Review of S. Weinberg, Third Thoughts, 2018 and related comments by N.D. Mermin, 2019


The book presents Weinberg’s viewpoints on a number of issues, from the history science including the history of Astronomy, major issues in science and cosmology including the trouble with quantum mechanics as well as other issues such as scientific funding, espousing political views, and being against manned space flight. He also presents additional personal views such as being wrong on the lack of symmetry in neutrino emission.

We will focus our review on Weinberg’s viewpoint in Chapter 14, The Trouble with Quantum Mechanics (an essay first published in The New York Review of Books, January 19, 2017), where he discusses the quantum measurement problem. Weinberg asks how do probabilities get into quantum mechanics. He discusses two classes of approaches, “instrumentalist” and “realist” approaches for which he states he is not satisfied with either. We will refer to Class C as those approaches which are neither instrumentalist nor realist. He later discusses GRW type of Lindblad approaches, which he seems to be more satisfied with, which is in Class C.

The instrumentalist approach is essentially that the wavefunction is not real but rather an instrument to be used to compute the probability of measurement once a measurement occurs.  In Weinberg’a description of the instrumentalist approach, the divide between measurement and unitary evolution occurs in the consciousness of the human observer.  He rejects this approach at this time, preferring not to assume that measurement occurs only in the observer’s mind, but rather to ascertain what is happening by fundamental laws. He states, “We may in the end have to give up this goal, but I think not yet.”

The realist approach is described in which the wavefunction is always a legitimate description of reality, and that the wave function evolves deterministically.  However, he rejects the realist approach for which non-local evolution and entangled states are predicted under Schrödinger evolution. That is he states, “how can something so non-local represent reality?”  Moreover, Weinberg examines Everett’s multiple world interpretation which can be derived as a function of the Schrödinger wave function evolution, but concludes that the Born rule has not been derived successfully using Everett’s model.

Weinberg concludes that the “present form of quantum mechanics may be warning us that the theory needs modification.” Moreover, “a new theory might be designed so that the superpositions of states of large things like physicists and their apparatus even in isolation suffer an actual rapid spontaneous collapse, in which probabilities evolve to give the results expected in quantum mechanics.”


Weinberg describes in Chapter 14 the basics of why there is a measurement problem. The description is put in terms of a particular outcome occurring probabilistically in measurement versus deterministic superpositions occurring via Schrödinger’s equation.  His description of the measurement problem indicates he has a good understanding of the problem. We agree that quantum mechanics in its present form will most likely require modification to account for measurement.


Weinberg believes that the instrumentalist approach is, “a surrender of a particularly unfortunate kind.” That is, Weinberg believes, “the instrumentalist approach turns its back on a vision that became possible after Darwin, of a world governed by impersonal physical laws that control human behavior along with everything else.” Weinberg quotes Wigner, “it was not possible to formulate the laws of quantum mechanics in a fully consistent way without reference to the consciousness.”  However, in our own research, we have not come across any statement from Wigner in which he states that “human” consciousness is necessary for measurement. That is, we expect that Wigner would have admitted that a dog or cat is conscious and makes measurements (i.e. through its sensory organs such as through vision, taste, hearing, touch, etc.). Also, later in life Wigner changed his mind and believed that consciousness is not a necessary condition for measurement. This is essentially Bohr’s position, in that the tinge of conscious experience is a sufficient condition for measurement, but other processes such as macroscopic amplification are also sufficient conditions.  Hence, we believe that reference to “human” consciousness as opposed to a dog’s consciousness is irrelevant at this stage of investigation. That is, any differences that occur from whether a “human” experiences a sense from a quantum phenomenon (such as the tinge that results from smelling a few molecules) or the tinge that a “dog” experiences from smelling the same few molecules, is likely a minor effect that is best left to investigation after the basics of measurement are understood. Admittedly if only human consciousness were to measure and not a dog’s consciousness, then the problem is much more difficult.

With that said, there is no surrender when considering the possibility that consciousness is involved in measurement.  For example, suppose that measurement occurs at the level of a bacteria or within a cell, and that there is a primordial tinge that results depending on certain cell functions that incorporate quantum measurement.  One might disagree that viruses or even individual cells exhibit tinge. But then there are multiple cell organisms that could be investigated, etc. a long way before the need to investigate “human” consciousness. Such theories for which consciousness is a sufficient condition for measurement and occurs at levels below human consciousness, somewhere between a virus and let us say for argument sakes a Condylonucula maya which is the smallest bi-valve of only 500 um in diameter and has functions for which it open and closes and for which measurement would seem to be a useful tool for the bi-valve to survive. The investigation of such theories falls into Class C, as they are neither instrumentalist nor realist.

On, “Can the scientist play a role in the laws of physics” by N.D. Mermin, Physics Today, January 2019, Pp. 53-54.

N.D. Mermin, in his review of Weinberg’s book Third Thoughts states:

There is an implicit assumption, shared by almost all physicists, that the scientist must be separated from the science. The usual appeals to measurement with classical outcomes, it seems to me, are unsuccessful attempts to objectify and impersonalize processes in which an individual scientist acts on and is reacted upon by the world. The collapse of the wavefunction after measurement represents nothing more than the updating of that scientist’s expectations, based on his or her experience of the world’s response to the measurement. Weinberg hopes to keep the scientist out of the laws of nature, but our chronic failure to agree on the meaning of quantum mechanics demonstrates the futility of his hope.

Now, while it is true that the scientist should be separated from the science, it does not mean that a scientist cannot perform experiments on other systems that might be conscious and also are measuring devices.  For example, a scientist might perform an experiment on a bi-valve or other sensory system of an organism that would be expected to create a tinge within the system in response to an external stimulus. That only a “human” is a bonafide measuring device is, in our view, utterly wrong.

The act of an external quantum phenomenon producing a tinge within an observing system is a sufficient physical condition for quantum measurement. We do not know whether or not consciousness is a necessary condition for measurement. Consciousness and its relationship to measurement can be investigated scientifically, and there would be no capitulation regarding the measurement problem.

A scientific investigation could center around what is necessary to physically constitute a system that exhibits tinge in response to an external phenomenon, which is characteristic of conscious systems.  We emphatically doubt that human consciousness as opposed to an earlier evolutionary form, is necessary in order to separate the scientist from the science. That is, the science that is being investigated by the scientist could be a primitive and less intelligent form of consciousness as compared to a human’s consciousness, and could still be a measurement device. Again, we do not know whether or not consciousness is a necessary condition for measurement. In fact, it may be that under certain non-conscious conditions, such measurements can occur.  However, we do not know the non-conscious conditions for measurement and we do know that conscious tinge is sufficient for measurement. It makes sense therefore to explore the potentially conscious architectures in Class C which have been already invented through the evolutionary process and have been in use for over a billion years. One does not need to reinvent the wheel; the solution to the measurement problem we expect has been already solved by natural selection because of the utility of measurement structures, a long long time ago. Furthermore, it is not only the utility of measurement structures that would have been beneficial in the evolutionary process, it is the potential necessity of measurement structures for the creation of consciousness and life as we know it.

Additionally, the non-deterministic and non-causal properties of measurement bear a striking resemblance to what one considers to be the animated and non-predictable properties of a living conscious system versus an inanimate system that appears to do little but evolve predictably and deterministically. And there have been other fundamentally non-deterministic theories proposed such as GRW theory. The idea that one should look to causal deterministic physics or laws in order to generate measurement is certainly plausible at this time, but then one has to ask why shouldn’t Schrödinger’s equation generate the quantum state evolution when the underlying physics is determinate?  That is, when the underlying physics is deterministic, Schrödinger’s equation is a perfectly satisfactory equation to be utilizing, and we see little reason to deviate from Schrödinger’s equation in such cases.

Hence it is most logical at this stage to primarily investigate phenomena that appear to be fundamentally non-deterministic within Class C, and secondarily investigate phenomena that are governed by deterministic equations but that evolve differently than Schrödinger’s equation. Whether the fundamental non-determinism comes from an external random process such as in GRW theory or whether it is intimately related to tinge and consciousness is unknown at this time, but consciousness as an explanation cannot be taken off the table in any logical sense at this time.

Leave a Reply

Your email address will not be published. Required fields are marked *