DETERMINISM AND THE ROLE OF THE OBSERVER:

Quantum mechanics replaces the deterministic universe described by classical physics with a

probabilistic universe. This is the idea that the behavior and various properties of subatomic systems

and particles cannot be predicted precisely, that only a range of probable values can be specified. If

you roll a series of marbles at a hill at less than a certain critical velocity, all the marbles will roll

back down, and if you roll the marbles at more than the critical velocity, all the marbles will make it

over the hill. In our classical macroscopic world, either they all get over or they all fall back. Things

are not so simple at the quantum level.*16

For instance, if subatomic particles such as electrons are fired at a potential barrier at a given

velocity, it may not be possible to say with certainty whether an individual electron will pass through

the barrier. Fire the electrons at a low enough velocity and most will be reflected, although a minority

will pass through; at a high enough velocity most will pass through; and at some intermediate velocity

about half will pass through and half will be reflected. But for any individual electron (out of a group

of apparently identical electrons), all we can specify is the probability that the electron will pass

through.

Another example of quantum randomness is radioactive decay. Say we have radioactive uranium

isotope A that decays into isotope B with a half-life of one hour. One hour later, half of the uranium

atoms will have decayed into isotope B. By all the known methods of physics, all of the uranium

isotope A atoms appeared to be identical, yet one hour later, half have decayed and half are

unchanged. The half-life of isotope A is highly predictable in a statistical sense, yet the precise time at

which any individual atom decays is completely unpredictable.

Probability enters here for a different reason than it does in the tossing of a coin, the throw of dice,

or a horse race: in these cases probability enters because of our lack of precise knowledge of the

original state of the system. But in quantum theory, even if we have complete knowledge of the

original state, the outcome would still be uncertain and only expressible as a probability.

(Philosophers refer to these two sources of uncertainty as subjective and objective probability.

Quantum mechanics suggests that in some situations probability has an objective status.)

Another surprising proposition was that subatomic particles do not have definite properties for

certain attributes, such as position, momentum, or direction of spin, until they are measured. It is not

simply that these properties are unknown until they are observed, instead, they do not exist in any

definite state until they are measured.

This conclusion is based, in part, on the famous “two-slit” experiment, in which electrons are fired

one at a time at a barrier with two slits. Measuring devices on a screen behind the barrier indicate the

electrons seem to behave as waves, going through both slits simultaneously, with patterns of

interference typical of wave phenomena: wave crests arriving simultaneously at the same place in

time will reinforce each other, but waves and troughs arriving simultaneously at the same place will

cancel each other (interference patterns result when two wave fronts meet, for instance, after dropping

two stones into a pond). These waves are only thought of as probability waves, or wave functions, as

they do not carry any energy, and so cannot be directly detected. Only individual electrons are

detected by the measuring device on the screen behind the barrier, but the distribution of numerous

electrons shows the interference patterns typical of waves. It is as though each unobserved electron

exists as a wave until it arrives at the screen to be detected, at which time its actual location (the place

at which the particle is actually observed on the screen) can only be predicted statistically according

to the interference pattern of its wave function.

If, however, a measuring device is placed at the slits, then each electron is observed to pass through

only one slit and no interference pattern in the distribution of electrons is observed. In other words,

electrons behave as waves when not observed, but as particles in a definite location when observed!*17

All quantum entities—electrons, protons, photons, and so on—display this wave-particle duality,

behaving as wave or particle depending on whether they are directly observed.

A variation of this experiment by physicists Bruce Rosenblum and Fred Kuttner3 makes this bizarre

point even more clearly. If a wave corresponding to a single atom encounters a semitransparent

reflecting surface (such as a thin film), it can be split into two equal parts, much as a light wave both

going through and reflecting from a windowpane. The two parts of the wave can then be trapped in

two boxes, as shown in figure 4.1.

Figure 4.1. The wave function at three successive times: t1, t2, and t3.

Since the wave was split equally, if you repeated this process many times, then each time you

looked into the boxes you would find a whole atom in box A about half the time and in box B about

half the time. But according to quantum theory, before you looked the atom was not in any particular

box. The position of the atom is thus an observer-created reality. Its position will also be the same for

all subsequent observers, so it is a reality that depends on an initial observation only.

You may be tempted to think that the atom really was in one box or the other before you looked, but

it can be demonstrated that before observation the atom as a wave was in a “superposition state,” a

state in which it was simultaneously in both box A and box B. Take a pair of boxes that have not been

looked into and cut narrow slits at one end, allowing the waves to simultaneously leak out and

impinge on a photographic film. At points where wave crests from box A and box B arrive together,

they reinforce each other to give a maximum amplitude of the wave function at that point—a

maximum of “waviness.” At some points higher or lower, crests from box A arrive simultaneously

with troughs from box B. The two waves are of opposite signs at these positions and therefore cancel

to give zero amplitude for the wave function at these points.

Since the amplitude of an atom’s wave function at a particular place determines the probability for

the atom to be found there when observed, the atom emerging from the box-pair is more likely to

appear on the film at places where the amplitude of the wave function is large, but can never appear

where it is zero. If we repeat this process with a large number of box-pairs and the same film, many

atoms land to cause darkening of the film near positions of wave function amplitude maximums, but

none appear at wave function minimums. The distribution of darker and lighter areas on the film

forms the interference pattern.

Figure 4.2. The box-pair experiment: (a) waves emanating from slits in the two boxes travel distances da and db and impinge

on a film at F; (b) the resulting pattern formed on the film from many box pairs.

The distribution of electrons on the film will show the interference patterns typical of two waves,

which overlap to cancel each other at some places. To form the interference pattern, the wave function

of each atom had to leak out of both boxes since each and every atom avoids appearing in regions of

the film where the waves from the two boxes cancel. Each and every atom therefore had to obey a

geometrical rule that depends on the relative position of both boxes. So, the argument goes, the atom

had to equally be in both boxes, as an extended wave. If instead of doing this interference experiment

you looked into the pair of boxes, you would have found a whole atom in a particular box, as a

particle. Before you looked, it was in both boxes; after you looked, it was only in one.

Rosenblum and Kuttner sum up the puzzle:

Quantum mechanics is the most battle-tested theory in science. Not a single violation of its

predictions has ever been demonstrated, no matter how preposterous the predictions might seem.

However, anyone concerned with what the theory means faces a philosophical enigma: the socalled

measurement problem, or the problem of observation … before you look we could have

proven—with an interference experiment—that each atom was a wave equally in both boxes.

After you look it was in a single box. It was thus your observation that created the reality of each

atom’s existence in a particular box. Before your observation only probability existed. But it was

not the probability that an actual object existed in a particular place (as in the classical shell

game)—it was just the probability of a future observation of such an object, which does not

include the assumption that the object existed there prior to its observation. This hard-to-accept

observer-created reality is the measurement problem in quantum mechanics.4

Up until the moment of measurement, certain properties of quantum phenomena, such as location,

momentum, and direction of spin, simply exist as a collection of probabilities, known as the wave

function, or state vector. The wave function can be thought of as the probability distribution of all

possible states, such as, for instance, the probability distribution of all possible locations for an

electron.*18

But this is not the probability that the electron is actually at certain locations, instead, it is the

probability that the electron will be found at certain locations. The electron does not have a definite

location until it is observed. Upon measurement, this collection of all possible locations “collapses” to

a single value—the location of the particle that is actually observed.

Physicist Nick Herbert expresses it this way:

The quantum physicist treats the atom as a wave of oscillating possibilities as long as it is not

observed. But whenever it is looked at, the atom stops vibrating and objectifies one of its many

possibilities. Whenever someone chooses to look at it, the atom ceases its fuzzy dance and seems

to “freeze” into a tiny object with definite attributes, only to dissolve once more into a quivering

pool of possibilities as soon as the observer withdraws his attention from it. The apparent

observer-induced change in an atom’s mode of existence is called the collapse of the wave

function.5

Measurements thus play a more positive role in quantum mechanics than in classical physics,

because here they are not merely observations of something already present but actually help produce

it. According to one interpretation of quantum mechanics popular among many theorists, it is the

existence of consciousness that introduces intrinsic probability into the quantum world.

This interpretation owes its origin to mathematician John von Neumann, one of the most important

intellectual figures of the twentieth century. In addition to his contributions to pure mathematics, von

Neumann also invented game theory, which models economic and social behavior as rational games,

and made fundamental contributions to the development of the early computers. In the 1930s, von

Neumann turned his restless mind to the task of expressing the newly developed theories of quantum

mechanics in rigorous mathematical form, and the result was his classic book The Mathematical

Foundations of Quantum Mechanics. In it he tackled the measurement problem head on and rejected

the Copenhagen interpretation of quantum theory, which was becoming the orthodox position among

physicists. Although it is somewhat vague, the central tenets of the Copenhagen interpretation seem to

be (1) that all we have access to are the results of observations, and so it is simply pointless to ask

questions about the quantum reality behind those observations, and (2) that although observation is

necessary for establishing the reality of quantum phenomena, no form of consciousness, human or

otherwise, is necessary for making an observation. Rather, an observer is anything that makes a record

of an event, and so it is at the level of macroscopic measuring instruments (such as Geiger counters)

that the actual values of quantum phenomena are randomly set from a range of statistical possibilities.

Von Neumann objected to the Copenhagen interpretation practice of dividing the world in two

parts: indefinite quantum entities on the one side, and measuring instruments that obey the laws of

classical mechanics on the other. He considered a measuring apparatus, a Geiger counter for example,

in a room isolated from the rest of the world but in contact with a quantum system, such as an atom

simultaneously in two boxes.