Can we directly measure something without changing it?

[KingNothing]

Can we really measure a physical property of something, without changing it in any way at all?

• In a circuit, you cannot use a voltmeter to measure a voltage without changing the circuit. The change may be small or negligible in most cases, but it is there.
• If you use a thermometer to measure the temperature of a sample, the thermometer has had to absorb/dissipate some heat from/to the sample in order to read a change on the thermometer, thus changing it.
• A common "measurement" procedure for measuring an electric field is to place test charges within it, and absorb the effect. However, the addition of these charges does in fact changing the local electric field, as charges have fields of their own
• Quantum cryptography is based entirely on this concept - that measuring the information (reading it) would change its value

The only counter-examples I can think of involve comparing the behavior of something to a "known value", such as holding a ruler up to an object or measuring mass with a balanced scale.

I believe there is a theorem (perhaps part of information theory) that essentially states that no transfer of information can occur without some transfer or change in energy. I don't remember what it's called or where to find it stated more eloquently.

 

[Physicist1231]

I was actually watching something about this in the quest for a Bose–Einstein condensate. They were trying to reach a temperature of absolute zero using lasers and trying to observe the reactions. The problem though is just the mere observation of the atoms involved light reflecting off of the atoms thus changing the desired results. That just makes it harder to have a completely controlled environment. I guess though that the mere knowledge of something is different... kinda hard to know without observing though.

 

[Bill_K]

Can we really measure a physical property of something, without changing it in any way at all?

Yes. Three examples:

You can measure the mass of an electron without changing it! Right now it is 0.51 MeV. Measure it again in 15 minutes and it will still be 0.51 MeV. Some things are constants of nature and for that reason can never change.

Now measure the spin of the electron. Let's say it is spin up. Leave the electron undisturbed and measure the spin again in 15 minutes. It will still be spin up. The spin can change, but once it is in a pure state (an eigenstate) it will remain in that same state indefinitely for future measurements. (Again, this is assuming you don't somehow disturb the electron in between.)

Quantum nondemolition measurements. This is a much more interesting example. For a description see http://physicsworld.com/cws/article/news/30913.

 

[BruceW]

It depends on the definition of a "direct measurement". I would define a direct measurement as interacting with the thing being measured, which generally does involve changing the thing being measured.
So I would say the answer is no, due to my definition of "direct measurement".

 

[Drakkith]

You cannot alter the fundamental properties of a particle that only have 1 value, such as mass, charge, ETC just by observing it, but you CAN alter the other properties such as position, velocity, spin, ETC.

 

[coelho]

^^ So, in a way, the only properties we can measure without changing them are the ones that are constant (like rest mass, charge, etc ) or the ones that can assume only a known set of values (like energy levels of an atom, etc), which means that we already know a lot (if not all) about them even before the measurement.

Yet, the properties that we can know nothing (or not very much) about before measuring them are exactly the ones that are changed by the measurement, so we can never know what was their "actual" (unchanged by the measurement) value...

In short, it seems we only can know what we already know, and can't know what we don't know... ;)