Today in Uncertainty Wednesday we are coming to the last limit on observations: their impact on reality. As a quick refresher at the beginning of the series I introduced the framework in which uncertainty is the result of limits on our ability to observe and explain reality. Since then we have been working through various limits on observations, starting with a foundational limit, the resolution of observations, measurement error and cost of observation (which following a reader comment I will refer to as coverage, since coverage, resolution and measurement error all drive cost). So far we have assumed that the observation does not by taking place alter the underlying reality. But there are important cases where the limit on our observation is that it creates a feedback into reality.

In order to have any kind of observation, there must be a transfer of information from reality to our measurement instrument. In the case of simply looking around us that happens by light hitting our eyes and sound hitting our ears. Now we might at first conclude that this doesn’t impact reality because after all the light was already being emitted by the object and the sound was already in the air. But here is an easy way to see that it is not so simple: imagine wearing a bright orange sweater and looking at a nearby car with a silver paint job. Depending on the angle, the paint may look slightly orange to you in some spots. Why? Because some light was reflected from your sweater onto the car and then back to you. So your own presence did in fact cause light to hit the object that wouldn’t otherwise have hit it (for a more extreme version of this example: what do you see when you observe a mirror?).

So we begin to see that limits on observations arising from the impact of information transfer on the underlying reality might be quite pervasive. The reason we often ignore them is that even though they exist we take them to be small. But there are at least two well known situations in which the effects are large: the quantum realm and beings who can adjust their behavior in response to being observed.

When the real object that is subject to our inquiry is small, say a single atom, then even transferring a single bit of information from it will alter its state (and possibly significantly so). This has to be so because while we think of information as something abstract, there is no actual embodiment or transfer of information possible without a corresponding physical process. Note that this is very different from saying that the underlying reality is random in and of itself (we will get to that question, which is still subject to super active debate). All we are saying right now is that one reason uncertainty exists about quantum systems is that our observations of them are necessarily limited because the act of observing (transferring information) changes reality materially.

Another area where the limit on observations is material is with any being that can adjust its behavior to the presence of observation. This is not just humans who will often behave quite differently the second they suspect that they are being observed. For instance, a number of conclusions about chimpanzee behavior have come under question because the presence of the researchers (including their own food supplies) may have materially altered the behavior of the apes.

In summary then: observations requires the transfer of information, which is a physical process. Such a process by definition is connected to an changes the underlying reality. Depending on the situation that may materially change reality and can impose a substantial limit on our ability to observe. That in turn is a source of uncertainty about the reality when it is not observed.

Next Wednesday we will turn to explanations and their limits. Eventually we will also look at the question of whether reality itself could be a source of uncertainty.