Multiverse theory is hot, both within physics and astrophysics, and within popular culture. Halpern’s title, while no doubt a marketing choice, conveys a theoretical pull toward multiverse theory in physics and astrophysics.
Halpern rightly traces the history of the “multiverse” back prior to current physical theories. The term originated with the philosopher William James’ moral theory in the late nighteenth century, but the idea goes much farther back, even to classical thought. The term “universe” itself has a complicated history, much less the idea of a multiplicity of them. Until the early twentieth century, galaxies, once discovered to be external to our own Milky Way galaxy, were referred to as “island universes.”
There are different types of multiverses, even within physics. Hugh Everett’s interpretation of quantum theory, called the “Many Worlds Interpretation,” refers to a branching of realities as quantum probabilities are resolved, producing new “worlds” in which each possible value of a measurement is realized. The Many Worlds theory gives rise to popular depictions of “alternative universes,” like in the novel, The Man in the High Castle, or in the Marvel Comics stories.
Everettian theory is increasingly mainstream within physics, and Halpern spends some time on it, but it’s not really the focus of his discussion.
His discussion follows a reconstruction of the conceptual developments that lead in physics to solutions to problems in Big Bang cosmologies. Those theories, particularly addressing problems in accounting for the uniformity of our observed universe, its flatness, and its fine-tuning of seemingly variable factors, like the strength of gravity or our universe’s vacuum energy, toward the formation of galaxies, planets, life, and us.
One solution to that apparent coincidence of fine-tuning, proposed in the later parts of the twentieth century is the “anthropic principle.” In various forms, that principle provides a filtering mechanism to explain how it is that our universe is so well-suited to life (and us). We could not exist in any universe that did not permit the development of galaxies, stars, planets, and life, so we must exist in one that does. There may well be many universes “out there” much more poorly suited for those things, or entirely excluding them. We must be in this one, or at least in some small subset of universes that permit our existence.
In the late twentieth century, Alan Guth and collaborators developed what was thought to be a solution that preserved a single universe theory and did not necessitate anything so boggling as the anthropic principle. Inflation, as a speculative description of the very first moments (really nano-moments) of the universe’s existence, explained many of those coincidences.
But as inflation theory developed, it actually led to a multiverse theory itself, “eternal inflation.” If inflation happened, there is no reason to suppose it must have happened only once, or that it must have stopped anywhere and everywhere. Our universe could, and likely by theory, would be one of many “bubble universes,” some of which continue to inflate, some of which exclude the development of galaxies, stars, planets, life, and us, and some of which afford that development.
Thus the “allure” of the multiverse. Theory keeps leading us back to it, even when we think we’ve found a way to escape it.
The supposition that ours is not the only “universe” strains, for many, the limits of properly scientific theory. By their very nature, we will never experience those other universes, Everettian or inflationary. No one will ever get into some sort of universe-traversing ship, visit another universe, and come back. So their existence remains entirely hypothetical, unconfirmed, and beyond empirical verification by its very nature.
Setting aside for the moment suggestions for how evidence of the existence of bubble universes other than our own might be detected (“bruising” or, less dramatically, detection of intersections or collisions between expanding universes), it’s interesting to think a little bit about this debate over the scientific status of multiverse theory.
The debate rests centrally on this point of whether or not multiverse theory is subject to empirical test. Scientific theories, in one mainstream view of the job of science, must in principle be falsifiable, i.e., there must be crucial empirical tests whose outcome will show the theory to be validated or invalidated. Otherwise, the theory is not scientific and not worth entertaining.
Can we be sure that an adequate theory of the universe (of “reality” if you want to call it that to avoid limitation to one universe) is subject to empirical test?
The universe may well be beyond observation. In fact it is, regardless of whether our universe is part of a multiverse. The universe, owing to its expansion, is larger than the portion of it that we can observe. And that observable portion becomes smaller and smaller, in relation to the whole, all the time.
More importantly, once we draw back to consider the general relationship between observability and reality, to say that multiverse theories cannot be valid descriptions of reality because they would defy observational test is very odd. The universe is what it is, and what we can observe is, or may be, something else. Having observability rule reality has the two in the wrong order.
That doesn’t mean speculation should be licensed without bounds. Sometimes physicists deride “unscientific” theories as “philosophy.” Philosophers do not abide unbounded speculation. In fact, there are works by philosophers of science that specifically address the evaluation of scientific theories on non-empirical bases. See for example Richard Dawid’s String Theory and the Scientific Method (Dawid is a former physicist who found physics needing resolution to philosophical questions to proceed on its way).
A consequence of a recognized rupture between reality and observability would be that physical cosmology (or “scientific” cosmology) would need to give up any claim to describe all of reality, if scientific theory requires test by observation.
That would be interesting.