Depends on what grain and standard you're using. If you're going down to the level of neuroscience, the actual "scanning" part of our sensory equipment works on a tiny timescale. On the other hand, if you're looking at the the ongoing "model of the world" that the brain's juggling as it goes, which involves memory, it's something like 7 seconds or so? Forget the exact figure, but it's something of that order.

Isn't the past/present/future thing a bit artificial?

What we call the present--Isn't it really just the recent past and immediate future? No point trying to quantify its duration, as if it were a real distinct division into actual different periods..

Michael Ossipoff

What we call the present--Isn't it really just the recent past and immediate future? No point trying to quantify its duration, as if it were a real distinct division into actual different periods.. — Michael Ossipoff

If you take the question to be about the maximum rate of change, then it makes sense.

So the present is commonly understood as the extent of the moment between the past and future. It is the instant when everything actually "exists" in some non-changed fashion. Things are momentarily fixed, suspended between a past that is some evolving history that is the causes of the events happening in the present, and the future where there are further possible changes, but those have yet to be actualised.

The present is thus our measure of actualisation or realisation. And it is imagined as being rather statically existent. There is a duration in which the actual is what is, and nothing else is changing. Then this actuality gets swept out of the present and into history once it itself becomes a cause of further actualities, the cause of further possibilities becoming realised.

So the question is really about how long does actuality endure in that present tense gap between first becoming stably real as an effect, a crystalisation of what had been a future possibility, and then stably real as itself a cause, or the now historic reason for further actualities.

Hierarchy theorist Stan Salthe dubs this the "cogent moment". Henri Bergson had a similar idea.

If the world is understood in terms of a hierarchy of processes, then they all will have their own characteristic integration times. Time for the Cosmos is not some Newtonian dimension. It is an emergent feature of being a process as every process will have a rate at which it moves from being just starting to form a settled state - reaching some sort of cogent equilibrium which defines it as having "happened" - and then being in fact settled enough to become the departure point, the cause, for further acts of integration or equilibration.

So this view of time sees it not as a spatial line to be divided in two - past and future - with the present being some instant or zero-d point marking a separation. Instead, time is an emergent product of how long it takes causes to become effects that are then able to be causes. For every kind of process, there is going to be a characteristic duration when it comes to how long it takes for integration or equilibration to occur across the span of the activity in question.

We can appreciate this in speeded up film of landscapes in which clouds or glaciers now look to flow like rivers. What seemed like static objects - changing too slowly to make a difference to our impatient eye - now turn into fluid processes. They looked like chunks of history. Now we see them as things very much still in the middle of their actualisation. They will be history only after they have passed, either massing and dropping their rain, or melting and leaving behind great trenches etched in the countryside.

So the present is our intuitive account of the fact that causes must be separated from their effects, and the effects then separated from what they might then cause. There is some kind of causal turnaround time or duration - a momentary suspension of change - that is going to be a physical characteristic of every real world process. Thus there is some rate of change, some further "time frame" or cogent moment, that gets associated with every kind of natural system.

At the level of fundamental physics, this turns out to be the Planckscale limit. Time gets "grainy" at around 10^-44 seconds. The Planck distance is 10^-35m. So the Planck time represents the maximum action that can be packed into such a tiny space - the single beat of a wavelength. That primal act of integrated change - a single oscillation - then also defines the maximum possible energy density, as the shortest wavelength is the highest frequency, and the highest frequency is the hottest possible radiation.

So the shortest time, the smallest space, and the most energetic event, all define each other in a neat little package. Actuality is based on the rate at which a thermal event can come together and count as a "first happening" - a concrete Big Bang act of starting to cool and expand enough to stand as a first moment in a cosmic thermal history.

Then psychological time for us humans is all about neural integration speed. It takes time for nerve signals to move about. The maximum conduction speed in a well-insulated nerve, like the ones connecting your foot to your brain is about 240 mph. But inside the brain, speeds can slow to a 20 mph crawl. To form the kind of whole brain integrated states needed by attentional awareness involves developing a collective state - a "resonance" - that can take up to half a second because of all the spread-out activity to become fully synchronised.

So there is a characteristic duration for the time it takes for causes to become the effects that are then themselves causes. Input takes time to process and become the outputs that drive further behaviour. Which is why I mention also the importance of bridging this processing gap by anticipation. The brain shortcuts itself as much as it can by creating a running expectation of the future. It produces an output before the input so that it can just very quickly ignore the arriving information - treat it as "already seen". It is only the bit that is surprising that then takes that further split second to register and get your head around.

But between this physical Planckscale integration time and this neural human information processing time are a whole host of other characteristic timescales for the processes of nature.

Geology has its own extremely long "present tense". Stresses and strains can slowly build for decades or centuries before suddenly relaxing in abrupt events like earthquakes or volcanoes.

Here is a good visual chart of the integration time issue in biology - http://www.cell.com/cell/pdf/S0092-8674(16)30208-2.pdf?_returnURL=http%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0092867416302082%3Fshowall%3Dtrue

The philosophy of time is still very much hung up on the old Newtonian model, where time is some global spatialised dimension, and St Augustine's psychological model of time, where it is now somehow all a subjectively-projected illusion.

There is some truth in both these views. The brain does have to construct duration. The Cosmos does have a characteristic global rate in terms of its thermal relaxation - the one described by c as a Planck constant.

But a process view explains time in a more general fashion by relating it to the causal structure of events. Every system has some characteristic rate of change. There is a cogent moment graininess or scale created by the fact that not everything can be integrated all at once. It requires "time" to go from being caused to being a cause. There is a real transition involved. And that happens within what we normally regard as the frozen instant when things are instead finally just "actual". That is, brutely existent and lacking change, not being in fact a transition from being caused to being a cause in terms of our multi-scale accounts of causal flows.