Felt Time

The feeling of duration often serves as an “error signal”: it is taking too long for dinner to be ready or for the bus to come. Human beings have a relatively precise sense of time for a span reaching from seconds to a few minutes at most. To date, the only kind of internal clock that has been found is the circadian rhythm: bodily and mental processes fluctuate systematically over the course of the day. People represent different chronotypes on the basis of when daily highs and lows of performance occur.

When do we feel “time”? In what situations do we become aware of the passage of time? Hours can pass without us ever making an explicit judgment of time. A typical situation in which time enters our consciousness occurs when we wait, especially for a desired event. For example: school is in session, the weather is beautiful, the sun is shining, the fragrance of spring floats in through an open window, and students are sitting in math class. A look at the clock indicates that the lesson and school day will be over in fifteen minutes. Soon, everyone will get to go home and eat (hunger and anticipation of the meal); the afternoon beckons: freedom to play with friends. For a brief spell, thinking about the afternoon distracts the pupils, then their attention turns back to class; a girl is standing at the blackboard, trying to solve an equation. A nervous glance at the clock reveals that there are still fourteen minutes remaining until the lesson ends. Time expands: the single minute since the last look at the clock seems to have lasted forever. Time inches forward at a snail’s pace. The sense of passing time becomes physical: hunger grows, variables and numbers become more and more uninteresting; another look at the clock indicates that salvation still lies thirteen minutes away.

Another example: French tourists in the United States, and American tourists in France. Time-cultures (among other things) collide. At a fancy restaurant in Los Angeles, the French couple places their order and looks forward to a romantic evening. They have barely settled into intimate conversation when the waiter shows up with the first course. This is much too fast for the French sense of time. In Bordeaux, an American couple sits down for a meal at a restaurant. They placed their order a long time ago, their conversation has now dried up, and they’ve eaten their baguette; as their hunger mounts, the feeling grows: “They’ve forgotten about us!” In fact, the food will arrive—but after a span of time quite different from what the Americans are used to. In either case, the guests at the restaurant become aware of time. In this way, the perception of time functions as an error signal indicating that something is amiss.

When waiting for an elevator or at a stoplight, two minutes can seem to take too long. Sometimes mere seconds prove irritating, for example, when the car ahead starts to back into a parking spot too slowly—why now of all times?—and keeps one from moving forward. When the feeling of time provides an error signal, it proves functional inasmuch as it calls for action. If, at lunchtime, too many people are using the elevator and it doesn’t even stop at one’s floor, it might be time to take the stairs. All the same, there are also situations when one feels trapped in time: there’s no escaping a wait in a traffic jam, for instance. Unpleasant feelings may arise. Some people start to feel aggression creeping up, and combined with a certain amount of impulsiveness, this can lead to violent actions. As we saw in chapter 1, impulsive individuals overestimate the duration of events. When impulsivity, the feeling of being trapped, and a potential for aggression occur together, certain people may erupt in violence in order to get out of a situation that is felt to be inescapable.

This may seem like an exaggeration. However, the Associated Press reported just such a case in 2007.¹ Attacks by drivers on construction workers in California got so out of control that the highway had to be closed to traffic entirely. Work was being done to expand the roadway, and frequent traffic jams were the result. As construction proceeded, again and again drivers threatened workers, threw objects such as burritos, and even, in one case, fired a shotgun. The straw that broke the camel’s back was when a man tried to run down a worker with his car. (He was subsequently arrested.) The spokeswoman for the state transportation department also interpreted incidents in terms of drivers’ experience of time; she observed that people who are saddled with excessive waiting periods will often lose patience and overreact.

To be sure, this is an extreme case. All the same, it demonstrates how closely the perception of time is tied to strong emotions, which can lead to dramatic responses. Needless to say, one might also discuss the cultural dimension of such events. The Californians’ reactions seem strange to Japanese. Indeed, intercultural comparison indicates that students in the United States attach less value to rewards that are deferred until a later time (see chapter 1) than their peers in Japan.² In other words, Japanese are able, on average, to wait more patiently than people in the United States.

The evaluation of temporal duration and the speed at which aspects of life occur varies from region to region throughout the world. Robert Levine, an American social psychologist, has summarized his findings impressively in his book A Geography of Time.³ The general picture emerges that industrialized nations view and manage time differently than countries that are less developed technologically; comparable variations occur between city and rural populations, big cities and smaller ones, as well as temperate regions in the northern hemisphere and more tropical lands. Dozens of studies conducted across the globe have verified these results empirically. Levine had his students—who came from all over the world—measure time in their home countries of, among other things, walking speeds and how long it takes to buy stamps at the post office. It turned out that the fastest cultures are found in large cities in industrialized, northern countries, where a great deal of attention is paid to time and punctuality and people live at a quicker pace, often feel rushed, and have difficulty dealing with delays. The other extreme involves cultures in rural parts of countries in the equatorial region, where people are in less of a hurry and work proceeds at a slower pace; here, one finds few public clocks—and even fewer that function—and people are more likely to pursue sociable forms of leisure, say, talking and drinking coffee or tea. Just as cultures may be distinguished in terms of political, economic, and historical factors, they can be defined in terms of how time is managed and subjectively experienced.⁴

Figure 5

A non-Western time culture: a gardener at Brihadeeswarar Temple in the Indian state of Tamil Nadu sits on the lawn and cuts the grass with garden shears.

Levine’s analysis yields two fundamental cultures of time: the first works according to event-time, the second according to the abstract time of the clock. In the first case, people are oriented toward the duration of an event. A meeting can only take place when some prior activity (e.g., a conversation or a meal) has been completed. In contrast, people who are oriented toward the abstract time of the clock will interrupt what they are doing in order to keep an appointment. Industrialization succeeded because, among other reasons, humans can be made to follow time precisely. Clocks made it possible to synchronize sequences of work efficiently. Societies that are more oriented toward abstract clock-time prove more successful economically. Merchants no longer agree to meet at some point shortly after sunrise; instead, they meet at exactly 7:30. “Clocking in” ensures that employees arrive on time and work for a fixed period. People who come from different time-cultures run the risk of missing each other when they make a date. Problems are practically preordained.

At first glance, it seems that bosses run chiefly on clock-time. When executives meet, the expectation is one of punctuality. In corporate environments, higher-ups demand that employees observe the schedule. Who can afford event-time is a question of power. An employee cannot afford to be ten minutes late for an appointment with the boss. The boss, on the other hand, can very well complete a telephone call or finish dictating memos to his secretary. Whoever holds power can make others wait.

The perception of time says something about the condition of the individual who judges that it is passing too slowly or too quickly. Time passes slowly when one is bored or yearns for the arrival of another; it passes quickly when one is “doing well.” Ninety minutes on the plane, when traveling to meet one’s beloved, can be excruciating; ninety minutes watching a good movie often fly by much too quickly. This is also the reason we see so many people staring at their cell phones on the commuter train: the Internet helps pass the time when traveling.

In such instances, time represents a vague feeling: either “This is taking too long,” or “Too bad it’s already over.” A poem by Christian Morgenstern, which lends the mechanisms of time vivid expression, illustrates how our perception of time functions in everyday life, the way we experience the phenomenon:

Although dressed up in playful form, the message is clear: when we pay attention to time, it passes (like an obedient sheep); when we are distracted from time, it runs away quickly (like a puma). In special situations, hours can sometimes fly by; they feel like minutes. Commonly, this involves the sensation of “flow,” which has been extensively researched by Mihaly Csikszentmihalyi, an American psychologist;⁵ often, it involves a sense of euphoria. The feeling of flow occurs when a person, in a state of utmost concentration and complete motivation, pursues an activity and reaches the limits of his or her capacity, without the challenge proving too great or too small. It might involve the conception and elaboration of a new idea, writing an important document, or performing music in a group. A high degree of skill, focus, and dedication must work in concert so that the actions required to perform the demanding task are completed as smoothly as possible. Such concentrated activity requires that time be disregarded; time flies by, as if it did not exist at all. Once work or play that has happened in a flow is over, one is surprised to find that it is already dark (or light) outside.

But how, precisely—in contrast to physical time or, alternatively, time as measured by the clock—do people assess temporal duration? Can animals also do so? We know that rats, mice, pigeons, and certainly monkeys can gauge time. Many animals have been shown to be able to assess short spans of time relatively precisely. For example, test animals are trained to press a disk or lever after fifteen seconds. If the animals do so after this interval (allowing for a small margin of error), they are rewarded with a piece of food. That said, the animals sometimes need dozens—or even hundreds—of trials before they consistently (i.e., with a certain degree of probability) exhibit the desired behavior.⁶ Animals cannot be given verbal instructions; they must learn what to do by trial and error, which can last quite a while. However, once they have learned, pigeons will use their beaks, and rats their snouts, to push the disk or lever after the interval has elapsed. In one particular task of time-assessment, monkeys learned to hold down a key down for periods varying between two, four, and eight seconds—depending on the color of a light signal; for example, they displayed the ability to press the key for eight seconds when the light was red, four when it was green, and two when it was yellow.⁷

When studies are conducted with human beings—for the most part, psychology students—some tasks for measuring the perception of time are common: acoustic or optical stimuli of a certain length are provided to determine the ability to assess temporal duration. In one experimental paradigm, once the stimulus has been given, subjects are asked to press a key for exactly the same span of time, that is, to reproduce the duration. In another experimental paradigm, when assessing the ability to discriminate duration, two stimuli of different length—tones, for instance—are provided; subjects must indicate which of the two was longer. The standard might be a tone that lasts exactly one second (1,000 ms). Then, the duration of the second tone is varied (1,050 ms, 1,100 ms, 1,150 ms, 1,200 ms, etc.; the sequence of the standard and comparison tones is also varied), and responses are collected. On the basis of these responses, it is possible to determine how large the difference between the two temporal durations must be for subjects to reliably determine which of them is longer. In other words, a threshold of duration discrimination is established. These two methods offer the advantage of not requiring verbal judgments of absolute temporal duration (e.g., “The tone lasted four seconds”); instead, researchers can determine subjects’ capacity for temporal perception down to the millisecond, either by means of analyzing the duration of the key press in the duration reproduction paradigm or by way of analyzing the accuracy in the decisions concerning the length of the two tones in the duration discrimination paradigm.⁸ Verbal judgments rely on spans broken down into seconds; moreover, subjects make crude estimates by rounding numbers up or down: “about ten seconds.”

Especially when the temporal durations last for several seconds, many subjects involuntarily experience the need to count. They want to apply a “time ruler,” as it were; in consequence, they can offer very accurate estimates. Preventing this is not easy. Some researchers trust their subjects and simply tell them not to count. Another possibility is to provide a second task to be performed at the same time; the distraction is meant to keep subjects from counting. But then the danger is that the second activity will interfere with the primary task of estimating time. Experiments demonstrated conclusively that subjects considered spans of time that were prescribed (i.e., experienced) to be longer than periods in which they were asked to perform a second task. Such results match what is described in Morgenstern’s poem: if one is distracted, time seems shorter.

Despite one hundred and fifty years of experimental research on the experience of time, there is still no consensus about how the nature of our capacity for perceiving and discriminating temporal duration. Michel Treisman, of Oxford University, first proposed a working model for a kind of internal clock in 1963:⁹ a pacemaker in the brain emitting pulses at regular intervals, which are collected in a counter; in turn, the number of pulses collected in the counter is supposed to define subjective duration. By means of this cognitive model, one can readily explain why a span of time that lasts longer in physical terms is also perceived as being longer in subjective terms: the hypothetical counter has collected more pulses.

Subsequent revisions of the simple model of pacemaker and counter introduced attention as a component:¹⁰ only when one is attending to time are impulses registered by the counter. If one is distracted, fewer pulses are counted and time seems to be shorter. This version of the model also explains why time seems to last longer when we are waiting. If one is not distracted and pays attention to time while waiting at the doctor’s office, twenty minutes can feel like a very long time. In contrast, if one happens to be reading an engaging novel and does not heed the time, the same twenty minutes pass rapidly. The pacemaker model fits well with the experience of time portrayed in Morgenstern’s poem. In the first case, a large quantity of pulses went into the counter; in the second, only a few.

The fact that researchers have proposed a vast array of alternative models indicates that this model hardly offers the final word.¹¹ For instance, temporal features of memory—specifically, the disintegration of memory traces—might function as internal signals for the passing of time.¹² Accordingly, the sense of temporal duration would be linked to how keenly one recalls events at the beginning of the temporal interval in question. The more time that passes, the weaker the recollection of what occurred at the outset will be. On this model, there is no separate mechanism responsible for assessing time—no “watch” in the brain; instead, the sense of duration arises as memory traces of events in the past subside.

Another view holds that the sense of duration comes about through the feeling of intellectual and emotional exertion.¹³ Novel events are felt to last longer because they make greater claims on the faculties of perception, thinking, and emotional assessment. In contrast, familiar items do not require special analysis or evaluation; perceiving them is not as intensive. Accordingly, the effort expended also contributes to the evaluation of temporal duration. This model also does without a special, internal clock that is “read” when time is gauged. Rather, we are constantly engaged in mental activity and judge situations as they arise. The sense of time emerges from the degree to which attention, memory, thinking, and feeling are activated.

In sum: at least so far, brain researchers and psychologists have not discovered an internal clock responsible for spans of time ranging from seconds to minutes. The pacemaker-counter model is just a metaphor, anyway—a ticking stopwatch that can be turned on and off. All efforts to find a physiological mechanism in the brain representing the sole instance for the experience of time have proven unsuccessful until now. Or, to put matters more precisely: there is no universally accepted theory of temporal perception (chapter 7 offers a new model combining information from the fields of philosophy, psychology, and brain research). Still, despite criticism of the idea of an internal clock as oversimplified and mechanistic, the pacemaker-counter model has proven adequate to describe many phenomena. According to this model, the subjective perception of time can be influenced by means of two mechanisms. The first involves directing attention, as discussed above, which has an effect on the pulses being counted. The second mechanism operates by way of the frequency at which pulses are emitted. If the rate of the pacemaker is increased, the pulses accumulate at a faster rate. This entails a subjective lengthening of duration.

In the early 1930s, the wife of American physiologist Hudson Hoagland was bedridden with a fever. Whenever Hoagland left the house even briefly—say, to go to the pharmacy—his wife complained about the length of his absence, which had lasted far too long in her mind. Every bit the scientist, Hoagland promptly performed a simple experiment on his spouse, instructing her to count to sixty. As it happened, she did so rather quickly—once, after just thirty-seven seconds. Hoagland concluded, on this basis, that the fever had produced a level of physiological activity higher than average, and that this had made her count faster and, by the same token, overestimate temporal duration significantly. This overestimation of time can be described with the pacemaker-counter model. The heightened level of physical agitation leads the pacemaker to operate at an elevated frequency; accordingly, more pulses are accumulated. Thus, the wife came to think that, say, ten minutes had passed, even though her husband had only been gone for five. In an article published in 1933, Hoagland already speculated about an internal clock running at different rates depending on physical conditions.¹⁴

We have still not asked one important question: What time frames should models focus on in the first place? When we say that someone has pressed the doorbell for too long, it is a matter of, say, four seconds. If a stoplight refuses to change, we evaluate temporal durations in terms of minutes. We make our assessments by voicing a feeling (“The light’s been red too long”). Only when a comparison is drawn to time on a clock do we use physically determinate units—seconds or minutes.

In La Jolla, near San Diego, there was an intersection that was notorious among my colleagues for the wait it imposed on drivers. Because the research group working under Martin Paulus had offices off campus, it was necessary to cross a large thoroughfare on foot or by car to reach various university facilities. The stoplight annoyed my colleagues and me because it took “so long” to switch to green. One day, I measured how long it stayed red; two days later, I double-checked the measurement. The light stayed red for two minutes and fifteen seconds. When I asked them how long the wait lasted—which they had experienced again and again over the years (and often several times a day)—most of my colleagues guessed too high. It should also be taken into account that one frequently arrives at a light when it is already red, and the switch to green occurs in less time. This slight vexation reveals two things. In a temporal extension that is only two minutes long, people already lose patience. “Two minutes” is a unit of time commonly used to indicate that one will do something “right away.” In this case, two minutes proved too long in subjective terms. In addition, this annoyance shows how imprecise assessments of time can be, even in the range of only two minutes. The question, then, is how much time human beings can estimate with relative exactness. The answer would demonstrate that a temporal mechanism—whatever its precise form—is at work.

The first item to note is that durations lasting up to about three seconds can be estimated rather precisely and accurately. If tests are conducted on the same person, the variance of subjective estimates of duration proves relatively small when intervals last up to three seconds. Only when durations are longer than three seconds do larger deviations between objective and subjective time emerge; estimations become less accurate.¹⁵ There is nothing magical about three seconds; this value simply means that perception and activity occur at an optimal level within this period of presence, the “now” (see chapter 3). Within these limits, we can assess temporal duration precisely, and perform our actions similarly. After all, timing was important for survival when responding to one’s surroundings—when fishing or hunting, for example, or when defending oneself against predators or other human beings. In today’s world, the issue is more likely to involve timely response to traffic. Music and sports are other spheres where temporally exact sequences are necessary. Here, fractions of a second can determine the success or failure of an operation. Even when we do not pay attention to time, the brain is constantly processing information about the duration of environmental stimuli, especially when recurrent events are at issue. If this fails to occur with a certain level of precision, a dancer may well step on his partner’s toes. When scientists researching temporal perception investigate the capacity to discriminate between two short tones, they tap into a finely honed system of perception and motoric operations that is focused on processing temporal duration experienced in the “now.” The same is made clear by research on how people perceive time conducted by means of imaging technology such as functional magnetic resonance tomography (fMRT), which makes brain activity visible: when subjects lying in the fMRT scanner are asked to distinguish between different temporal intervals, areas of the brain that are involved in motoric planning and execution display heightened activity (see figure 6).

The French psychologist and time researcher Paul Fraisse (1911–1996) distinguished between perception and the estimation of time.¹⁶ When time is perceived, temporal durations up to about three seconds are processed as a related whole. If the duration of a stimulus exceeds three seconds, the event becomes too long to be perceived as a temporal whole; then, duration must be assessed with the aid of short-term memory. The duration is no longer experienced as a total event of “now.” A few seconds ago, a note was struck, but it can still be heard. The beginning, which happened three or four seconds before now, is no longer present, yet the memory function still retains it. Such circumstances make it clear why the accuracy of temporal evaluation decreases at intervals lasting more than three seconds: because the interval is no longer temporally present as a whole, the operations of short-term memory are required to assess duration.

Figure 6

Brain activity of a subject performing a task to determine duration discrimination with one-second stimuli, compared to the state of rest. The basal ganglia (left) and cerebellum (right) are especially active—regions that play a pronounced role in controlling sequences of action.

Even if our estimations of durations longer than three seconds prove increasingly inaccurate and variable, we still have a sense of this longer duration. What is the maximum interval that we can experience with full awareness—that is, focus on attentively—as continuous and uninterrupted? Researchers can instruct subjects to press a key every hour, in other words, to produce a series of one-hour intervals. The intervals that are produced in this way differ markedly from physical time, but the order of magnitude still proves correct. Despite this, when the period lasts an hour, we can no longer say that people are having a continuous experience. Many things can happen over the course of an hour; attention is constantly drifting. One needs to remind oneself over and over that one is supposed to estimate how long it takes for the hour to pass. Ultimately, it proves a matter of making estimates on the basis of all that has been experienced—guessing the duration of activities that occurred during the time that has lapsed (“It took about twenty minutes to wash the dishes”). Memory constructs the duration of the hour that has gone by. Or, put somewhat differently: the mechanism for assessing time does not “tick” for an hour (or operate in any comparable fashion). There must be a natural limit up to which we can perceive temporal duration as a continuous and coherent occurrence. Simply cooking a three-minute egg probably already takes us to the outer limit of temporal perception, if not beyond. On a Sunday morning, we can test our abilities by means of the consistency of a three-minute egg, which reveals the imprecision of our estimates. It proves extremely difficult to focus attention on the pot for three minutes. Associative thoughts arise, which keep us from concentrating on the time. Indeed, there is great danger that something important will come up, causing us to get back to the egg too late. All the same, this little exercise in concentration may succeed—provided that one overcomes boredom. Someone who is practiced in meditation will have an easier time managing the task. People who meditate regularly are used to maintaining attention over a longer span of time—to being mentally present (see chapter 3).

The length of our short-term memory sets a natural limit for our continuous experience of temporal duration. When one tries to remember a telephone number, it is quickly forgotten (within a few seconds) if one does not repeat it to oneself. Working memory is involved when different things must be processed during the span of short-term memory—for example, when one is reading a text aloud but must also retain a series of numbers. Systematic experiments have demonstrated that subjects are no longer able to remember the greater part of new numbers or letters after just twelve seconds, if they are kept from actively repeating them. Working memory continues to function over longer spans of time, but it seems that the greatest loss of conscious content—at any rate, abstract content—occurs within ten to twelve seconds. A fixed upper limit cannot be drawn; it depends on what is supposed to be retained, as well as context. Still, a minute seems to be a generous estimate—half a minute might be more realistic.¹⁷

As far as the contents of consciousness are concerned, and in the context of working memory, we can speak of a “sliding window”: we are always experiencing new things as time passes; simultaneously, however, we forget what has just happened. Experience relies on this temporal window of consciousness, our mental presence. The boundary conditions of working memory also set an upper limit on how long we can experience temporal duration continuously, without interruption. Accordingly, the mechanism of perception related to the continuous experience of spans of time—which still has not been discovered—is restricted to a range of seconds up to just a few minutes.

As we have seen, researchers have not yet found an internal clock underlying our perception of time in the range of seconds to minutes. All the same, there is an internal clock; it exists in organisms from single-cell organisms up to human beings, and it steers physiology and behavior in cycles of twenty-four hours. This mechanism corresponds to the change between light and dark that occurs on the Earth over a single day: in this temporal rhythm, algae move between different depths of the ocean (alternately absorbing nutrients at lower levels and undergoing photosynthesis above, closer to the light), plants open and close their blossoms, and mammals rest and wake. Circadian rhythm (Latin: circa = about, dies = a day) is evident in many biological functions, including gene expression. In human beings, periodicity can also be demonstrated in moods, thoughts, and the perception of time. For example, the expression of feelings, the capacity for mental calculation, and reaction time are subject to systematic fluctuations over the course of a day.¹⁸ Many cognitive operations—that is, the speed and exactness of thinking—improve before noon; at midday, they prove optimal. Incidentally, this is a reason not to schedule core subjects at school or important deadlines between eight and ten o’clock in the morning. In turn, the capacity for performance decreases over the course of the early afternoon.

Circadian rhythms are endogenous: one or several pacemakers in the organism regulate periodicity over the course of a day. This occurs even without the influence of the sun. Plants that extend their leaves to the sun in a daily cycle also do so when one puts them in a dark room. The endogenous nature of the internal clock has also been demonstrated in human subjects. During the 1960s, in the legendary “bunker” experiments conducted by Jürgen Aschoff at the Max Planck Institute for Behavioral Physiology in Erling-Andechs, volunteers were locked up for three to four weeks (they included students who wanted to study for exams).¹⁹ Even without any form of social contact that might indicate daily routine (needless to say, wristwatches had been removed), and without the natural change from day to night (there was only electric light, which subjects could turn off or on at will), patterns continued as before: sleeping, waking, fluctuations of body temperature and other bodily functions. However, one difference from the cycle of twenty-four hours in a natural setting was noted. Endogenous rhythms had longer period durations—about twenty-five hours, on average. That is, under conditions of isolation, the subjective day—getting up, going to bed, and all other bodily parameters—took a little longer. Moreover, the phase differences between individual parameters entailed a consequence: whereas different rhythms normally occur in conjunction with each other—that is, they are phase-connected—in the “bunker,” desynchronization emerged over time. Normally, for example, body temperature is lowest in the early morning and highest in the early afternoon; in this way, it proves tightly connected to the sleep-wake cycle. Under conditions of isolation, however, these parameters became unlinked and diverged temporally.

Separate studies that focused on temporal perception also revealed fluctuations of daily periodicity.²⁰ During waking periods, subjects were asked to press a button whenever they felt that an hour had passed; that is, they were to mark intervals of an hour individually. In addition, either before or after pressing the button to mark an hour, they were instructed to press another button for precisely ten seconds, so that their accuracy in producing a short temporal interval could be measured. As it turned out, the circadian fluctuation that occurred in producing an interval of ten seconds was tied to the daily cycle of body temperature: when body temperature was highest, the shortest interval resulted. If we use the pacemaker-counter model discussed above, elevated temperature leads to a heightened frequency of the pacemaker; this entails shorter durations when temporal intervals are produced, because the number of pulses representing temporal duration is achieved more quickly—further experimental proof of what Hudson Hoagland observed when his wife was sick with fever.

Another picture emerged when participants were asked to identify one-hour intervals: the fluctuations that were measured over the course of the day (the subjects produced markedly different intervals, varying between one and almost three hours) correlated with time spent awake, but not with body temperature. These findings show, yet again, that our ability to identify temporal intervals of different lengths—in this case, the range between a few seconds and an hour—relies on different mechanisms.

Under more normal conditions of life, the many clocks in the body operate independently, but through pacing effected by light, they achieve synchronization into a stable rhythm that lasts for twenty-four hours. The experimental conditions of the “bunker” demonstrated how important light is for stable periodicity and the synchronization of body rhythms. Other studies have shown that light functions as a pacemaker for mammals, providing the exact rate of timing. Change between light and dark in the environment is registered via the suprachiasmatic nucleus (SCN), a small region in the hypothalamus located above the optical nerve; this ensures the synchronization of the separate physiological rhythms of organs.²¹

We achieve heightened awareness of the steering function that our internal clock performs after intercontinental flights, when we have crossed many time zones. Under these conditions, the timing of our internal clock, which is still “set” to home, and the course of the sun at the new location are phase-delayed with respect to each other. During the day, one is drowsy and incapable of thinking clearly; at night, one sits wide awake in bed. Here, moreover, the rule of thumb for adapting to a new time zone applies: be active outdoors and stock up on as much daylight as possible in order to let the body know what time it is and adjust to the new setting.

Some people go through this kind of phase delay on a daily basis. As holds for all human qualities, circadian rhythm varies from person to person. Accordingly, the internal clock does not “tick” the same for everybody; the rhythm of sleeping and waking can differ markedly between individuals. A spectrum of chronotypes exists, which extends from people who rise extremely early (“larks”) to those who get up extremely late (“owls”). They all display a rhythm that takes twenty-four hours, but high and low points of activity are phase delayed. Early risers grow tired more quickly in the evening, and they go to bed sooner; late risers still feel wide awake at night, and they go to sleep later.²² This leads to problems for people who get up extremely late. All human beings, irrespective of chronotype, need eight hours of sleep (give or take an hour). Because society is calibrated to the chronotype of the early riser—school and work start between eight and nine o’clock, at the latest, and sometimes even earlier (e.g., hospitals and bakeries)—even a moderately late riser must live at odds with his or her inner clock. In this context, Till Roenneberg, a chronobiologist at the Institute for Medical Psychology at the University of Munich, speaks of “social jetlag.”²³ Late risers experience desynchronization between their inner clocks and societal structures. Their natural rhythm of sleeping and waking makes it impossible for them to rest early at night; then, the alarm clock wakes them up too soon. For this reason, a sleep deficit accumulates over the course of the week, which can only be compensated in part by sleeping in on the weekend. Research shows that late risers drink more caffeinated beverages over the course of the day (to stay awake) and more alcohol in the evening (to have an easier time falling asleep). These modes of behavior represent typical forms of a self-medication. On the whole, circumstances prove less favorable—in terms of subjective well-being and quality of sleep—for late risers than for early risers.

Social jetlag assumes even more dramatic dimensions for adolescents: for developmental reasons, they are extremely late risers; only after the age of twenty do the interindividual differences emerge that situate adults somewhere on the continuum between early and late risers.²⁴ It follows that young people—who, for the most part, have to get up before seven o’clock to go to school—are still in a phase that, in physiological terms (i.e., according to their internal clock), amounts to the depths of night. And just a short time later, they’re supposed to cram vocabulary and solve equations. Needless to say, this does not mean that scholastic problems can simply be ascribed to conflicts between biological disposition and cultural expectations. Many factors play a role. Although direct connections can be demonstrated between late chronotypes, subjective disposition, and behavior, they are not that pronounced. All the same, research presents a clear picture: a connection exists between chronotype, on the one hand, and forms of behavior to be viewed as self-medication, mental exhaustion, and depression, on the other. Large-scale studies have shown that young late risers, in particular, drink more caffeinated beverages and more alcohol; it is also more likely that they will be smokers. A “small” change might produce a dramatic effect. A long-term study performed in the United States in 1997, which involved shifting the start of classes from 7:15 to 8:40, showed what would happen if school began later:²⁵ on the whole, these pupils were less sleepy, more alert, and less depressive—and even had better grades over the medium term—than peers who were not permitted to make the adjustment. Because of the complexity of social interests (e.g., those of teachers, the work schedules of parents), it is probably not realistic to demand that schools start at nine o’clock. All the same, it should be possible not to schedule core subjects before ten.