In this edition of TOK-TALK we will explore if it is in principle possible to measure anything accurately. How does a measurement change the value of that what you want to measure? Listen to find out!

Here in front of me, I have a cup of hot water, and over here we have a thermometer. Let’s put the thermometer into the glass, we have to wait a bit for the temperature reading to adjust. For our listeners, it’s a digital thermometer with a metallic probe. You use similar thermometers to measure the inside temperature of a cake to check if it is finished baking.
Now the temperature is slowly climbing and leveling off. We have a reading of 76.2C. We can see that it is fluctuating a bit, now we can see 76.1C. Of course we have to be careful here, because this is the temperature reading of the side of the glass. The center is much warmer. Let’s stir the water a bit. OK, we are up to 82.5C.

We could agree that we want to measure the average temperature of the water, so we have to stir the water. But this would of course increase the cooling rate. But then there is the problem that we are measuring a value that is continually changing. The water is cooling down.

We could try to completely isolate the water from its colder environment. But we first have to bring up the temperature of the container to the same temperature as the water. Otherwise the container will cool down the water. The isolating container will prevent the cooling of the water, but first the container has to assume the same temperature as the water. So we have to determine the temperature of the container as well.

And the situation becomes even worse. Of course the thermometer itself also has a certain temperature. By sticking the thermometer into the water, we are already changing the temperature of the water. Every measurement changes the thing that we want to measure. We could of course try to compensate this error mathematically, but for that we have to know the exact temperature of the thermometer that we are using. So we need another thermometer to measure the temperature of the first thermometer. But then this measurement would of course change the temperature of the first thermometer just like the thermometer would change the temperature of the water. We are pushing the problem out.

We could make the mass of the water very large compared to the mass of the thermometer. If we use a large amount of water and also a large thermometer, then we do not gain anything. And can you imagine it if we attempt to use a very large thermometer to measure the temperature of a drop of water? The thermometer would influence the temperature of the drop more than the drop the thermometer. This is what we call the „Observer effect“. When we observe something, we end up changing the thing that we observe.

Besides temperature measurement, there are other examples as well. In my younger days I used to play around with electronics a lot. I liked to assemble radios and other electronic gadgets. I bought myself a nice voltmeter to measure volage, resistance and current of electrical circuits. The only problem was that the device, being electronic itself, actually influenced the object that I wanted to measure. The ammeter measured the current flowing through a cable, but the device of course consumed some of this current itself. So it changed the current that I wanted to measure. Of course manufacturers try to keep this effect small, but the effect does exist. The current measured must be substantially larger than the current consumed by the device.

The issue starts to become interesting when we want to measure the position or velocity of very small objects. How do we determine the position or speed of movement of a quantum particle, such as an electron? For measuring very small objects we need to have very small measuring devices. So we need a second particle, a photon, that we can use as a probe. We need shoot this photon at the particle that we want to measure, our target electron. The photon will interact with the electron and the photon will be deflected, it will bounce off. If we now measure the deflected photon, then we will gain some information about the nature of the electron. There is a problem, however. The photon that we shot at the electron has changed the position and movement of the electron.

The question now is, how we measure the deflected photon. Maybe you have already guessed it. We need to let it interact yet with another particle! So we are again encountering a similar problem. We are pushing out the problem.

There is a actually a practical use to the observer effect. It is possible to encode messages using quantum cryptography. When an encrypted message is passed from person Alice to Bob and a third person, an intruder, tries to intercept and read the message, then the message is instantaneously modified. The reading of the transmitted message requires a measurement of the message, and this changes the message. The communicating parties Alice and Bob will then be able to know that somebody tried to eavesdrop on their communication.

So far I mentioned the measurement of temperature, current and particles. But the „observer effect“ is not limited to physics. Imagine a zoologist who wants to study the behavior of wild animals in their natural environment. The animals will not behave naturally anymore when they sense the presence of an intruder. The more closely the zoologist wants to study the animals the closer he has to come to the animals. Something similar exists in the social sciences, it is called the „Hawthorne Effect“ in Psychology.

The Hawthorne Effect states that the observed people will not behave naturally if they know that someone observes them. I like to give extreme examples to make the point clear. Imagine a company boss telling his employees the following: „Today we will investigate how often you play computer games during your work time. This is an entirely scientific study, it will not impact on your salary. Please behave normally and play as frequently as always.“ Of course you can not expect any meaningful outcome of this study, even if the employees do not have to face negative consequences.

And I can give you one last example: I sometimes run a diagnostic software on my computer. I want to measure the perfomance of my computer, the memory that is used, the speed, and so on. The program that diagnoses my computer, however, also consumes some of these resources and it also slows the computer down. The program while monitoring my computer also changes the perfomance of the computer. So we are again changing the thing that we want to measure.

I just wonder, whether we can generally ever know the „true nature“ of things. But then I guess it is a question that our measurements and observations just have to be „good enough“ for our purposes. When I measure my body temperature to check for fever, I am not really worried about the fact that the cold thermometer cools down my arm pits and will therefore give a lower reading.

Now the thermometer is reading 71.5C and what does this tell us? This tells us that it is time for a little quote again. This one is from from Johann Kaspar Lavater, a Swiss philosopher and priest of the 17th century.

He alone is an acute observer, who can observe minutely without being observed. (Johann Kaspar Lavater)

Questions for Discussion:

• When our observations and measurements change the object being observed or measured, can we ever know the “true state” of the object? Can we ever know the “true nature” of reality?
• Can a “true nature” of reality exist in the first place without an observer?
• If I am watching the behavior of a group of people from the distance (and they do not know that I am watching them) then their behavior does not change, does it? So is it now possible to observe without changing the the behavior of the observed? How do we know?