OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

1
Introduction
Steven French and Juha Saatsi

Quantum physics represents our best and most fundamental physical science. Our
understanding of numerous physical phenomena and our knowledge of the nature
of light, electricity, solid matter, elementary particles, and even parts of chemistry, is
rooted in quantum physics. But exactly what kind of knowledge does it provide us?
This question gains significance from weighty epistemological issues that forcefully
arise in this context—issues that are also at the heart of a more general debate on
‘scientific realism’ in the philosophy of science. This volume aims to advance both
the realism debate as well as our understanding of the nature of quantum physics by
bringing the two together in a productive dialogue.
Scientific realism was famously announced dead already back in 1984 by Arthur
Fine, an American philosopher of science. He explained that its demise “was hastened
by the debates over the interpretation of quantum theory, where Bohr’s nonrealist
philosophy was seen to win out over Einstein’s passionate realism,” and that its death
“was certified, finally, as the last two generations of physical scientists turned their
backs on realism and have managed, nevertheless, to do science successfully without
it” (1984, p. 261). Fine’s diagnosis appears flawed, however, as more than thirty-five
years later realism doesn’t just linger on, but thrives in discussions about quantum
physics! But debates over the interpretation of quantum theory have not become
any calmer in the hundred or so years after its inception, even though Bohr’s ideas
have been debunked many times over (Becker 2018). If anything, it is currently even
harder to find a consensus about critical interpretive issues, as the range of seriously
considered alternatives has steadily increased amongst broadly realist approaches
to quantum physics. In this state of affairs there is considerable pressure to better
articulate not just what “realism about the quantum” amounts to, but also what justifies
a realist perspective over alternatives that rescind one or another realist theses. This
volume collects together new work from the cutting edge of this active area of current
research.
As indicated, the core realist intuitions about the physical sciences are resilient and
hard to deny, but what exactly does “realism” stand for? As Richard Healey explains in
his contribution to this volume, this term of philosophical trade has many meanings.
In the extensive philosophical and foundational literature on quantum theory, realism
has most typically signified the notion that we can specify what the mind-independent
physical world could be like so as to render quantum theory (approximately) true.
Painting a picture of reality compatible with the truth of quantum theory is a business
Steven French and Juha Saatsi, Introduction In: Scientific Realism and the Quantum. Edited by:
Steven French and Juha Saatsi, Oxford University Press (2020). © Steven French and Juha Saatsi.
DOI: 10.1093/oso/9780198814979.003.0001
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

2 steven french and juha saatsi

of interpreting the theory, its mathematical formalism and models. Realists like
Einstein traditionally placed intuitive constraints on plausible pictures of reality com-
patible with quantum theory—in particular, that they should conform to a principle
of locality according to which physical systems have determinate local properties not
influenced by action at a distance.
Today’s realist interpretations are a much more motley bunch, filling more of
the space of logically coherent possible ways the world could be to make true one
or another formal variant of quantum theory. These interpretations have emerged
over decades of work by both philosophers and physicists engaged in foundational
research. This work has largely aimed to tease out quantum theory’s metaphysical
and ontological implications, but hitherto much less attention has been paid to the
concomitant epistemological issues. It is these latter issues that form the primary focal
point of the present volume, which aims to engage more directly with the relevant
epistemological questions that are also debated within general philosophy of science,
concerning the status of our best scientific theories as a source of knowledge about
unobservable reality or as furnishing representations of it.
The realist attitude towards well-established scientific theories is widely shared,
seemingly common-sensical, and presupposed by the broadly accepted idea that such
theories indeed do provide us genuine scientific understanding of natural phenomena
through explanations in terms of how the unobservable world is structured and how
it “works”. Considerations in favour of realism tend to capitalize on the empirical
success of science, variously manifested in triumphant theoretical predictions and
the way science ever-increasingly supports our ability to manipulate the world to
our liking through powerful interventions and applications that put to concrete use
quantum theoretical notions such as ‘spin’.
But while realists proclaim optimism about science’s ability to tell us how things
stand behind the veil of observable appearances, a very long tradition steadily opposes
any such optimism on the basis of varied considerations regarding science at large.
Two sources of scepticism have been particularly pervasive. First, there are historically
driven concerns about the status of our current scientific theories’ as ‘approximately
true’, based on the historical track-record of radical and unexpected (r)evolutions in
foundational scientific theorizing. Secondly, there are general “underdetermination”
worries about the possibility of there being empirically indistinguishable—either in
principle or for all practical purposes—theories that represent the world in radically
different ways. These two broad sets of concerns have been raised time and again
against scientific realism in various specific ways, sometimes individually, sometimes
in unison.
Like much of general philosophy of science, the debates surrounding such concerns
have been traditionally largely conducted in rather broad and abstract terms, quite
independently of specific scientific detail. In a significant recent trend philosophers
have become increasingly troubled about the potential limitations of sweeping,
general arguments for or against realism, due to the variability of evidential, method-
ological, and explanatory contexts and practices that seem relevant for the assessment
and outcomes of these arguments. As a result, there has been increasing emphasis
within the realism debate of the importance of discipline or domain specific scientific
details. In this spirit more ‘local’ analyses of the key issues animating this debate
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 3

have been undertaken in relation to disciplines such as, e.g. cosmology, economics,
geology, molecular biology, paleontology (see, e.g. contributions to Saatsi 2018,
Part IV).
Quantum physics of course also provides a very natural locus for such a ‘local’
analysis, as the contributions in this volume nicely demonstrate. On the one hand,
realism towards quantum physics is very easy to motivate in the light of its truly
outstanding empirical successes. On the other hand, the theory is well known for its
exceptional interpretational challenges and the resulting bifurcation regarding what
it is taken to tell us about reality. This bifurcation powerfully brings to life the kind
of underdetermination that many anti-realists have tended to worry about in the
abstract. Relatedly, many classic philosophical questions concerning the relationship
between science and metaphysics—the latter being deeply problematic according
to some prominent anti-realist philosophers—are also nicely brought into focus in
this context. In addition to throwing new light on such well-known issues, there
are also entirely new ideas to be considered that have recently emerged specifically
in the context of the philosophy of quantum physics, such as quantum pragmatism
(advocated by, e.g. Richard Healey), and quantum Bayesianism (advocated by, e.g.
Christopher Fuchs). The fourteen chapters that follow engage with all these issues
and many, many more.
* * *
Let’s now turn to the contributions to this volume. A theme running through many of
them is to respond to the above problems by changing the terms in which realism
is articulated. PART I presents two proposals for accomplishing this by explicitly
rethinking what scientific realism amounts to.
Carl Hoefer articulates and defends ‘Tautological Scientific Realism’ (TSR). It
eschews standard ways of defending and delineating realist commitments with regard
to Inference to the Best Explanation and considerations of what might be preserved
across theory change. Nevertheless, TSR, much like standard realism, maintains that
our current scientific picture of the world is to a significant extent correct and will
be retained through future changes at the theoretical level. But such a realist stance
is only appropriate, Hoefer argues, with respect to those areas of current science for
which we simply cannot seriously imagine future developments that would show that
we are seriously mistaken in our current ontological commitments—in the way that
we were with regard to phlogiston, for example. These ‘safe’ areas of science embrace
the core properties of atoms and the way they combine, as well as our knowledge of
electronics, for example, but not, crucially, quantum physics. Hence, Hoefer argues—
more contentiously—for a new way of delineating realist commitments, according to
which our current ‘fundamental’ theories, such as quantum mechanics and quantum
field theory, are specifically excluded from the scope of TSR. The grounds for this
are two-fold: first, quantum physics is subject to the kind of underdetermination
indicated above (and as discussed in one way or another by most of the papers
in this collection); and secondly, it is expected to be replaced by a theory capable
of unifying quantum theory and general relativity. Thus, Hoefer argues that the
appropriate attitude towards quantum physics is one of anti-realism: agnosticism
about its ontology, coupled with instrumentalism about its theories.
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

4 steven french and juha saatsi

Juha Saatsi also proposes an alternative articulation of realism, focusing his

discussion on the exemplary quantum property of spin. As is well-known, spin has no
classical analogue and, as Saatsi notes, it not only lies at the heart of many quantum
theoretic explanations, but has come to be understood as increasingly manipulable in
a way that allows it to feature in a number of exciting new technological developments
(e.g. ‘spintronics’). These features strongly motivate a realist stance towards spin in a
way that is, Saatsi argues, analogous to the motivations behind Hoefer’s TSR, thereby
questioning the latter’s exclusion of quantum physics from its scope. Yet spin also
lies at the heart of the ‘interpretational conundrums’ of quantum theory and spelling
out what spin is involves ‘deep’ metaphysical commitments that go beyond what
is necessary to account for any theory’s explanatory and predictive success. Here
the underdetermination that bedevils realism raises its ugly head again: even the
comparatively straightforward setup of silver atoms passing through a Stern–Gerlach
apparatus arguably comes to be characterized in radically different ways depending
on one’s theoretical approach. Yet, these different ways seem to add nothing to the
epistemic virtues of the theory—hence, Saatsi argues, they involve ‘deep’ metaphysics
that remains unjustifiable by the realist’s lights.
However, rather than abandoning a realist stance towards spin altogether, Saatsi
argues that we should step away from such deep metaphysics and modify our realism
accordingly. Thus he suggests we should give up the epistemic ambitions of what he
calls ‘truth-content’ realism, grounded as it is in notions of reference and approximate
truth. Instead we should accept and articulate a form of ‘progress’ realism which, in
the case of spin, does not commit to claims about what spin is like, but nevertheless
acknowledges that the relevant models function as representations of reality and
to that extent can be considered to ‘latch onto’ the world in ways that ground the
empirical success of the theory. This maintains a representational role for these
models and, in naturalistic terms, allows for physics theorizing itself to explain the
success of spintronics, for example. It also constrains future theorizing by pointing to
those well-known exemplars of inter-theoretic relationships that motivate claims of
theoretical progress and emphasizing that this is how physics can make sense of its
own empirical success.
PART II contains three chapters that further explore the challenges faced by realism
in the quantum context, focusing on the interconnected issues of underdetermina-
tion and interpretation. As already noted, there are well-known alternative realist
approaches to quantum theory, such as Bohmian (‘hidden variables’), Everettian
(‘many worlds’), and dynamical collapse formulations, as well as specific interpreta-
tions, such as Quantum Bayesianism, the ‘transactional’ approach, and myriad others.
Which, if any, should a realist embrace, and on what grounds? Or does the problem of
underdetermination completely undermine scientific realism in relation to quantum
theory?
Craig Callender voices a degree of pessimism about realists’ prospects for quaran-
tining the blight of underdetermination. He helpfully characterizes the different foun-
dational/interpretational approaches in terms of Lakatosian ‘research programmes’,
delineated by their hard cores and negative and positive heuristics. As he also points
out, given the flexibility inherent in each programme, no crucial experimental test
between them is likely in the near future. Furthermore, as Callender goes on to
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 5

argue, the underdetermination here cannot be dismissed as artificial (with the various
interpretations construed as philosophers’ toys or mere notational variants); hence,
the realist faces a genuine problem.
A natural move for the realist is to try to find some common ground between these
different programmes in which she can plant her flag. Unfortunately, as Callender
spells out, all we seem able to find are small disconnected ‘islands of reprieve’.
Following a suggestion by Alberto Cordero (2001), Callender looks at basic ‘textbook
level’ cases of quantum models, to see what common ground can be found. He
concludes that insofar as quantum models of, e.g. the water molecule underwrite
incontestable claims, these claims are not distinctly quantum in nature. And much of
the distinctly quantum theoretic content of models of, e.g. quantum tunnelling or the
hydrogen atom, turns out to differ between different variants of quantum mechanics
(e.g. the Bohmian theory vs. standard textbook presentations)—a point that chimes
with Saatsi’s claim about different accounts of the Stern–Gerlach apparatus. Even
relaxing what one means by ‘quantum’ and shifting to the semi-classical level offers
little hope for the realist, as Callender shows that most of what we say about the
quantum realm is dependent on the chosen foundational perspective. And unless
the realist substantially reins in her ambitions, along the lines suggested by Hoefer
and Saatsi, for example, it seems she must make such a choice. But which? As
Callender notes, it’s not just a matter of balancing competing theoretical virtues, but of
reconciling different attitudes towards such virtues and their relationship to empirical
confirmation. In a sense, he concludes, we have a kind of philosophical gridlock.
David Wallace locates the disagreement between the different realist approaches to
quantum theory at a more fundamental level: what is the theory that one is trying to
interpret in the first place? In other words, when it comes to the issue of identifying the
‘best interpretation of quantum theory’, it is not just a matter of debating theoretical
virtues, or what we mean by an ‘interpretation’, but how we identify quantum theory
itself. Wallace argues that ‘abstract’ quantum mechanics, with its formalism of Hilbert
space and self-adjoint operators representing observables and so on, should not be
conceived of as a scientific theory at all, but as a theoretical ‘framework’ within
which concrete quantum theories—e.g. quantum particle mechanics, quantum field
theory, or quantum cosmology—can be expressed. The latter theories have limited
applicability, depending on the type of system, the energy level involved, and so on.
(This is analogous to classical mechanics, according to Wallace, where an overarching
framework, provided by Hamiltonian or Lagrangian mechanics, encompasses a wide
range of concrete theories.) These concrete theories are inter-related in various ways,
and Wallace argues that in the quantum case these inter-relationships should be seen
not as establishing a hierarchy, but something more akin to a patchwork (although not
necessarily a disunified one). Given this understanding of quantum theory, he asks:
what should we expect from an interpretation thereof?
The answer, Wallace argues, is an interpretive recipe that tells us how to understand
any specific quantum theory, in a manner that is compatible with the relevant inter-
theory relations. (Again, this is analogous to classical mechanics.) Such an interpretive
recipe is arguably provided by the Everett interpretation. Other interpretational
approaches, he claims, fail to similarly make sense of the relationships between spe-
cific, concrete quantum theories. Dynamical collapse and hidden variable approaches,
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

6 steven french and juha saatsi

for example, typically focus only on non-relativistic particle mechanics and, further,
under the fiction that it is a fundamental and universal theory. The theoretical com-
mitments of such approaches that in one way or another modify ‘standard’ quantum
mechanics are disconnected from the actual practice of physics, and incapable of
accounting for the successes of quantum theories as non-fundamental, effective theo-
ries applicable in a given domain or energy regime. In the same spirit Wallace argues
that interpretational strategies that take the quantum state to be non-representational
(e.g. Richard Healey’s quantum pragmatism, see Ch. 7) fail to make sense of how
quantum physics captures a quark-gluon plasma, for example, involving an interplay
of many concrete quantum theories and their relationships.
J. E. Wolff adopts a broader perspective on the question of what it is to interpret
quantum theory. She contrasts a ‘naturalistic’, science-driven philosophical stance
towards theories with that of the more principled, ‘empiricist’ stance, as represented by
van Fraassen. Regarding the former, the Everettian interpretation favoured by Wallace
is a clear example of an attempt to naturalistically ‘read off ’ ontology from the theory.
However, van Fraassen raises a challenge for the naturalistic stance. Following Maddy
(2007), a ‘naturalistic native’ is someone so deeply immersed in scientific practice that
she approaches all questions, including interpretive ones, from within that practice.
Van Fraassen questions the idea that a naturalistic philosopher can consistently regard
a paradigmatic participant of current science as such a naturalistic native. If such a
native is incapable of adopting a distinctly philosophical interpretation of her own
scientific practice, how will she cope with situations where scientists are more or
less forced to step back and reflect on their aims and methodologies? Paradigmatic
cases of such a situation were involved in the development of quantum mechanics,
and hence a naturalist philosopher must here face van Fraassen’s challenge: on the
one hand, if the naturalistic native cannot engage in such reflection, then she cannot
function as the paradigmatic participant in science, as she will be unable to handle
crisis situations; on the other hand, if she is allowed to step back and reflect, then
she cannot be really characterized as a ‘naturalistic native’. Thus, van Fraassen insists
(from within his empiricist stance) that interpreting theories necessarily requires
stepping outside of science itself, since interpreting a theory like quantum mechanics
involves considering how the world could possibly be the way the theory says it is and
that involves investigating alternatives.
As Wolff suggests, the naturalist, in response, might question whether interpre-
tation necessarily involves stepping outside of science in this way. To this end she
identifies three different ‘moments of interpretation’ that arose in the development
of quantum mechanics: interpretive questions that featured in that very development;
the presentation of alternative views in competition with the ‘orthodox’ interpretation;
and articulating what the world would be like were the theory to be true. With regard
to the first, Wolff argues that this did not require scientists at the time to step back from
their scientific practice or engage in particularly philosophical reflection, so there is
no issue for naturalism here. When it comes to the second ‘moment’, the naturalist
could maintain that the hidden-variables and the dynamical collapse approaches are
actually different theories rather than different interpretations per se. As for the third
project of interpretation, which is the one van Fraassen primarily has in mind, this
does seem to present a problem for the naturalist, insofar as it invites metaphysical
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 7

speculation. One option for the naturalistic philosopher identified by Wolff is to deny
that there is a plurality of such speculative interpretations worth engaging with. Thus
she could follow Wallace, for example, in adopting a ‘literalist’ line and arguing that the
Everett interpretation is the only one that takes the theory literally, thereby rejecting
the basis of van Fraassen’s challenge.
However, Wolff continues, the risk then is that the naturalist might be unable to
accommodate the way that such interpretive projects aid our understanding of the
theory. After drawing a distinction between ‘symbolic’, ‘objectual’, and ‘explanatory’
forms of understanding, she argues that the last, in particular, is not closed off to
the naturalist. Focusing on de Regt and Dieks’ contextualist approach to this form of
understanding, Wolff notes that by characterizing it as an epistemic aim of science—
something the empiricist would reject, of course—this approach would surely look
attractive to the naturalist. And in the quantum context, both the hidden-variables
and dynamical collapse approaches, for example, can be viewed as offering such
explanatory understanding. This underwrites them as appropriate for the naturalist’s
consideration, and thus the naturalistic native is ultimately not precluded from
engaging in various forms of interpretive endeavour.
PART III comprises three chapters that focus on pragmatism about quantum the-
ory, representing a step further beyond traditional conceptions of scientific realism,
but without embracing traditional anti-realism either.
Richard Healey is a key advocate of such a pragmatist interpretation, at the heart
of which lies the rejection of the ‘representationalist’ assumption that a scientific
theory can give us a literally true account of what the world is like only by faithfully
representing the world. As Healey notes, those parts of quantum physics that are
actually used in incredibly successful technological developments such as spintronics,
for instance, are independent of foundational and interpretational issues. If we think
carefully about how quantum mechanics is actually applied to physical systems,
Healey argues, we should see that the name of the game is not the representation of
quantum reality but rather to give us, the users of the theory, appropriate information
about the significance and credibility of claims regarding non-quantum physical
features associated with those systems. Thus, for example, according to Healey the
primary role of the notion of the interpretationally troublesome ‘quantum state’ is
not to represent, or describe, some system, but rather to prescribe how we should
determine the probabilities associated with various measurable eventualities (by
applying the Born rule).
However, Healey insists, this is not a form of instrumentalism or empiricism (of the
sort advocated by van Fraassen, for instance), since it does not rely on any distinction
between ‘observable’ and ‘unobservable’ features of the world; rather, various claims
about unobservable physical magnitudes are significant and true or false, depending
on how the world is. It is the function of (non-quantum) magnitude claims to
represent the relevant features of reality and ultimately it is the truth or falsity of
such claims that we care about. This is still compatible with a ‘thin’ version of the
correspondence theory of truth, in the sense of one that eschews some form of causal
account of reference, so that we’re not misled into thinking that terms appearing in
magnitude claims refer to their subject matter via some form of causal connection.
Indeed, Healey maintains, recent arguments regarding a form of the Bell inequalities
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

8 steven french and juha saatsi

put paid to such thinking. Nevertheless, we can still accept the existence of a physical
world that is independent of our thinking about it. What the pragmatist adds to this
conception is a broader perspective on how we gain knowledge of that world: this
is achieved not via representation per se but, in effect, by taking the theory’s advice
on how the world might be meaningfully represented and what the likelihood is
of such meaningful representations being true. Furthermore, from this pragmatist
perspective quantum theory still helps us to explain a range of otherwise puzzling
phenomena by showing that they were to be expected and also what they depend on.
The following two chapters focus on critical issues about the pragmatist attempt to
construct a ‘middle road’ between realism and instrumentalism.
Lina Jansson focuses on the issue of explanation and the close ties that it has
with scientific realism and argues that from this perspective, Healey’s pragmatist
interpretation comes with significant costs. She begins by inviting us to consider the
putative truism that genuine explanations posit true explanans. This intuitive idea has
to be immediately qualified, however, due to well-known challenges arising from the
roles of idealizations, distortions, and fictions within scientific explanations. A realist
can try to hang onto the gist of the putative truism by appropriately distinguishing the
explanatory from the non-explanatory roles played by different aspects of scientific
explanations, in such a way that the latter’s ontological commitments are tracked.
However, such moves are not open to Healey who rejects, as we have seen, the
claim that quantum models explain by virtue of representing quantum reality, arguing
instead that they explain by virtue of telling us what to expect regarding non-quantum
magnitudes, together with what such magnitudes depend upon. A crucial issue is how
to make sense of this explanatory dependence by the pragmatist’s own lights.
As Jansson suggests, one possibility is to adopt a popular counterfactual approach
based on ‘what-if-things-had-been-different’ questions (in the spirit of James Wood-
ward), while also allowing for non-causal dependencies. However, without causation
to rely on, there is no straightforward way of distinguishing the explanatory theoret-
ical posits from the non-explanatory roles played by idealizations and the like. The
way to proceed, she avers, is to carefully distinguish different kinds of dependence
within the epistemic dependence approach to explanation and, in particular, to look
to what it is that allows us to make the relevant inferences about the counterfactually
robust connections between the initial input of the explanans and the explanandum.
Idealizations, distortions, and fictions can serve to do that, without acting as the
relevant input into the explanans. In Healey’s account, since the quantum state is not
taken to represent the system in question, it cannot serve as such an initial input but
it may nevertheless be indispensable to us in offering the appropriate explanations.
As a result, Jansson argues, crucial information about the physical grounds for the
appropriate assignment of such states has to be effectively ‘black-boxed’, a feature that
she highlights as one of the costs of adopting this form of pragmatist stance.
Peter Lewis also examines the costs of pragmatist approaches—here taken to
embrace also Simon Friederich’s (2015) account—not only with regard to explanation
but also when it comes to our understanding of the content of propositions. In
articulating their position the pragmatists appeal to an inferentialist account of
meaning according to which the meaning of a proposition lies with the material
inferences that it supports, rather than in its representational content. It arguably
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 9

follows from such an account that claims concerning, e.g. quantum states, spin etc.,
are best viewed not as describing physical systems, but rather as prescribing degrees
of belief in non-quantum magnitude claims that do have descriptive content. Lewis
illustrates this by reference to a quantum state associated with a particular molecule.
This quantum state can licence appropriate probabilistic inferences regarding, e.g.
the molecule’s location upon encountering a silicon surface through the application
of the Born rule (underwritten by decoherence). A claim concerning the molecule’s
location on the surface is an example of a non-quantum magnitude claim that has
descriptive, empirical content that is worth asserting, since it supports material
inferences about, e.g. image formation in an electron microscope. The quantum state
itself allegedly does not have such content; any claim about the molecule’s location at a
diffraction grating, for example, would lead to erroneous inferences concerning which
slit the particle is going to go through, for instance. Hence, apart from prescribing
probabilistic inferences supported by the Born rule in situations where the quantum
state decoheres, arguably the state ascription has no content, especially when the Born
rule is inapplicable.
As Lewis notes, one might worry that the distinction between prescriptive and
descriptive content is not supported by the inferentialist account itself and here
perhaps appeals to further elements of the pragmatist toolbox must be made. More
acutely perhaps, Lewis raises the issue that it is not clear how counterfactual inferences
are to be treated on the pragmatist approach: if a diffraction grating were to be
replaced by a silicon surface, we would shift from a situation in which no credences
regarding location can be assigned, the relevant claims being taken to be devoid of
content, to one where definite probabilistic prescriptions can be made, the relevant
claims being contentful. But given that latter point, if counterfactuals contribute to
the content of quantum state attributions, then, Lewis argues, the former claims
should also be understood as having at least some content, contrary to what the
pragmatists assert. One possible response to this worry would be to reconsider the role
of decoherence with regard to this shift in context—rather than delimiting the range
over which claims have content it should be understood as delimiting the range over
which our material inferences can unproblematically draw on our classical intuitions.
Resolving these sorts of worries, together with those concerning explanation, Lewis
concludes, will crucially determine whether this sort of pragmatist approach has
enduring advantages over its realist rivals.
PART IV comprises three chapters that focus on various issues concerning the
nature of the quantum state, standardly taken to be—in contrast to the pragmatist
approach—represented by the wave function. Indeed, advocates of so-called wave
function realism argue that this representational role should place the wave function
at the centre of the scientific realist endeavour.
Alisa Bokulich challenges this view and the ‘hegemony of the wave function’ in
general by presenting a formulation of quantum mechanics that doesn’t make use
of it. This is ‘Lagrangian quantum hydrodynamics’ according to which the state is
represented via the displacement function of a continuum of interacting ‘particles’
following trajectories in spacetime. Schrödinger’s equation is then recast as a sec-
ond order Newtonian law governing such trajectories. Although this formulation is
helpfully motivated by classical hydrodynamics, Bokulich is at pains to emphasize
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

10 steven french and juha saatsi

that it does not require commitment to some notion of a ‘quantum fluid’. Instead
the fundamental state entity via which one can understand the time evolution of the
system is given by the congruence of the trajectories.
As she goes on to note, the centrality of these trajectories in this formulation
suggests an obvious comparison with Bohmian mechanics. However, there are crucial
differences, most notably with regard to the role played by the wave function in
the latter. Furthermore, Bokulich insists, Bohmian mechanics is an interpretation,
whereas Lagrangian quantum hydrodynamics is a formulation, and as such has
entirely different interpretive ambitions (with regard to the measurement problem,
for example). Interestingly, as Bokulich outlines, this alternative formulation reveals
a previously obscured symmetry of quantum mechanics, associated with the infinite-
parameter particle relabelling group, which implies the conservation of quantum
forms of circulation, density, and current. Controversially, perhaps, when transposed
to the relativistic context, the conservation law allows for the definition of global
simultaneity manifolds. More significantly, it is partly because it allows the articula-
tion of this relabelling symmetry that the Lagrangian formulation should be regarded
as more fundamental than the apparently equivalent Eulerian formulation of quantum
hydrodynamics (associated with Erwin Madelung), which retains the wave function
representation of the quantum state.
What does this imply for the various realist projects adopted and pursued in
the context of quantum theory? As Bokulich notes, the formulation not only chal-
lenges the hegemony of the wave function, but also offers a new perspective on
experiments—such as those involving protective measurements—that are invoked as
evidence for its reality. More generally, the existence of the Lagrangian formulation
encourages us to be cautious in reading off our realist commitments from features
of the standard mathematical presentation of quantum mechanics. Finally, one could
also adopt a realist stance towards this formulation itself. Here Bokulich identifies
three possible ways forward. One is to render it an interpretation of the theory, as
in the ‘Many Interacting Worlds’ or ‘Newtonian QM’ views. Another is to adopt a
‘duality’ line towards the quantum state, with the wave functional- and trajectory-
based aspects regarded as a new take on the (in)famous wave-particle duality. The
third approach is what Bokulich calls ‘inferential realism’ which urges a shift in
realist focus from asking ‘what is the world like?’ to ‘what true things can we learn?’
instead. Drawing on Ernan McMullin’s emphasis on the role of metaphors in the
realist enterprise, Bokulich insists that inferential realism is more about developing
a plurality of fertile interpretations than finding the one true picture of the world,
and both the trajectory-based and wave function-conceptions of state feature in
this plurality.
Valia Allori similarly seeks to decentre the wave function in realist approaches to
the theory. Like Bokulich she argues that we should not simply read off our realist
commitments from a given formulation, but instead start the interpretive project with
a ‘primitive ontology’ and construct our interpretation around that. In Newtonian
mechanics the primitive ontology is that of particles, for example, represented by
points in three-dimensional space and our understanding of the theory is grounded
in this. Shifting to the quantum domain, Allori argues that we should retain the
same approach, dropping the representational role of the wave function, not least
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 11

because of the issue of how to understand superpositions. Instead, she maintains,

we should begin with a primitive ontology located in space-time, and select an
appropriate law of evolution for the relevant entities and aim to understand the
theory on that basis. Different such primitive ontologies can then be combined with
different laws of evolution, and Allori considers three kinds of the former: particles,
matter fields and ‘flashes’. This array of alternatives can accommodate a whole slew
of theories, as she sets out. Within this interpretive framework the role of the wave
function is to help implement the law that governs the spatio-temporal evolution
of whatever primitive ontology has been chosen. Thus the wave function can be
regarded as having a nomological character, a suggestion that appears more palatable
to many if understood from a Humean perspective, according to which law statements
are simply the axioms and theorems of our best theoretical system, representing
regularities found in the world. Given the choice of modifying our conception of what
counts as ontology or that of what counts as a law in the quantum context, Allori
prefers the latter.
This general approach meshes well, she argues, with selective realism about the
‘working’ posits of the theory that are responsible for its explanations and predictions.
Here the primitive ontology would supply the working posits, the wave function
counting as a merely ‘presuppositional’ auxiliary that is necessary for the theory’s
mathematical formulation (however, see Bokulich above) but not to be understood
realistically. Nevertheless, in some of the theories canvassed here, there remains a kind
of dependence of the primitive ontology on the wave function and Allori suggests
that this yields a useful way of categorizing solutions to the measurement problem:
in theories of type 1 the primitive ontology and the wave function are independent,
as in particle theories; in theories of type 2 the two are co-dependent and these
include flash and matter density theories. Armed with this distinction Allori goes on
to explore how such theories differ with respect to their super-empirical virtues (e.g.
empirical coherence, simplicity, and relativistic invariance), arguing that Bohmian
interpretations with particles as their primitive ontology and GRW approaches with a
flash ontology should be viewed as the leading contenders, with the former, according
to Allori, just nosing ahead. More importantly and generally, she concludes that once
we get the wave function off centre stage we can more easily explore the different ways
quantum mechanics can be made compatible with realism.
Wayne Myrvold considers a broader set of reasons for denying that quantum states
represent something physically real and argues that at best these provide grounds for
pursuing theories in whose ontologies quantum states don’t appear. Such reasons may
draw on certain classical ‘toy’ models in the context of which apparently quantum
phenomena can be reproduced, such as the existence of pure states that cannot reliably
be distinguished. However, Myrvold notes, these phenomena are at best only ‘weakly’
non-classical and such models cannot capture the Bell inequalities, for example, which
are regarded as exemplars of quantum behaviour. Likewise, he argues, the fact that
quantum mechanics exhibits classical behaviour under certain restrictions is better
regarded as a prerequisite for taking the theory as comprehensive in the first place,
rather than as evidence that quantum states are not real.
Myrvold then goes on to give positive reasons for an ontic construal of quantum
states, within the context of the information theoretic ‘ontological models framework’.
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

12 steven french and juha saatsi

From this perspective, he explores the importance of two theorems that constrain
the set of possible theories that could account for quantum phenomena. The first is
due to Barrett, Cavalcanti, Lal, and Maroney and shows that quantum states cannot
be construed as probability distributions over an underlying state space in such a
way that the operational indistinguishability of such states can be accounted for in
terms of overlap of the corresponding probability distributions. The motivation for
constructing an interpretation under which the quantum states are not ontic is thus
stymied. The second theorem, due to Pusey, Barrett, and Rudolph, demonstrates that
distinct pure quantum states are ontologically distinct. Crucially this assumes the so-
called Preparation Independence Postulate (PIP), which has to do with independent
preparations performed on distinct systems. Myrvold problematizes PIP in relation
to quantum field theory, and proposes that it be replaced with what he calls the
Preparation Uninformativeness Condition (PUC), which, he argues, suffices to show
that distinct quantum states must be ontologically distinct.
Given these results, he concludes, the project—which goes back to Einstein, of
course—of understanding the quantum state in epistemic terms must be abandoned.
Myrvold’s argument hinges on the requirement that the ontological lessons we draw
from a theory should rely only on premises that could reasonably be expected to be
preserved when we shift to the successor theory, in this case quantum field theory.
This raises the further question: how does realism fare when we move to consider
quantum physics beyond the realm of non-relativistic quantum mechanics?
PART V examines various responses to this question, specifically in the context of
quantum field theory.
Doreen Fraser focuses on the example of the Higgs boson as exemplifying the use
of certain formal analogies holding between mathematical structures in the absence
of any physical similarity between the relevant models. This, she suggests, represents a
major challenge to the support that is typically adduced in favour of scientific realism.
The construction of the Higgs model proceeded by drawing formal analogies with the
relevant order parameters in certain models of superconductivity: with regard to the
latter, it is the effective collective wave function of the superconducting electrons,
which distinguishes the normal state of the metal from the superconducting state,
and in the case of the Higgs model, this is the complex scalar quantum field associated
with the Higgs boson.
Fraser argues that the physical dissimilarities of the various elements means that
these analogies must be regarded as purely formal. Thus, for example, the internal
relationships these elements enter into in each model are quite different: the transition
to a superconducting state, involving spontaneous symmetry breaking, is a temporal
process, but there is no analogue of this in the Higgs model.
What then is the explanation for the successful application of such analogies?
Fraser argues that they opened up the space of mathematically conceivable models
by showing that it was possible to incorporate spontaneous symmetry breaking into
one’s model accompanied by massive bosons. Furthermore, because of the physical
disanalogies, crucial features of the superconductivity model were open to experi-
mental investigation, allowing the formal analogies to play a heuristically useful role.
This then presents a fundamental challenge to realism, as the purely formal rea-
soning used to construct the Higgs model was instrumentally successful, yet, Fraser
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 13

insists, the truth or falsity of the theoretical statements asserting the appropriate causal
connections cannot be relevant to explaining its success because there is no plausible
physical analogy underpinning them. Thus we seem to have an example of scientific
success that cannot be accommodated in realist terms.
Fraser concludes by noting that realists and anti-realists alike tend to draw on
diachronic sequences of theories in defence of their opposing claims. However,
she argues, what the Higgs case study demonstrates is that when it comes to the
development of specific quantum theories (and here we might recall Wallace’s point
above), it is also synchronic relationships that need to be considered, involving new
sets of challenges.
James Fraser is more sanguine, insisting that despite the challenges, we can give
a realist reading of quantum field theory. However, he argues, restricting our atten-
tion to perturbative or axiomatic treatments is unhelpful in that regard. It is the
former approach that lies behind the striking empirical predictions of the theory,
including many of those tested at the Large Hadron Collider, for example. Yet the
underlying strategy is famously problematic and has been widely regarded as lacking
in mathematical rigour, depending as it does on the removal of certain infinities in
a suspiciously ad hoc manner. Indeed, it leaves the realist unable to specify what the
theory says about the world. In desperation, perhaps, one might turn to the so-called
axiomatic approach that at least gives a clear set of theoretical principles for the realist
to work with. Unfortunately, as is well known, these principles can only be used as the
framework for certain physically unrealistic ‘toy’ models.
All is not lost, however. As Fraser notes, developments in renormalization theory
offer a way forward and here he sketches the core features of the momentum space
approach—in particular the way in which certain coarse-grained transformations
induce a ‘flow’ on the space of possible theories which offers information on the
behaviour of systems at different scales. These systems, modelled by QFT, display a
feature known as ‘universality’, whereby models that display very different behaviour
at high energies manifest very similar physics at lower energy levels. What this means
is that if the high energy degrees of freedom are removed, as in a ‘cut-off ’, this will leave
the lower energy behaviour more or less unaffected. This in turn allows the realist
to ‘bracket off ’ what the world is like at the fundamental level, while still accurately
modelling its lower energy properties.
This then helps to justify the various steps of the perturbative renormalization
procedure. Thus, for example, it justifies the absorption of the physics beyond a
certain ‘cut-off ’ point into an effective action and reveals that what this procedure
is really about is ensuring the right kind of scaling behaviour exhibited by the system
in question. Finally it pragmatically justifies taking the cut-off to infinity, which yields
significant computational benefits.
Given this procedure, Fraser argues that the renormalization group offers a way
of developing a selective realist reading of QFT, according to which we should be
realist about those constituents that underwrite a theory’s predictive success. In
particular, it reveals that certain features play no role in that success and can be set
aside as far as the realist is concerned. However, it also helps the realist articulate
the relevant positive theoretical commitments, with regard to relatively large-scale,
non-fundamental aspects of the world. It shows, for example, that they are largely
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

14 steven french and juha saatsi

insensitive to the details that obtain at high energies and hence can be considered
‘robust’ and thereby worthy of realist commitment.
Nevertheless there remain challenges. Thus, for example, even granted the robust-
ness of low-energy features of the relevant models, it remains unclear what aspects of
the world they are latching onto. Here, Fraser argues, the realist needs to pay further
attention to such claims about the non-fundamental and consider more carefully the
terms in which they are characterized. This is in addition to the more well-known
concerns regarding what the world is like according to quantum theory, as canvassed
in this volume, as well as the additional puzzle posed in the context of QFT by the
existence of unitarily inequivalent Hilbert space representations. As Fraser concludes,
such puzzles and concerns highlight the need for a comprehensive re-examination of
realist strategies in general.
Laura Ruetsche agrees that the development of the Standard Model and the
quantum field theory that underpins it present a range of new challenges to realism.
She focuses on one of the strategies indicated by James Fraser, namely adopting a
selective attitude as embodied in what she calls ‘effective realism’. As she notes, this
takes seriously the point that our best current theories are merely ‘effective’ in the
sense that they’re not true across all energy regimes and uses the renormalization
group as a means of motivating the core ‘divide et impere’ move of such an attitude.
Unfortunately, she points out, the action of the renormalization group is defined on
a specific space of theories and whatever the virtues are of our best current models, if
the true, final theory lies outwith that space, then, as she puts it, all bets are off.
Even more worryingly, Ruetsche notes that it is not clear how realist ‘effective
realism’ is! Through an examination of various features of effective theories, she
concludes that in order to distance herself from the empiricist, the effective realist
must approach such features in the light of certain interpretive projects. So, for
example, the effective realist might endorse particles corresponding to fields that are
robustly present in the relevant Lagrangian at a certain length scale, but to do so she
must engage in interpretive manoeuvres that are typically articulated in terms of a
theory’s truth conditions and which she supposedly repudiates.
However, Ruetsche suggests, the effective realist needn’t disavow such interpretive
work per se, as long as she is mindful of the distinction between asking what the
world is like according to a given theory, and asking why that theory is so successful.
According to the view Ruetsche labels ‘fundamentalism’, the answer to the second
question is given in terms of the answer to the first: the theory is so successful
because it accurately describes how the world is. When it comes to effective theories,
however, she argues that this intertwining of the answers is a mistake because what
explains the success of an effective theory is something exogenous to it. Recognizing
that and rejecting fundamentalism then brings effective realism closer to what she
calls a ‘humble’ form of empiricism that explains a theory’s success in terms of
its approximation to the predictions of some final theory within experimentally
accessible regimes. The humble empiricist accepts that we can give an explanation
of a given theory’s success, just not in terms of its truth and that we can entertain
the possibility of some true, final theory but that we should adopt an agnostic stance
towards a given effective theory’s set of unobservables.
 OUP UNCORRECTED PROOF – REVISES, 27/12/2019, SPi

introduction 15

To conclude, Ruetsche suggests that whether the commitments of effective realism

actually count as realist or not depends on how they’re understood. It is better,
she maintains, to embrace a stance of humble empiricism that has the resources to
accommodate the myriad ontological subtleties of quantum physics.
* * *
This collection of chapters is thus ‘book-ended’ by attempts to shift realism away from
the traditional conception in the face of the multiple problems posed by quantum
physics. The nature and extent of that shift varies from author to author—in some
cases it involves a move away from the foundational, in others dropping the emphasis
on the truth-content of theories, and in yet others it requires some form of non-literal
‘reading’ of quantum theory. There are a variety of options ‘on the table’ and what
realists in general need to do now is not just take quantum physics seriously but to
continue articulating, defending and contrasting these epistemic alternatives along
the lines presented in this volume, which we hope will come to be seen as a significant
step in the right direction.

References
Becker, A. (2018). What Is Real?: The Unfinished Quest for the Meaning of Quantum Physics.
New York: Basic Books.
Cordero, A. (2001). Realism and underdetermination: some clues from the practices-up.
Philosophy of Science 68, S301–S312.
Fine, A. (1984). The natural ontological attitude. In J. Leplin (Ed.), Scientific Realism,
pp. 83–107. Los Angeles: University of California Press.
Friederich, S. (2015). Interpreting Quantum Theory: A Therapeutic Approach. London: Palgrave
Macmillan.
Maddy, P. (2007). Second Philosophy. New York: Oxford University Press.
Saatsi, J. (Ed.) (2018). The Routledge Handbook of Scientific Realism. London: Routledge.