This is the third entry on Ladyman and Ross' Every Thing Must Go (2007), though the scope of the argument here will also extend to positions distinct from (though related to) the one advocated in the latter.
Once again, the project undertaken in the foregoing book is the rejection of metaphysics as it is generally practiced in contemporary philosophy, which aims to give us an account of the most general features of reality more or less independently of the sciences, in favor of a much more modest kind of metaphysics that does nothing beyond showing how different scientific subfields, like physics and biology, connect to each other.
One thing we have not yet seen is that at the heart of the book's proposal lies a "verificationism", which (partly) consists in the claim that
"no hypothesis that the approximately consensual current scientific picture declares to be beyond our capacity to investigate should be taken seriously" (p.29)
The authors make it clear that "investigate" here refers specifically to scientific investigation. Indeed, if it follows from our scientific theories that a certain question cannot be answered by the methods of science, we should refrain from "[looking] to an institution other than science to answer such questions" and instead "forget about the questions" (p.30). The hypothesis that God caused the Big Bang, because it cannot even in principle be confirmed scientifically, should therefore not be affirmed or denied, but simply ignored. Thus, the "verificationism" of Everything Must Go entails that claims outside the reach of science are not worth our time.
The term verificationism is an explicit tribute to logical positivism, on which the authors openly model their approach. Of course, the verificationism of the logical positivists was significantly different, though similar in spirit, to the one advanced here. The members of the Vienna Circle famously (or infamously) argued that scientifically unverifiable statements (unless true by definition) are meaningless. Ladyman and Ross make the comparably weaker claim that scientifically unverifiable statements should be ignored.
As is well known, the logical positivists faced the damning objection that their own principle undermined itself: there is no way to scientifically verify the claim that scientifically unverifiable statements are meaningless (unless true by definition), meaning that their verificationism was meaningless by its own lights. It is less well known that Rudolf Carnap, a leading figure of logical positivism, produced a sophisticated answer to this problem in his The Logical Syntax of Language (1934). The verificationist principle, he argued, only applies to propositions, claims capable of being true or false. But the principle itself is not a proposition, but a kind of prescription, something like 'regard as meaningless any proposition that is neither scientifically verifiable or true by definition'. Such a sentence couldn't possibly be true or false, and thus is not excluded by the verificationist principle (i.e. by itself). Carnap argued that whether or not we choose to follow the verificationist prescription is a pragmatic matter. Of course, he contended that there were many pragmatic benefits to be gained from following the principle, not least that it avoids us getting bogged down by trivial metaphysical debates, allowing to focus on the far more fruitful endeavors of scientific investigation.
Interestingly, the verificationism of Everything Must Go also seems to take the form of a prescription (notice the "should" in the above quote), in which case it too evades putative accusations of incoherence and self-undermining. Moreover, it is also partly (though not exclusively) justified on pragmatic grounds, as the authors make clear that "identification of the nature of justifiable metaphysics, is pragmatic in character" (p.28) where "justifiable metaphysics" is the kind of modest metaphysics that merely draws out the connections between the different sciences, as opposed to the more traditional 'beefy' metaphysics that seeks to plug the holes that science cannot fill. The modest metaphysics is pragmatically beneficial, because it gives us (or aims to give us) a more unified picture of the world, in which the various claims of different sciences mutually support one another, rather than "stranded" hypotheses which, unless connected to a nework of hypotheses, remains "a mystery", and we sure don't like mysteries. It is more pragmatic for metaphysicians to engage in this task, because scientists need to be highly specialized in order to be efficient.
But is it really 'pragmatic' to accept one version or another of verificationism? I take it that it is 'pragmatic' to follow a prescription if it is somehow in our interests to do so, e.g. if it secures some good that we desire. And here it seems that verificationism, both in its positivist and revised forms, is decidedly unpragmatic. Human beings are interested in all sorts of subject matters that are beyond the reach of natural science. They care about questions of right and wrong, of what constitutes a good or well-lived life, of whether there is any sense in which they have free will, whether there is a God, and so forth. And yet the prescriptions of verificationism tells them to dismiss such questions, either because they are meaningless, or because they are just not worth our time. Defenders of either variety of verificationism could respond that, even if it is true that human beings in interested in matters beyond the reach of natural science, it is nonetheless more pragmatic to ignore the questions pertaining to these matters, since (1) they simply cannot be answered with any significant degree of certainty, or (2) because we can never agree on them. The claim that one cannot answer an unscientific question without any significant degree of certainty would have to rely on the authors' arguments for the unreliability of nonscientific (in this case, metaphysical) methods, some of which have been addressed in the previous two entries. But even if we accept (1), it still doesn't follow that it is in our best interests to dismiss scientifically intractable questions. Lacking in certainty as to the answer to an important question may be psychologically unsatisfying, but perhaps still less unsatisfying than the outright dismissal of the question. Likewise for the unsatisfyingness of not being able to agree on important nonscientific questions ( (2) ).
Let me emphasize that pragmatism is not the only consideration that Ladyman and Ross raise in support of their brand of verificationism, though it seems to do most of the work in Carnap's earlier brand. Nevertheless, it seems to me that the moment one raises pragmatic benefits as a reason to support a theory, one must be prepared to answer for all the ways in which said theory is unpragmatic. I take this to be the most serious weakness of positivism and its descendants, far more serious than the self-undermining issue: positivist theories pride themselves in the pragmatist spirit by which they reject 'fruitless' metaphysics, but, because they ultimately disregard human interests beyond scientific explanation and prediction, themselves turn out to be woefully unpragmatic, if not positively dehumanizing.
Carnap, R. (1934). Logical Syntax of Language.
Ladyman, J. and Ross, D. (2007). Every Thing must go. Oxford: Oxford Univ. Press.