Dr Louisa Kulke from the University of Göttingen has investigated how our eyes and brain react when we see emotionally charged or neutral faces. She combined eye-tracking and electroencephalography (EEG).
The result: reflex-like eye movements are independent of the expression a face shows; our attention is drawn to them just as fast. The study was published in the journal Neuroscience.
In everyday life, people are exposed to numerous stimuli, just as when walking through a town, there are people and faces, clothes and shop windows. From this onslaught of information, relevant content must be filtered and reacted to.
“This shift in attention is often accompanied by a movement of the eyes,” says Kulke, a researcher in the Department of Affective Neuroscience and Psychophysiology at the University of Göttingen.
In the current study, she combined two methods to investigate what happens in the brain during this attention shift: eye-tracking and EEG. With the Eye-Tracker, research volunteers sat in front of a device that records eye movements.
Kulke then showed them standardised faces with different emotional expressions. At the same time, EEG measured brain waves via electrodes placed on their head.
“We investigated how quickly our test subjects look at the faces that appear on the screen in different places,” says Kulke.
The result shows that quick and immediate eye movements were occurring independently of the facial expression.
“We quickly look at people in our environment with the same speed, regardless of whether they look cheerful, angry or indifferent.
The study also showed that the first reactions of the brain are independent of facial expression. Only later – after the eye movement is completed – do the reactions of the brain show strong responses to the emotional expression of the face.”
This is an image showing brain responses to happy, neutral or angry faces shown at the periphery of vision. The image is credited to Louisa Kulke.
According to Kulke, the result is also interesting for follow-up studies.
The question is whether the processing of emotions also takes place later, when the eyes do not react reflexively, but are consciously controlled using a tangible task.
Kulke explains: “For example, do we process facial expressions that we see out of the corner of our eye before we move our eyes?
When given the explicit goal of only looking at certain faces, such as our friendly fellow student, and ignoring others, such as our annoying neighbour, how does that affect our reactions?”
To scientists who study perception as well as to those who study emotion, the idea that emotion routinely alters perception may seem completely foreign. Most of us assume quite reasonably that as we look at a hill, for example, the steepness of the incline in our visual image is more or less the steepness of the hill in the world.
Moreover, our perception of the steepness will change from one occasion to the next depending on our mood . For example, when we are feeling sad, we will perceive the hill to be steeper than when we are feeling happy. Such findings indicate that the perception of spatial layout is in fact influenced by non-optical factors, including emotion.
In this article, we review evidence of a variety of emotional influences on visual perception. Rather than a single, general mechanism that explains them all, a number of processes appear to be involved. Thus, we discuss candidate explanations as we review specific findings.
The emotional phenomena discussed include effects on early visual processes, global vs. local perceptual focus, susceptibility to visual illusions, and perceptions of natural environments.
In addition, since emotions have both bodily and motivational components, we also touch on perceptual influences of bodily and motivational states. For example, both emotion and motivation appear to prepare the visual system to detect relevant aspects of the environment by making them easier to see [4, 5].
And both emotional and bodily states appear to regulate visual perception of spatial layout. We propose a functional view in which emotional influences on perception can be seen as evolving in the interest of minimizing negative and maximizing positive outcomes, a view consistent with the “affect-as-information” hypothesis [6, 7]. More generally, we propose that emotion influences perception in the interest of resource maintenance.
The reader will note that, although this article concerns Emotion and Perception, we consider only emotional influences on perception and not the reverse. However, it should be understood that perception is also fundamental to emotion.
Indeed, many emotions arise immediately upon the perception of emotionally evocative stimuli, some requiring more interpretation (rising gas prices) and some less (snakes, spiders). But exploration of those phenomena requires a separate treatment [for a review of relevant conceptions, see 8].
Fear and Orientation Discrimination
Emotional arousal guides attention so that people’s attention tends to be drawn to objects that are arousing. Indeed, some years ago, Herbert Simon  proposed that a chief function of emotion is to interrupt and reorder processing priorities.
Thus, even the most avid chess player is likely to stop his game upon noticing that his house is on fire. As we shall see, some of the important influences of emotion on perception are mediated by attention, but emotion can also influence pre-attentive perceptual processes.
This fact was discovered by Phelps, Ling, and Carrasco  using an orientation discrimination task. The task consisted of showing four sinusoidal gratings simultaneously, in which three of the patterns (distracters) were oriented vertically and one (the target) was tilted eight degrees clockwise or counter-clockwise.
The grating contrast levels were varied on each trial. Participants were to locate the target as quickly as possible.
The patterns were preceded by a rapidly presented face displaying either a fearful or a neutral expression. The logic of the experiment was that if emotion enhances perception, discrimination should improve following exposure to a fearful face. The results supported that hypothesis: contrast sensitivity at threshold improved by three per cent following a fearful face. Examples of stimuli at the contrast thresholds for the fearful and neutral groups are shown in Figure 1 along with the fearful and neutral faces (which show the psychologist Paul Ekman posing the two expressions).
In a second experiment, the authors  asked whether emotion had really changed perceptual sensitivity or whether instead the fearful and neutral faces had differentially influenced covert attention.
To find out, they took attention out of the equation by presenting the fearful or neutral face in the same quadrant as the subsequent target grating would appear (instead of in the middle of the screen as in Experiment 1).
But again, exposure to a fearful face increased contrast sensitivity, even though attentional shifts were no longer involved. Hence, the authors could conclude that , “…the mere presence of a fearful face increased contrast sensitivity.…. [and] Emotion actually affects how people see.” In addition, they found that the facial expressions had to be emotionally meaningful — if the faces were inverted, the effect disappeared.
The authors proposed that this effect is probably the result of feedback from the amygdala, to the early visual cortex, as well as to regions that enhance attention.
The amygdala responds to significant stimuli, including fearful faces, rapidly and prior to awareness. A fearful face indicates that there may be a threat in the environment, but it gives no information as to its form or location, so enhanced contrast sensitivity might aid in detecting the threat .
Other research shows that high-level goals can also influence the responsivity of the amygdala to affective stimuli. The amygdala is believed to respond to affective stimuli more or less automatically, but evidence shows that responses depend on the current relevance of affective stimuli .
In an imaging study, participants were asked to rate either the positive or negative aspects of 96 famous names (Adolph Hitler, Paris Hilton, Mother Theresa, George Clooney).
The results showed that when evaluating positive aspects, the amygdala responded only to the names of people that a given participant liked, and when evaluating negative aspects, the amygdala responded only to the names of disliked people. Thus, the automatic affective reactions of the amygdala were guided by the current goal of the individual.
It is unclear whether the amygdala itself filtered information for motivational significance or whether top down processes did so before information reached the amygdala. But the results encourage a view of the brain in which high- and low-level processes continually interact—a view within which it becomes less surprising that emotion can affect perception.
Mood and Global-Local Perception
People sometimes say that a person “can’t see the forest for the trees.” In doing so, they imply an incompatibility between perceptions of details and perceptions of wholes. Some conditions encourage global perception and some encourage local perception, but people generally show a tendency to process globally.
Emotion also influences whether people focus on the forest or the trees. After hitting his head during a parachute jump, the psychologist Easterbrook  noted that his spatio-temporal field seemed to shrink .
ith this experience in mind, he later proposed that stress narrows atttention. Fifty years later, many findings support this idea as well as the extension that positive emotion broadens attention [16, 17].
Relevant research sometimes employs standard tests to measure global and local perception. On the Kimchi test, respondents are shown a target geometric figure and asked which of two comparison figures is most similar to it .
As shown in Figure 2, the target might be three small squares arranged in the overall shape of a triangle.
People then choose which of two comparison figures is most similar. One comparison figure is a triangle composed of small triangles, and the other is a square composed of small squares.
A local response would be to choose the figure with squares, because the target figure had been composed of squares. A global response would to choose the figure with triangles, because the overall shape of the target figure had been a triangle. When investigators induce happy or sad moods (for example, by having participants spend a few minutes writing about a happy or sad event from their lives), participants in happy moods often adopt a global perceptual style, whereas those in sad moods adopt a local perceptual style .
Another standard method, the Navon procedure, involves measuring reaction times to large or small letters . For example, a large “L” might appear made up of many smaller “T’s.” Respondents might be asked to indicate as quickly as possible whether they see an “L” on a given trial. Comparing the reaction times to detect letters appearing as global stimuli and those appearing as local stimuli yields a measure of whether global or local perceptual styles are dominant.
Some research using this measure suggests that although global processing occurs in generic positive moods, states in which a specific object elicits approach motivation (e.g., hope) can lead to local rather than global responses .
Still other research suggests that rather than a dedicated relationship between affect and perceptual style, positive affect may facilitate and negative affect may inhibit whatever orientation is most accessible in a given situation.
In many situations, a global focus is dominant , a tendency sometimes called the “global superiority effect.”
It is possible then, that affect influences whether one focuses on the global forest or on the local trees simply because positive affect says “yes” and negative affect says “no” to the (generally more accessible) global focus.
A test of that hypothesis used cognitive priming techniques to alter whether a global or a local orientation was momentarily more accessible . The results showed that when local responding was made especially accessible, the usual result was reversed. Positive affect then led to a focus on details and sad moods to a focus on the big picture.
It appears, therefore, that positive affect may facilitate whatever the dominant orientation is rather than being specifically tied to a global focus. Thus, whether one focuses on the global forest or the local trees is indeed influenced by one’s current emotional state. However, rather than reflecting a direct connection to perception, these data indicate that positive affect can empower (and negative affect can inhibit) either a big or a small view, depending on which is dominant in a given situation.
Emotion and Attention
The studies we have reviewed show that emotional and motivational factors can regulate global vs. local orientations to visual stimuli and whether or not perception includes or excludes contextual stimuli. But, as noted above, the same is also true when the stimuli are conceptual rather than sensory.
For example, individuals feeling happy are more likely to use stereotypes and other categorical information when forming impressions of others. By contrast, when forming impressions, people feeling sad focus on behavioral or other detailed information and tend not to use global categories [e.g., 25, 26, 27]. Such results may indicate that the influences of affect on global-local perception and conception are mediated by attention.
Attention is sometimes thought of as a spotlight that directs limited processing resources to the most relevant stimuli [28, 29]. If affect signals value  or motivational significance , then we might expect affect to influence attention. For example, activating an affective attitude leads to attitude-consistent judgments  by biasing attention toward attitude-relevant stimuli .
Studies examining the role of emotion in attention have sometimes employed a spatial probe task for measuring attention.
In the spatial probe task, two words are presented briefly in different locations followed by a dot probe in one of the locations. If an emotion-relevant word attracts attention, the dot appearing in that location will be detected faster than a dot appearing in the other location. Speed of response to the dot is thus a useful measure of selective attention .
Some of the research using this technique has been conducted by clinical psychologists interested in the effects of anxiety. The general finding is that fear and anxiety bias attention toward threatening stimuli, including words and pictures [e.g., 34, 35]. Selective attention may thus serve to facilitate the processing of threat information . But of course, if affect governs attention and attention in turn governs affect, then when affect draws people’s attention to possible dangers, it is likely to induce stress and anxiety .
Positive affective reactions signal opportunities rather than dangers, raising the question of whether positive affect also directs attention. Indeed, evidence from dot probe studies indicates that positive moods bias attention toward positively valued stimuli .
As a result, positive affect should make rewards easier to detect, just as anxiety facilitates threat detection. Of course, attending to the upside, rather than the down-side, of events is also likely to elevate mood and subjective well-being.
Top-down Effects of Motivation and Emotion
Traditionally, the study of perception has stressed low level, bottom-up visual processes. But research suggests that higher level processes may play a role as well. A recent study demonstrated top-down effects of emotional information on face perception .
The study involved a binocular rivalry task, in which a different image is presented to each eye—for instance, a face and a house. In that task, only one image is consciously experienced at a time, and which image is seen tends to alternate every few seconds.
The images essentially compete for dominance, the more important or relevant image being perceived relatively longer.
In this experiment, faces became more dominant in the rivalry task after being paired with descriptions of negative social information, such as, that the person lied, stole, or cheated. The results suggest that gossip and other social information may tune the visual system, aiding in the detection of persons who should be avoided without requiring any direct negative experience with them. This idea that the emotional significance of objects may make them easier to see has a long and interesting research history, as we see next.
Years ago, the “New Look” in perception proposed that perception should be influenced by motivation. For example, Bruner and Goodman  reported an experiment in which a sample of poor children from the Boston slums perceived coins to be larger than did children from wealthier Boston families.
The same effect did not appear for similarly-sized cardboard disks, leading the authors to conclude that motivation can influence perceptions of size, making motivationally-relevant objects easier to see.
At the time, the idea that visual perception, our window to objective reality, might be guided by subjective desires was seen as quite unacceptable. Moreover, when the New Look was elaborated to include predictions from Freudian theory, it was soundly rejected by many investigators.
But the basic hypothesis that motivation might affect perception has since been revisited. Recent evidence shows that, for example, people who are thirsty perceive a glass of water as taller than those who are not thirsty .
And when typically neutral goals, such as gardening, are made positive by pairing them with positive stimuli, tools associated with the goal (such as a shovel) appear larger .
Similarly, smokers deprived of cigarettes tend to overestimate the length of a standard cigarette . Other findings also indicate that ambiguity in visual stimuli (e.g., a stimulus that could be seen either as the letter “B” or as the number “13”), will tend to be resolved by seeing the stimuli in a way that leads to reward in an experimental situation .
In related research, participants who had agreed to walk on their campus wearing a large, embarrassing sign underestimated the distance to be walked .
The authors reasoned that the misperception of distance was a way of reducing the cognitive dissonance of having freely chosen to engage in such an unpleasant action.
Consistent with the original New Look logic, such data again suggest that goals can tune the visual system to see the world in motivationally-consistent ways [for more on the social psychology of perception, see 43].
Whereas most of the emotional effects we have discussed have been evident only in limited, somewhat artificial laboratory settings, this experiment and the research to be discussed in the remainder of this review concern perception in the world.
Emotional effects in real-world environments may be more pervasive than most people realize.
It is often assumed that one of the primary goals of the visual system is to recreate the environment, forming a representation in the brain that is as accurate as possible.
However, research over the past ten or fifteen years has demonstrated that this is not the case. Rather than reproducing pictures inside the brain, research results indicate that what we perceive is a systematically altered version of reality. Part of what we “see” are the opportunities for and costs of acting on the environment.
For example, the ground is perceived relative to its walkability and to the bioenergetic costs that this action would incur.
However, these non-visual influences are not limited to energy-related factors: emotions too are a source of non-visual information that affects visual perception.
Moreover, the influences of such non-visual information generally appear oriented toward such beneficial consequences as conserving energy, attaining goals, or avoiding danger.
In the following sections, we first review research showing the role of extra-visual influences in the perception of spatial layout.
We then review research indicating that emotions may serve a similar function and are integrated into perception in a similar manner.
University of Gottingen
Louisa Kulke – University of Gottingen
The image is credited to Louisa Kulke.
Original Research: Closed access
“Neural Mechanisms of Overt Attention Shifts to Emotional Faces”. Louisa Kulke.