it as a face, rather than as ambiguous grey blotches. You may also note that certain areas of the image seem brighter than others, despite the background being a uniform colour. In ads such imagery would be embedded in ads and what we perceived would be influenced by the surrounding context and our previous knowledge.
If successfully embedded then we would not consciously notice embedded figures such as those believed to be embedded in the following extracts from Marlboro adverts. See below for additional commentary related to these extracts, in particular the image on the left.
The perceptual system is an organizing system. Where there is simply a meaningless array of objects in the 3 dimensional world the perceptual system produces organization. The following illustrations demonstrate that although we could simply 'see' arrays of individual objects what we perceive are organized sets or clusters of objects.
The simplest figure shows some crosses and squares. They could be 'seen' as pairs of crosses and squares or as 12 different shapes. But they are not. We see them as two lines of crosses and squares. Similarly when presented with a larger array of the same shaped objects we will perceive four matching sets of crosses and squares, rather than 48 images, or whatever.
Some of this mental organization arises because of proximity as the illustration on the right shows. The gap between the column on the left and the other columns is sufficient to produce an appreciation of one single column and a group of four columns.
With unusual patterns there is still the superimposition of perceptual organization and this can lead to the perception of 'letters' and 'words'. The illustration on the left, for example, does not produce an appreciation of a random display of shapes, we perceive instead four lines. Other rules relating to Common Fate, Closure and Symmetry can also be identified in the way in which we make judgements about images laid out in 2 dimensional images. All of these processes can be manipulated to produce particular perceptions and influence judgements. Consider for example how little effort it would require to organise the shapes on the continuity image into an approximation of the letters SEX, especially if they were to overlap. Check out the rollover to see. Four dots have been removed and one added. It still looks like a collection of lines but the example now included 3 overlapping letters. If such an example were embedded in an ad on its own it would probably have little impact. If it were presented in conjunction with other 'letters' and/or relevant cues, the ad could make quite an (unconscious) impact.
Various illustrations based on sets of colours such as shown below can also be used to demonstrate that what we 'see' is dependent upon the context in which we see them. If the colours illustrated in the boxes on the right could have been reproduced accurately on screen it could be shown that both contain exactly the same sets of coloured squares. The two sets are simply laid out in different orders. However, when each square is compared with its identical counterpart, because of the different surrounding colours, each colour is 'seen' as being somewhat different. More information about such phenomena can be found in the book Visual Intelligence by Donald D. Hoffman and other books on perception. A larger example, giving you the opportunity to move the various colours around to compare 'differences' can be accessed by clicking here.
The fact that perception is partly influenced by context helps explain why one can clearly 'see' a semi-subliminal embedded figure within the context of an ad or a work of art. However, when it is extracted and presented on its own the image does not seem the same. The image may, in fact, become unrecognisable as the various cues available in the surrounding area are no longer present to help guide the perceptual decision making. In effect, one is seeing a different image. Other factors relevant to this phenomenon are related to various rules as noted below.
A second major part of an explanation for the differences that are perceived when viewing segments of ads in and out of context relates to how we perceive the edges of objects/images in two dimensional representations. Common-sense 'tells us' that we see the edges of objects. Research indicates otherwise. We actually 'construct' the objects that we 'see'. We learn to 'recognize' discrete, sharp, edges even though the information we receive through our visual sense (the eye) is vague and uncertain. The edges of many of the objects illustrated on web sites, in paintings and ads, etc are actually fuzzy as you can verify by enlarging them slightly or viewing them with a magnifying glass. Nevertheless, despite the fuzziness around the edges we 'see' distinct edges. Similarly, even when there is slight variation in colouring we perceive solid colours.
Despite the limitations of our visual system we need not get concerned. Our judgement remains secure even although it is based on a fallacy, the fallacy that we actually 'see' what exists in the real world.
Our conscious perception of what we are seeing is much more clearly delineated - and thus much more useful - than the indistinct and fuzzy visual input received at our eyes. The illustration included in the next paragraph gives some insight into this process and how it is influenced by pre-existing knowledge by focussing upon the judgement of perceived contours.
The impression in this illustration is of ridges and valleys. It seems, therefore, that the shading ought to vary across each segment of the illustration i.e. the area between each of the noticeable boundaries ought to vary in brightness, depending upon how much light it was receiving. This is, in fact, what is perceived. Light seems to be catching the illustration from the right hand side. The right hand side of each segment is thus perceived to be darker than the left hand side. Yet this is an illusion produced by the visual system.
If you cover up the edges of any of the segments or run the mouse over the image to see the rollover image you will note that the shading is pretty much the same all the way across. With perfect reproduction of the basic image that was scanned the same effect can be produced by segments that are uniform in colour, other than the contoured edge. It is the existance of the contoured edge at the side of each segment that leads us to perceive dips and ridges. And, since these dips and ridges are seemingly there, our judgement of what we are 'seeing' must take this into account. Accordingly, we 'see' variations in shading that do not exist in reality.
Accurate variation in shading can only be recorded by a light meter. People with normal vision are incapable of such accurate judgements because our perceptual systems are designed to enhance any edge/contour. We then can clearly 'recognize' objects even when the visual input is actually fuzzy and indistinct. When these and other powerful rules of perception intended to be applied to the real, three dimensional, world are applied to two dimensional images we are often deluded.
The type of 'errors' produced by the perceptual system, in conjunction with various other rules applied to the information extracted from the visual system (see Hoffman's book on Visual Intelligence), mean that it is possible to 'see' embedded figures in adverts even if these figures are incomplete. The perceptual system, in effect, 'completes' them just as it 'fills in' the blind spot in our vision produced by the lack of receptors in the centre of the human retina.
Various perceptual rules that are applicable to 3 dimensional reality are applied to 2 dimensional ads in order to extract meaningful information or to make sense of the input. For example rules regarding curvature are applied in conjunction with knowledge of contours. These lead to judgements as to whether an object is convex or concave. A perfect application of these rules to a two dimensional advert can be found in the extract from a Marlboro ad reproduced on the right. If it seems familiar, then so it ought to be, it was initially presented near the top of this page in another context.
The 'face' that can be perceived has no distinct edge but the essential facial features can be 'seen', embedded within the background. This 'face' can be 'seen' because of judgements based on the 'rules of perception'. The rules are applied to the variations in texture evident in the image even if these do not produce a complete image. This is an automatic set of processes that are applied whenever a normally sighted person is viewing a two dimensional or three dimensional object, providing they do not allow their judgement to be 'biased' by oversimplistic or concrete thinking and make snap judgements about the ad as a whole.
Other rules are also applicable. As was noted above, where objects are partially obscured or where contiguous or aligned edges exist, there is an assumption that one object is in front of another (as evident in the Kanizsa triangle illusion) and that the object has a particular shape. Additionally, when there are curves, these are taken to indicate rims (edges or outlines of objects). Concave creases point in to an object whereas convex creases point out of an object. The 'eyes' of the Marlboro 'face' are darker and have concave 'creases' - they are further away. The 'nose' has convex 'creases' and is lighter - it is closer. The 2 dimensional image thus assumes a 3 dimensional form based on the application of the rules of visual perception allied with existing knowledge of the world.
On the basis of intuitive judgements, appraisal of a number of Malboro ads and knowledge of the rules of visual perception, it is therefore possible to state with a considerable degree of certainty that this 'face' is not simply a judgement based on aberrations in the colouring of the ad. Nor is the 'image' simply the projection of ideas around a set of ambiguous visual information. This embedded image (face) was intended to be perceived, unconsciously, as a face or as some other meaningful entity by some of those who would view the ad.
A few possibilities seem likely. The author prefers the primary interpretation to be related to death or dying because of other contextual features of the ad e.g. depressive colouring, helicopters in the background, various connotations associated with the caption, etc. One possible interpretation leads to 'identification' of the face as that of a dead, buried and decaying individual. Others, equally reasonable, could be face-like imagery related to the Star Wars movie. The 'face' shares some of the same essential structure as the headgear of a Star Wars trooper (illustrated above) and, to a lesser extent imagery from other science fiction movies e.g. the alien in Predator, or a latter day soldier in anti biological warfare gear as illustrated on the left.
What is ultimately 'recognized' is not simply what is in the ad, it is what is the outcome of mental processes triggered by the ad. The 'recognition may occur intuitively and consciously, as in the author's analysis of the ad, or preconsciously in instances of normal viewing when individuals are predisposed on the basis of previous knowledge or expectations. An ex-soldier, for example, might respond to the mood of the ad. If a depressed mood were enhanced by the embedded imagery, this could lead to depressive thoughts. If the viewer were a Star Wars fan the response might be quite different. To paraphrase the popular saying 'Horses for courses', one might state that ads, in particular embedded elements of ads, present 'Images for Individuals'.
The construction of specific images are likely to be based on the type of psychographic analysis discussed in textbooks on marketing and consumer behaviour. Language is also an important factor as the next section demonstrates. Complementary information can be found on the Imagination page.
For more on Faces in Ads see the Faces pages.
The two lists of images and titles are the basis for a simple experiment that can demonstrate that how we name images affects how we remember them and also how we might draw what we remember. Note that each list has identical images. What differs is simply the label that is attached to each image.
When people are shown either one or other list and then asked to recall what they have seen it is almost invariably the case that what they recall is influenced by the label associated with it. The first image in each case is a rough diamond shape within a square but note that the lines of the 'diamond' are slightly curved. When viewers recall this list the lines are straightened in any drawing that they produce. In contrast when viewers are given the list with the label curtains in window they draw even more enhanced curves to make their drawing much more like 'curtains in a window'. Similar results apply to most of the items on the list. When individuals are tested the results are generally not as powerful as when two groups of people are compared. However, there usually are sufficient differences even with an individual trying to recall images to make this a powerful reminder that we do not simply see things.
The conclusion that is drawn on the basis of such studies is that we do not simply remember shapes, we associate them with language. And, if we label objects, then it is the label or name that may influence our judgements of what we 'see'. Hence, when viewing ambiguous stimuli or embedded words in ads it is our previous knowledge that will, in part, determine what we perceive. Advertising agencies are, of course, familiar with this and other aspects of psychology and they would be remiss if they did not make use of their knowledge to facilitate sales. However, when they cross the boundary into using embedded and other manipulative techniques then this is surely unacceptable.
Last Revised: 3rd January, 2003