Lab Talk

spontaneous activity noise

Why Do Brains Have Spontaneous Activity?

The brain spontaneously produces activity regardless of mental activity and stimulus input.  Current theories treat this activity as noise or simply ignore it. What does it do?

The way we usually conceptualize the functioning of the nervous system is as if it is exclusively an input response system: It produces output only if there is first an input to provoke a computation of the output. It is like a hand-held calculator: it makes no sense for a calculator to compute on its own. It should compute only when asked to do so.

Relating Brain Activity to Inputs

One peculiar thing about the brain is that it does not really behave like a hand-held calculator. If left alone without any input, the brain will still produce activity—a lot of activity. In fact, the brain is always doing something, no matter whether inputs are coming in or not. A difference in the amount of activity during the absence of any inputs and while inputs are incoming almost does not exist. This difference is so tiny that it can often be detected only by using refined experimental procedures requiring repeated trials combined with elaborate statistical methods.

There is also no strong correlation between the richness of your internal mental activity and the activity of the brain. For example, you may sit bored in front of a blank screen and then suddenly a picture may appear. This picture may make you think about something related. Or the picture may present a math problem and you may then be engaged in solving it. In that example, your internal mental life was first doing nothing, being bored, and then there was suddenly rich activity. One may think that there is a similarly big difference in the electrical activity of the brain or in its blood oxygenation signals. This is not the case. The difference is miniature. The brain may be no more than 10% more active while you are doing something mentally intensive as opposed to when you are just relaxing .

Activity is an intrinsic property of brain tissue

This is because of the extensive spontaneous activity. The spontaneous activity of the brain has been well documented and intensively studied. For example, spontaneous activity normally produces the statistics of avalanches: there are a lot of small events and only few big events (Beggs & Plenz 2003; Hahn et al. 2010).

See a related post Neuronal Avalanches:  What are They and What do They Mean?.

The brain generates spontaneous activity all the time. No matter at which place in the nervous system you stick an electrode, something will be going on. Even if you cut a small chunk out of brain (a slice) and place it in a petri dish, it will be active for as long as it is alive (e.g., Beggs & Plenz 2003). Spontaneous activity is a pervasive property of the brain.

see related post Brain in a Dish

Just noise or something else?

This raises an interesting question: Why does spontaneous activity (in EEG this is called ‘resting state’) exist in the brain in the first place? Is it something that has a function and purpose? Or is it some sort of a noise in the brain?

If spontaneous activity was just noise, then the brain would want to get rid of it. With less noise, the brain would operate better, more accurately. But if the spontaneous activity had a purpose, it would be a bad idea to remove the “noise”. Removing spontaneous activity would debilitate the brain’s functioning.

Interestingly, our mainstream theories on how the brain creates mental operations do not assign to spontaneous activity any significant computational role. In my previous posts I discussed how brain theories have problems explaining certain properties of working memory and perception. We can add one more problem to those theories: they do not incorporate spontaneous activity as in integral part of the computation. These theories tacitly assume that spontaneous activity is of no use to the brain—that it is just noise.

A phenomenon closely related to spontaneous activity (which is referred to as resting state in the EEG world) is response variability. If we present the same stimulus to the brain twice in a row, we will almost never get the same results. Each time the result will be slightly different. If the brain is not just waiting for a stimulus but is constantly doing something, it means that the brain changes its state. And changing states means responding differently to identical stimuli (Arieli et al. 1996).

If this is the case, necessarily, behavioral responses will vary. And this is exactly what happens. For us, it is nearly impossible to repeat twice an identical behavioral act. Try getting consistently a long distance (three-point) shot in basketball. If I did it once right, why can’t I just repeat exactly the same movement over and over again, and get the ball each time in? Not even superstar basketball players can do it. Similarly, why can’t professional golf players get their swings always perfectly right? Why can’t a tennis player serve an ace each time?

Is it the case that evolution simply could not come up with a better way for sending signals across the nervous system without ample amounts of debilitating noise? Are we being imperfect athletes because we are simply paying a price for quirks of biology? In my own studies, I found that, a single neuron cannot reliably provide information about the presented stimulus, but that a population of a hundred neurons or so can do a pretty good job (Nikolić et al. 2009). So, are we simply wasting resources due to an unreliable computational system within our heads?

There are reasons to believe that actually the brain does everything right. First, let us consider the evidence from anesthesia. These injected chemicals can reduce pain when applied locally and in small amounts, make us drowsy in larger amounts applied globally and can even completely remove our consciousness. If spontaneous activity is a residual and behaviorally irrelevant aspect of the system you would expect that it increases the amounts of spontaneous activity, but it does not  (e.g., Vogel et al. 1989). If anything, there is a small reduction in noise and spontaneous activity. Anesthesia does not produce its tranquilizing effects by adding noise to the system.

see related post Tracking Anesthesia

Moreover, we know that biological systems are capable of generating nearly noiseless neural activity when they need it. For example, accurate sound localization can be achieved only through sub-millisecond precision of timing of action potentials (e.g., Oertel, D. 1999). There are also other examples of high precision in the brain activity (Hasson et al. 2010). Therefore, the biology seems able to produce noiseless activity when there is a demand.

Useful or mysterious?

There is even more indirect evidence that the brain needs spontaneous activity for functioning. Data tell us that the higher we go in the brain hierarchy, the more “noise” there is in the neural activity. Subcortical areas are not nearly as noisy as the cortex is. And the higher we go in the cortical hierarchy, the more spontaneous activity and more variability in responses we find (Gur et al. 1997; Shadlen & Newsome 1998). What this means is that the more advanced mental operations a part of a brain is responsible for, the more “noisy” this part is. This suggests that what we see is no noise after all. By all likelihood, what we see in the form of spontaneous activity and variability in responses is a result of deliberate information processing that contributes to mental operations and our intelligence.

Neuroscience is in a need of a theory that can explain these empirical findings and account for spontaneous activity and the variability of responses in an encompassing way. The theory needs to tell us what is going on, why and how these activities occur and how they contribute, if at all, to our capacity to perceive, decide, act and think.



Arieli, A., Sterkin, A., Grinvald, A., & Aertsen, A. D. (1996). Dynamics of ongoing activity: explanation of the large variability in evoked cortical responses. Science273(5283), 1868-1871.

Beggs, J. M., & Plenz, D. (2003). Neuronal avalanches in neocortical circuits. Journal of neuroscience23(35), 11167-11177.

Gur, M., Beylin, A., & Snodderly, D. M. (1997). Response variability of neurons in primary visual cortex (V1) of alert monkeys. Journal of Neuroscience17(8), 2914-2920.

Hahn, G., Petermann, T., Havenith, M. N., Yu, S., Singer, W., Plenz, D., & Nikolić, D. (2010). Neuronal avalanches in spontaneous activity in vivo. Journal of neurophysiology104(6), 3312-3322.

Hasson, U., Malach, R., & Heeger, D. J. (2010). Reliability of cortical activity during natural stimulation. Trends in cognitive sciences14(1), 40-48.

Nikolić, D., Häusler, S., Singer, W., & Maass, W. (2009). Distributed fading memory for stimulus properties in the primary visual cortex. PLoS biology7(12), e1000260.

Oertel, D. (1999). The role of timing in the brain stem auditory nuclei of vertebrates. Annual review of physiology61(1), 497-519.

Shadlen, M. N., & Newsome, W. T. (1998). The variable discharge of cortical neurons: implications for connectivity, computation, and information coding. Journal of neuroscience18(10), 3870-3896.

Vogels, R., Spileers, W., & Orban, G. A. (1989). The response variability of striate cortical neurons in the behaving monkey. Experimental brain research77(2), 432-436.

2 thoughts on “Why Do Brains Have Spontaneous Activity?

  1. Hi Sue, in my future posts I plan to suggest some answers. They will not be based on fuzzy logic but on something else. Stay tuned.

Leave a Reply