“We demarcate good science-around lines which are inevitably fuzzy near the boundary-by reference to institutional factors, not directly epistemological ones.” (Every Thing Must Go p. 33)

‘Every Thing Must Go’ relies heavily on the notion of scientific institutions. Their justification for this reliance is pragmatic; i.e. these institutions have been highly successful in the past so we should trust them. They have a pragmatic view of science and argue that attempts of philosophers to explicate necessary and sufficient conditions of something being science will inevitably be inferior as the practices that have been honed through practicing scientists working in the field will yield more reliable results. 

“To reiterate; we assume that the institutions of modern science are more reliable epistemic filters than are any criteria that could be identified by philosophical analysis and written down.” (ibid p. 37)

Ladyman et al’s primary target is analytic metaphysics which they argue constitutes a crude caricature of scientific modelling. They applaud the clear and rigorous style of argumentation used by analytic metaphysicians, however they argue that despite the argumentative style such metaphysicians are engaging in crude pseudo-scientific reasoning.The majority of analytic metaphysicians would describe themselves as naturalists and would consider themselves to be theorising about the world in a way consistent with contemporary science; though using different techniques. 

Despite describing themselves as naturalists such metaphysicians rely on toy models of reality that are wildly at odds with contemporary discoveries in physics. Using examples from contemporary metaphysical papers, Ladyman et al demonstrate that analytic metaphysicians regularly rely on assumptions about the physical world that are not only pre-Quantum Physics, or pre-Relativity Theory but are pre-Newtonian. Self described physcialists gleefully speak of reality of at base being composed of atoms banging against each other. The ontology they presuppose is closer to Cartesian Physics than it is to anything discovered by contemporary physics. Furthermore aside from the crude ontology that is uncritically assumed, a key methodological assumption of analytic metaphysics is that a plus point for a theory is if it chimes with our intuitions. Ladyman et al, note that this methodological assumption is wildly at odds with the assumptions in our best science which assumes that our brains which were designed by evolution to mediate our interaction with middle sized objects are terrible at intuiting the nature of reality. 

Ladyman et al aren’t arguing a la the logical positivists that metaphysics per se is a meaningless discipline. Rather they are working within the tradition of people like Chomsky and Quine who argue that metaphysics is a perfectly sensible discipline; but that our best way to draw a metaphysical picture of reality is through understanding the world as our best science describes it. They offer two heuristic constraints on metaphysicians (1) The Principle of Naturalistic Closure: A new metaphysical claim is justified if and only if it involves the conjunction of two or more sciences (one of which is fundamental physics) jointly explain more than was explained by the two hypotheses taken separately (2) The Primacy of Physics Constraint: any theory we can construct must be consistent with our best contemporary physics. 

Ultimately the metaphysical picture they offer in place of analytic metaphysics is a form of structural realism. They argue for this picture based on two famous philosophy of science arguments. (1) The Pessimistic Meta Induction: Scientific Theories in the past with great predictive and explanatory success referred to entities we know today do not exist; e.g. Caloric and Phlogiston.  Therefore it is likely that our current scientific theories refer to entities that do not exist. Hence we shouldn’t be realists about the entities presupposed in scientific theories. (2) The No Miracles Argument: Given the pragmatic and predictive success of our best scientific theories; if they don’t pick out real things in the world their success is a miracle. If germ theory isn’t picking out real things e.g. germs then its success is inexplicable. The no miracles argument and the pessimistic meta induction are in tension with each other. Ladyman et al try to resolve this tension by arguing that we can have our cake and eat it. We can be realists about our scientific theories and account for theory change by arguing that what is preserved across scientific theories is not things but structure. 

The metaphysical theory they argue for is compelling and is a parsimonious way of thinking about the history of science. Furthermore their criticisms of analytic metaphysics for its reliance on intuitions and pre Newtonian models of physics are to the point. While I share their respect for science and their naturalism I am unsure of what to make of their claims about the institutions of science. Science is clearly our most successful way of interpreting reality and it gives us a level of prediction and control that is far superior to any other academic discipline. Nonetheless scientific institutions aren’t always the best metric as to what counts as a reasonable question to ask. 

Ladyman et al suggest that asking what hypothesis’ would be accepted for a grant proposal is a reliable determiner of the validity of the hypothesis:

“Here, then, is such a norm. Almost all successful participants in ‘bona fide institutional science’- on which we will say more below-learn in graduate school,  or soon after, which sorts of hypotheses one cannot propose as the targets of investigation in a grant proposal to a ‘serious’ foundation or funding agency with non-zero prospects of success.”  (ibid p. 33)

They even jokingly ask us to imagine the prospects for an analytic metaphysician seeking funding in a particle physics department. The prospect of a metaphysician appealing for funding in a physics department is hilarious; one could imagine them being laughed out of the room. But one wonders whether this would be a reliable metric?

In the Harvard psychology department  in the late forties if you were to seek funding for studying the neural correlates of consciousness you would have been laughed out of the room. With B.F. Skinner’s brand of behaviourism being the dominant strand of psychology in America of the time, studies of subjects such as consciousness were deemed to be beyond the pale in terms of science. Pre Skinner behaviourists would have rejected studying consciousness on methodological grounds. They would have argued consciousness existed but couldn’t be studied scientifically because of its subjective nature. So early behaviourists would leave it to philosophers to speculate over the nature of consciousness. Skinner’s radical behaviourism held the position that consciousness was real but was best understood interms of behaviour. He would have argued that to study consciousness the best approach would be to study how the contingencies of reinforcement were arranged to help us develop tools to speak about our private experiences. The key point is that during the heyday of behaviourism certain questions about the nature of consciousness would have been deemed silly and would have been laughed out of court by the dominant scientific institutes of the time. 

Even after the rise of cognitive science in the 50s most psychologists or neuroscientists would have been afraid to seek funding for a study on consciousness. It was only in 1990 when people world famous scientists such as Francis Crick started studying consciousness that the taboo was broken. When Ladyman et al make the following claim:

“We seek a principle, referenced to the institutional factors that make science  epistemically superior, for distinguishing well-motivated from ill-motivated metaphysical proposals; we do not seek a principle for distinguishing sense from nonsense. (ibid p. 34)

One wonders whether they would argue that the institutional factors that kept consciousness a taboo subject matter in science for over fifty years were justified?

And it’s not just in psychology that institutionally sanctioned taboos are enforced. In his book ‘What is Real? Adam Becker tells the story of how physicists who wanted to interpret Quantum Mechanics were actively ran out of the field. The dominant mantra in Physics was ‘shut up and calculate’, physicists who questioned this approach and offered an interpretation to explain the experimental results in Quantum Mechanics risked serious harm to their career prospects. So here again one wonders whether Ladyman et al think that the institutional factors in physics that were used to weed thinkers like Tim Maudlin, and David Albert out of the physics department and into the philosophy department were justified?

While I admire Ladyman et al focus on scientific practice and criticisms of philosophy that is wildly at odds with what we know about the world through experimentation. I think their views of institutional science are too rationalistic. Thomas Kuhn’s historical approach to understanding scientific revolutions reveals scientific institutes as a much more human and fallible and tribal scene than the objective picture presented by Ladyman et al. And as we have seen with our brief discussion of behaviourism and quantum mechanics sometimes dominant scientific institutes can stand in the way of understanding not enhance it.