To illustrate the difficulties facing the proposal of collapse by conscious perception, we will look at a Stern-Gerlach experiment on a spin silver atom. (See also Ref. 2, Ref. 6.) There is a detector on the trajectory of the – branch but none on the trajectory of the + branch, with the detector reading “yes” for detection and “no” for no detection. Suppose the observer is initially gazing off into space and then, a couple of seconds after the silver atom passes the detector, she focuses her attention on the detector.

Presumably, for an instant, there are two versions of the observer, perceiving both the yes and the no branches. In this instant, there must be something—we will call it a “Mind”—outside quantum mechanics, which realizes ‘the observer’ is perceiving a multi-version reality. That is, the “Mind” must somehow become aware that there are two separate versions of the observer’s brain wave function. And then, in the next instant, the “Mind” collapses the wave function to just one option.

There are three problems with this approach. They do not rule it out, but they still need to be addressed. The first is that conscious perception—focusing the eyes on the dial in this case—is a complicated neural process. So it is not clear at what point in that process the “Mind,” perceiving the state of the brain, knows there are two versions.

For the second problem, suppose the observer focuses on the dial the whole time, from before the atom is shot through the apparatus to after it passes the detector. And suppose the final result is that the detector is perceived as continuing to read ‘no,’ so there is, in the conventional sense, no change in the dial reading from beginning to end. In that case, the observer has not consciously (in the conventional sense) perceived anything that could have cued the collapse. This implies the definition of “conscious perception” must be amended from its conventional meaning to accommodate collapse by “conscious perception.” It also implies, as indicated above, that the “Mind,” associated with the individual brain, is perceiving more than one version.

To illustrate the third problem, we again use a Stern-Gelach experiment, with detectors on both branches. The state is
Now each version of the observer’s brain is independent of the coefficients; for example, the neural firing pattern for each version does not depend on a(+) or a(-). However, the “Mind” must collapse the wave function in accord with the |a(i)|² probability law, so the “Mind” must have access to the values of those coefficients. But it is not possible to get those values strictly from the wave function of the observer’s brain. So some process must be proposed for how the “Mind” ‘senses’ the value of the coefficients. This observation virtually guarantees that the collapsing “Mind” must perceive more than just the wave function of the individual brain. To obtain information on the coefficients, it must apparently perceive the whole wave function, and then, by some unknown process, sort out the values of the coefficients. And it then rolls the dice.

In summary, proposals for observer-induced collapse need to be made more explicit.