SHOUT for Quanta Magazine

Our brains have an extraordinary ability to monitor time. A driver can judge just how much time is left to run a yellow light; a dancer can keep a beat down to the millisecond. But exactly how the brain tracks time is still a mystery. Researchers have defined the brain areas involved in movement, memory, color vision and other functions, but not the ones that monitor time. Indeed, our neural timekeeper has proved so elusive that most scientists assume this mechanism is distributed throughout the brain, with different regions using different monitors to keep track of time according to their needs.

Over the last few years, a handful of researchers have compiled growing evidence that the same cells that monitor an individual’s location in space also mark the passage of time. This suggests that two brain regions — the hippocampus and the entorhinal cortex, both famous for their role in memory and navigation — can also act as a sort of timer.

In research published in November, Howard Eichenbaum, a neuroscientist at Boston University, and collaborators showed that cells in rats that form the brain’s internal GPS system, known as grid cells, are more malleable than had been anticipated. Typically these cells act like a dead-reckoning system, with certain neurons firing when an animal is in a specific place. (The researchers who discovered this shared the Nobel Prize in 2014.) Eichenbaum found that when an animal is kept in place — such as when it runs on a treadmill — the cells keep track of both distance and time. The work suggests that the brain’s sense of space and time are intertwined.

The findings help to broaden our understanding of how the brain’s memory and navigation systems work. Perhaps both grid cells and other GPS-like cells aren’t tuned only to space but are capable of encoding any relevant property: time, smell or even taste. “It probably points to a broad thing the hippocampus does,” said Loren Frank, a neuroscientist at the University of California, San Francisco, who studies memory and the hippocampus. “It figures out the relevant axis for encoding experiences and then uses the cells to map those experiences.”

Pastalkova, E et al Science 2008

Different cells fired at different times during the 15-second stretch that rats ran on a wheel. Each row represents activity of an individual neuron (red indicates peak activity).

These maps in turn construct a framework for memory, providing an organizing system for our never-ending series of past experiences. “The hippocampus is this grand organizer of memories in space and time,” Eichenbaum said. “It provides a spatiotemporal framework onto which other events are applied.”

Time Tiles

To study how the hippocampus monitors time, scientists train rats to run on a wheel or tiny treadmill. This setup holds the animal’s location and behavior constant, so that researchers can focus on the neural signals linked to time. (Rats are too fidgety to sit still, so running helps standardize their normally twitchy behavior.) Electrodes implanted deep in the brain record when different cells fire.

In Eichenbaum’s experiments, a rat runs on the treadmill for a set period — say, 15 seconds — and then gets a reward. As the animal repeats the cycle over and over, its brain learns to track that 15-second interval. Some neurons fire at one second, others at two seconds, and so forth, until the 15 seconds have elapsed. “Each cell will fire at a different moment in time until they fill out the entire time interval,” Eichenbaum said. The code is so accurate that researchers can predict how long an animal has been on the treadmill just by observing which cells are active. Eichenbaum’s team has also repeated the experiment, varying the treadmill’s speed, to make sure the cells aren’t simply marking distance. (Some of the cells do track distance, but some seem linked solely to time.)

Although these neurons, dubbed “time cells,” are clearly capable of marking time, it’s still not clear how they do it. The cells behave rather like a stopwatch — the same pattern of neural activity repeats every time you start the clock. But they are more adaptable than a stopwatch. When researchers change the conditions of the experiment, for instance by extending the running duration from 15 to 30 seconds, cells in the hippocampus create a new firing pattern to span the new interval. It’s like programming the stopwatch to follow a different time scale altogether.

Moreover, time cells rely on context; they only mark time when the animal is put into a situation in which time is what matters most. When other variables come into play, the same cells behave differently. Allow a rat to explore a new environment, for example, and these same cells will map themselves to space; a particular cell will fire every time the animal is in a specific location rather than doing so at a certain time.

Rohan Chitracar

Howard Eichenbaum, a neuroscientist at Boston University, is exploring how parts of the brain that map space can also track time.

The Brain’s Space-Time Matrix

Eichenbaum’s work dovetails with a 15-year trend in neuroscience research that suggests the hippocampus is more flexible than scientists expected. Researchers traditionally thought of it as a mapmaker — place-encoding cells were discovered 40 years ago — but growing evidence suggests it can encode other types of information as well. According to the newest picture, place cells can map not just space but other relevant variables. Time is one of them, but others are possible. For example, “a wine taster might have a space of wine tastes and smells,” Frank said.

But many scientists still view the hippocampus as a largely spatial structure. According to their argument, neural circuitry evolved to keep track of location, and everything else is just recorded on top of it. “The hippocampus provides a code that is fundamentally spatial in nature,” said Bruce McNaughton, a neuroscientist at the University of California, Irvine.

Eichenbaum’s findings challenge this viewpoint, but they don’t bury it. “What’s definitely clear is that place cells can represent information beyond place,” said David Foster, a neuroscientist at Johns Hopkins University. “But what’s less clear is whether they can code for the pure passage of time.”

In the timed treadmill experiments, the rats appear to be doing something very much like counting. But are these cells marking the passage of time itself, or are they responding to something else that merely looks like time? “We don’t know the driving principle that tells cells to fire at a specific point, but I don’t think it’s time,” said Eva Pastalkova, a neuroscientist at the Howard Hughes Medical Institute’s Janelia Research Campus in Ashburn, Va. “It’s not precise enough; they are not like ticking clocks.”

György Buzsáki, a neuroscientist at New York University’s Neuroscience Institute whose lab did some of the first experiments exploring how the hippocampus tracks time, proposes that rather than monitoring time itself, these cells are doing something else — remembering a path through a maze or plotting the animal’s next move. Both memories and future plans unfold in time, so time cells may simply reflect this mental activity.

“That’s the number-one problem for me: Are there dedicated neurons in the brain doing nothing else but keeping track of time?” Buzsáki said. “Or do all neurons have functions that happen in sequential order, which for the experimenter can be translated into time?”

Buzsáki points out that it may not even make sense to think of hippocampal cells as independently coding for space or time. The human brain often considers time and distance interchangeably. “If one asks how far New York is from LA, the answers you get vary: 3,000 miles, six hours by flight,” he said. “In older language, distances were typically given by time — the days it takes to go from one valley to another — since it was not distance but the number of sunsets that was easy to calculate.”

For Buzsáki, the issue goes beyond neuroscience and reaches into physics. Physicists consider space-time as a cohesive, four-dimensional entity, a fabric upon which the objects and events of the universe are embedded. “Neuroscience must converge back to the old problem of physics: Are there place and time cells? Or is there only a single time-space-continuum representation in the brain?” Buzsáki said.

Memory Maps

Eichenbaum is less concerned with these abstract questions. His goal is to unpack the role time plays in forming memories. “When you recall what you did this morning, you remember events in the order in which they occurred,” he said. “How does the hippocampus organize memories in time?”

People with damage to the hippocampus often can’t make new memories — the famous patient H.M., who had a lobotomy to remove much of that part of the brain, introduced himself to his doctor over and over again each day. But these patients also have trouble remembering the sequence of words or objects presented in an experiment. “How does the hippocampus support the ability to remember the temporal order of a sequence of events?” Eichenbaum said.

Eichenbaum envisions time cells as providing a timeline onto which sequential events are attached to represent an experience. If memories are a movie, he said, time cells are what puts the individual frames in order. His team is planning experiments that will intersperse time delays with different events, to see how time cells modify their code to remember the order in which the events occurred. “I don’t think the hippocampus is a clock,” he said. “But it’s using a clock to map out when things happened in a memory to keep them in order.”

This article was reprinted on,, and

View Reader Comments (15)

Leave a Comment

Reader CommentsLeave a Comment

  • A non-scientist, I have often thought of memory, and place and time keeping, as part of what I've called the holographic function of the brain. Individual areas for various sense-interpretations, the hippocampus for time/geographic localization: taken together, sense impressions, particularly the stronger "memorable" events, imprint as unique fingerprints or holograms, basically because event A has certain unique characteristics which activate a unique set of neutrons, and then, when we encounter a similar pattern (for instance, a smell), the original or something approaching the original pattern will be triggered, thus: recall. We can voluntarily, consciously, bring up a reconstruction by recalling, for instance, a subset of the hologram: a visual clue (photograph) or a sound or even just replicating some abstract set of triggers in our imagination, which will then trigger the holographic recall. Or am I totally off-base? I mean, I'm just a bus driver.

  • Buzsáki points out that it may not even make sense to think of hippocampal cells as independently coding for space or time. The human brain often considers time and distance interchangeably. “If one asks how far New York is from LA, the answers you get vary: 3,000 miles, six hours by flight,” he said.

    You dont have to get very exotic to do that, any analog clock display is just a way of converting distance around a dial to time.

  • In many ways I don't find this surprising, after all time is just another dimension and as the article suggests we often treat it as having properties such as length and distance. We speak of a 'long' time and a 'short' time, so maybe it would actually be odd if the brain did handle the concept of the time dimension differently to the spatial ones.

    When something is moving we just perceive it to be at different locations in space and that is how we can judge speed and distance so it makes sense that the same part of the brain deals with it all.

    In fact you can't really think about time, distance and speed in any other way than by using two of them to define the other. If ther was no time nothing would move and if nothing moved there would be no concept if time.

    I think it makes total sense that the brain does interpret the four dimensions all in one place and from that mechanism or process comes our perception of the world as a 3+1 (space and time) entity. Nice to see evidence for it.

  • have they tried linking these while subject is awake & asleep? I remember multiple dreams a night & note time between awake periods. The mind is much faster asleep than awake.. considering what I do in my dreams.

  • I am an alum of the Eichenbaum lab and a current member of the Buzsaki lab. In my mind there is a natural bridge between

    "That’s the number-one problem for me: Are there dedicated neurons in the brain doing nothing else but keeping track of time? … Or do all neurons have functions that happen in sequential order, which for the experimenter can be translated into time?” -Buzsaki


    "It [time cells] provides a spatiotemporal framework onto which other events are applied.” -Eichenbaum

    The key is that the hippocampal sequences indeed reflect the processing of other kinds of information. But not only the experimenter can translate these sequences to an estimate of elapsed time. Any other brain regions that listen to these sequences would be able to decode that information as well. Marc Howard has some nice models for how this could be done. Experiments still need to be done to link 'time cells' with temporal memory or temporal estimates, but there are certain physiological properties of time cells which are consistent with the psychophysics of temporal perception and memory, such as Weber law decreases in temporal precision with elapsed time for both physiology and cognition.

    Well written article. Thanks!

  • interesting – i do not have good long-term memory, as a rule, unless i've read something. it's like my memories are beads, and they're jumbled together in a bowl. the beads do not have time stamps on them, and each bead is an image, a tableau, rather than a video. unless there's something in the image to give me a hint as to date or time, i have no idea when it happened. *However*, i have excellent directional sense and once i've been somewhere, i rarely get lost getting to or from that location. go figure!

  • I think a concept like "situational memory" could bring time and space aspects into a single framework. The hippocampus, I suspect, has mechanisms for encoding situational memory in such a way that relevant neurons can be subsequently sorted in a post-hoc fashion for short-term and long-term memory formation (and in higher mammals, narrative construction and conscious recall). There might be a kind of structural vocabulary for these situational memories that allows them a recombinant potency often associated with language; e.g. for anticipation based upon similarity of current situations and memorized situations; likely transformations of situations over time based upon memorized temporal sequences; etc. I don't see the time vs. space designation being as productive as the situational perspective that inherently combines temporal and spatial aspects through mechanisms that permit cross-referencing, anticipation, recombination, and post-hoc sorting.

  • They should study musicians who mark time to the music they play. This has to be very accurate and even coordinated in ensemble play.

  • In his "Critique of Pure Reason," which is a combination of rationalism and empiricism, Kant differentiates time and space in the following way: 1. Different times are not coexistent but successive; 2. Different spaces are not successive but coexistent.
    Furthermore, he defines both as conditions of both empirical and ideal perception, via the 'sensual manifold,' differentiating 'space' as a condition of external sensibility (through the five-plus outer senses) from 'time' as a condition of internal sensibility (through the inner sense.) The latter is then 'common sense' or making sense of the senses and of ourselves and the beginning of consciousness.
    A 'map' is traditionally defined as an ordering of space, not of time. Therefore, there is a lack conceptual clarity here and undifferentiated mixing of perceptual categories.

  • Anyone who who takes this article too seriously will be mislead. Zeno of Elea proved that physical time doesn't exist. If time were composed of infinite instants, it would be impossible to cross finite space in infinite time. If time was composed of finite instants then nothing could move in a single instant, because, if it did, it wouldn't be an instant. Einstein proved this with special relativity, which necessarily implies that space-time is a block universe in which our consciousness moves in one direction, taking us with it. The Wheeler-DeWitt Equation which describes the quantum state of the universe lacks Time. Therefore everything is frozen. Nothing moves! Back to Zeno. Conclusion? We are all travelers in time on a journey through eternity. Death is just a stopover. Read Psalm 139 if you want to understand time. As for science, it is an irreligious attempt to counterfeit God's Creation, consequently, its theories contradict logic, and will continue to do so. Theories are excellent for mimicking the mind, but to suppose that the brain creates the mind contradicts the logic of time. As the great physicist Sir James Jeans concluded, the universe is more like a great thought than a machine.

  • The comment by John T reminded me that, many years ago, I conducted the local youth orchestra. One day I was warned that a well known conductor would be there for a concert. Afterwards, when I was introduced to him, he commented on the fact that, despite significant variations in tempo, the last piece had ended up at exactly the same speed as that at which I'd started it. I expressed surprise that he hadn't expected that: wasn't that what should happen? I certainly wasn't conscious of needing any special effort to achieve it.

  • From a purely personal perspective, I have damage to my hippocampus (rendering me epileptic) and one of the consequences is that I cannot sequence my autobiographical time line very well. Was event A before or after event B? That's a hard question for me. I also experience a feeling of temporal disonance. That is, I can't easily connect to the idea of having a 'past' or having a 'future'.
    So, the hippocampus does seem to be involved in temporal activities.

  • Time is just a repetitive motion. Its also based on counting. It requires knowledge of distance of objects and its CHANGE or motion, because without knowledge of distance we cannot measure motion. When motion becomes repetitive, we call it time. All these position (distance), motion (change in distance), and repetitive motion (time) etc. are interrelated.

    Time must be also related to memory. There must be link between a event (a particular unique change happened) with a repetitive change (time), which allows us to recall based upon the time. E.g. Morning, afternoon and night are repetitive, so if we link the afternoon with writing this article (which is unique), then you will always remember that you wrote the article in the afternoon.

    So we link things and store the data. If the data is exactly same (i.e events linked in it all the same), we cannot distinguish it. If events linked are different, than we can distinguish it from one another.

  • I am pharmacist interested in neuroscience.
    I would rather try to see if neurons fire in certain ways and work as networks.
    In physics spacetime is considered as emergent phenomena and this emergence is linked to entanglement, which make network.
    I believe the same representation may take place in hippocampus.
    Is there any cross talk between other hippcampus cells in the way that time cells would fire in different way depending on other neurons activity and such activity would identify how to fire according to what rats are doing?

  • The idea of time is used to explain the experience of things changing. It is not a 'thing' in itself. As all our subjective experience of a universe, including ourself as part of it, is a neurological construct that we can't leave, it uses spacetime as a framework so that we don't experience everything in the same place and at the same time. My prefered understanding of time is that there is only an ever present now, not a past present and future.

Comments are closed.