Hinges, Disagreements, and Arguments: (Rationally) Believing Hinge Propositions and Arguing across Deep Disagreements

Abstract

Wittgenstein famously introduced the notion of ‘hinge propositions’: propositions that are assumptions or presuppositions of our languages, conceptual schemes, and language games, presuppositions that cannot themselves be rationally established, defended, or challenged. This idea has given rise to an epistemological approach, ‘hinge epistemology’, which itself has important (negative) implications for argumentation. In particular, it develops and provides support for Robert Fogelin’s case for deep disagreements: disagreements that cannot be rationally resolved by processes of rational argumentation. In this paper, I first examine hinge epistemology in its own right, and then explore its implications for arguments and the theory of argumentation. I argue that (1) the Wittgensteinian approach to hinge propositions is problematic, and that, suitably understood, they can be rationally challenged, defended, and evaluated; (2) there are no well-formed, coherent propositions, ‘hinge’ or otherwise, that are beyond epistemic evaluation, critical scrutiny, and argumentative support/critique; and (3) good arguments concerning hinge propositions are not only possible but common. My arguments will rely on a thoroughgoing fallibilism, a rejection of ‘privileged’ frameworks, and an insistence on the challengeability of all frameworks, both from within and from without.

Full text

Vol.:(0123456789)
1 3
Topoi
https://doi.org/10.1007/s11245-018-9625-6
Hinges, Disagreements, andArguments: (Rationally) Believing Hinge
Propositions andArguing acrossDeep Disagreements
HarveySiegel1
© Springer Nature B.V. 2019
Abstract
Wittgenstein famously introduced the notion of ‘hinge propositions’: propositions that are assumptions or presuppositions of
our languages, conceptual schemes, and language games, presuppositions that cannot themselves be rationally established,
defended, or challenged. This idea has given rise to an epistemological approach, ‘hinge epistemology’, which itself has
important (negative) implications for argumentation. In particular, it develops and provides support for Robert Fogelin’s
case for deep disagreements: disagreements that cannot be rationally resolved by processes of rational argumentation. In this
paper, I first examine hinge epistemology in its own right, and then explore its implications for arguments and the theory
of argumentation. I argue that (1) the Wittgensteinian approach to hinge propositions is problematic, and that, suitably
understood, they can be rationally challenged, defended, and evaluated; (2) there are no well-formed, coherent propositions,
‘hinge’ or otherwise, that are beyond epistemic evaluation, critical scrutiny, and argumentative support/critique; and (3) good
arguments concerning hinge propositions are not only possible but common. My arguments will rely on a thoroughgoing
fallibilism, a rejection of ‘privileged’ frameworks, and an insistence on the challengeability of all frameworks, both from
within and from without.
Keywords Deep disagreements· Hinge propositions· Justification· Pritchard· Wittgenstein
1 Wittgenstein and‘Hinge Epistemology’
1.1 What Are ‘Hinge’ Propositions?
Hinge propositions, also called ‘framework’ propositions,
are propositions that are the presuppositions of particular
‘language games’ or ‘systems of judgment’. Examples of
hinge propositions include ‘the earth exists’, ‘every human
being has parents’, and ‘there are physical objects’ (Witt-
genstein 1969, §§208, 211, 35). They are, in another Witt-
gensteinian metaphor, the ‘bedrock’ beyond which rational
judgment cannot go. On the Wittgensteinian view, they are
epistemically fundamental. More than that, they are episte-
mologically primitive, in that they are immune to rational
scrutiny and critique: “…all enquiry on our part is set so
as to exempt certain propositions from doubt, if they are
ever formulated. They lie apart from the route traveled by
enquiry” (Wittgenstein 1969, §88, emphasis in original).
Since they are presuppositions of rational thought, they can-
not be challenged without being relied upon to launch the
challenge, and so are beyond rational challenge.
Putting it in the way I just did makes it sound as though
Wittgenstein’s argument for the fundamental status of hinge
propositions is either ‘transcendental’ (that is, relying on the
necessity of presupposing the hinge proposition in order for
it to be possible to challenge it) or a reductio ad absurdum
(in that supposing that a hinge proposition can be challenged
requires presupposing that proposition itself, thus contradict-
ing the supposition). But both these readings are misleading.
Wittgenstein’s claim is rather that hinge propositions are
beyond rational challenge: they belong to “the scaffolding
of our thoughts” (Wittgenstein 1969, §211, emphasis in
original) and are “part of the whole picture which forms the
starting-point of belief for me” (Wittgenstein 1969, §209,
emphasis in original). Such propositions are exempt from
enquiry, Wittgenstein argues, both because they are presup-
posed by enquiry and because they are more certain than
any reason we might have to doubt them (Wittgenstein 1969,
§§32, 111, 125, 162–163, 243, 247–253, 282, 307, 341).
* Harvey Siegel
hsiegel@miami.edu
1 Department ofPhilosophy, University ofMiami, P.O.
Box248054, CoralGables, FL33124-4670, USA

H.Siegel
1 3
On this view, “justification comes to an end” (Wittgenstein
1969, §192; cf. Wittgenstein 1958, 1: §§1, 485)—in the end
one has to say simply “my spade is turned”; “this is simply
what I do” (Wittgenstein 1969, §217). For these reasons,
challenging hinge propositions is ‘unintelligible’ and ‘inco-
herent’. His argument involves what is sometimes called ‘the
structure of justification’. I turn to that argument next.
1.2 Wittgenstein’s Argument forHinges
Wittgenstein’s epistemology is, like most positions in phi-
losophy, subject to alternative interpretations and analyses; I
won’t discuss the complex literature his account has spawned
here. Instead, I will rely, in addition to Wittgenstein’s writ-
ings themselves, mainly on Duncan Pritchard’s recent treat-
ments (Wittgenstein 1958, 1969; Pritchard 2005, 2009,
2011, 2016). Pritchard’s discussions are I think unusually
clear and insightful, and will be helpful in articulating the
key Wittgensteinian claims. Pritchard is mainly concerned
to use Wittgenstein as part of a general strategy to defeat
epistemic skepticism1; while skepticism will invariably arise
in what follows, I will mostly bypass skeptical concerns in
order to focus on Wittgenstein’s claims concerning the limits
of rational evaluation, since these are the ones most relevant
to argumentation and the possibility of rationally resolving
deep disagreements.
1.2.1 Groundlessness
Perhaps the most central Wittgensteinian claim is that our
most fundamental beliefs are groundless: not capable of
rational justification (or rational doubt), and yet not in need
of such justification. As Wittgenstein puts it, “The difficulty
is to realize the groundlessness of our believing” (1969,
§166).2 As Pritchard puts it, “Wittgenstein maintains that
that which we are most certain of must be by its very nature
rationally groundless” (2016, p.64).
Why does Wittgenstein think this? A key reason is that
on his view justification must travel from the more certain
(which is, because more certain, less in need of justification)
to the less certain (which is, because less certain, more in
need of justification), and that our most fundamental (hinge)
beliefs are by their very nature maximally certain, so that
they cannot be justified in terms of anything less certain. As
Pritchard summarizes the case:
…not only are these Moorean certainties [e.g., ‘here
is one hand…’] necessarily groundless, according to
Wittgenstein, but it also seems that they are by that
same token immune to rational doubt. For any rational
basis for doubting the Moorean certainty would be
necessarily less certain than the optimally certain
Moorean certainty, and hence one would have more
reason to doubt the grounds offered for doubting the
Moorean certainty than to doubt the Moorean certainty
itself. At the very least, what Wittgenstein seems to
be suggesting in this passage is that there could be no
rational basis that would mandate doubt of a Moorean
certainty, since one rational response to the presenta-
tion of this ground for doubt could simply be to doubt
the ground itself. That claim falls short of the stronger
thesis that rational doubt of a Moorean certainty is
impossible, but it is even so a dramatic claim to make.
It soon becomes clear, however, that Wittgenstein
wants to defend the stronger thesis. That is, that not
only are Moorean certainties necessarily groundless,
but that also rational doubt of a Moorean certainty is
simply impossible (i.e., as opposed to being merely
rationally unmandated). Wittgenstein claims that the
very idea of a rational evaluation, whether positive or
negative, presupposes a backdrop of Moorean certain-
ties that are themselves exempt from rational evalua-
tion. To attempt to rationally evaluate a Moorean cer-
tainty is thus an attempt to do something impossible.
In particular, Wittgenstein repeatedly urges that the
very idea of rationally doubting a Moorean certainty
is incoherent. Such a doubt, he writes, would “drag
everything with it and plunge it into chaos.” (Pritchard
2016, pp.65–66, emphasis in original, quoting Witt-
genstein 1969, §613)
I will continue to explicate Wittgenstein’s view below. But
there is already reason for doubt. For one thing, the case
for the groundlessness of hinge propositions and the limits
of rational justification just rehearsed is less than compel-
ling. In a passage from On Certainty that Pritchard discusses
in these pages, Wittgenstein writes: “If a blind man were
to ask me ‘Have you got two hands?’ I should not make
sure by looking. If I were to have any doubt of it, then I
don’t know why I should trust my eyes” (1969, §125). Let
us grant Wittgenstein his claim that if I were to doubt that
I had two hands, checking by looking might not quiet my
doubt. (Even this is dubious; whether or not it would help to
quiet my doubt would I think depend on the source of that
doubt. If I doubted upon waking in a hospital bed with the
1 Pritchard’s ‘biscopic’ approach to defeating skepticism utilizes
hinge epistemology to defeat closure-based skeptical arguments,
while McDowell-inspired disjunctivism is used to defeat underdeter-
mination-based skeptical arguments. Pritchard’s complex anti-skep-
tical case is well worth detailed study, though it will not occupy us
here.
2 Pritchard pairs this quotation from On Certainty with a parallel
from John Henry Newman: “None of us can think or act without the
acceptance of truths, not intuitive, not demonstrated, yet sovereign.”
(Pritchard 2016, p.61; 2005, p. v.)

Hinges, Disagreements, andArguments: (Rationally) Believing Hinge Propositions andArguing…
1 3
lower portions of both my arms appearing shorter and heav-
ily bandaged, and hearing whispered conversations about
amputation, I might well quiet my doubt by asking that the
bandages be removed and looking.) Nevertheless, I could
legitimately and helpfully answer the blind man by answer-
ing ‘Yes’ and basing that answer on my sight and other sen-
sory evidence. (‘I can see them, feel them, grasp things with
them, touch you with them (can’t you feel my hand on your
arm?)’, etc.) So even if Wittgenstein is right that appealing
to my sight won’t help me quiet my own doubts, so appeal-
ing may well be a legitimate, informative, and justificatory
answer to the blind man. It is not impossible to justify my
belief in that particular hinge proposition.3 If so, then it is
false that hinge propositions always are, let alone must be,
groundless.4
1.2.2 Direction ofJustification
But a more basic worry is that it is not clear why justifi-
cation must always go, as Wittgenstein’s argument presup-
poses (in fairness, along with many others), from the more
certain to the less certain. For example, I can be justified in
believing that you have a mind on the basis of our conver-
sations (your responsiveness, facial expressions, etc.) even
though I have no general solution to the problem of other
minds, and even though I acknowledge that that basis does
not disarm skeptical doubts of the ‘but your evidence can’t
distinguish between my mind’s being real and its being a
mere appearance, an imposition of an evil deceiver; that
evidence is equally compatible with both scenarios’ vari-
et y. 5 Non-philosophical folk who have no knowledge of that
problem, or of skeptical hypotheses, surely are justified in
believing that their friends and family members have minds,
even though the evidence for those beliefs is not more cer-
tain than the skeptical alternatives. If so, then it is not true
that the direction of justification must always flow from the
more to the less certain. Moreover, even minimally justi-
fied reasons or evidence can add to the justification of more
highly justified (and so more certain) propositions. ‘I think
I saw it too’, said of a faint streak of light indicating the
path of a comet momentarily visible through a telescope,
can enhance the justificatory status of propositions, e.g., ‘a
comet just passed near Ursa Major as viewed from North
America’, that are more certain—enjoy higher justificatory
status, because many others, in lots of locations, using many
different telescopes, viewed it too, and their collective sight-
ings confer substantial justification—than the minimally jus-
tified justifier itself. Here, too, the less certain can enhance
the justificatory status of the more certain. The ‘direction of
justification’ thesis Wittgenstein relies on is false.6
1.2.3 Locality ofJustification
In any case, Wittgenstein holds that rational evaluation is
necessarily local, in that it can take place only against the
backdrop of hinge propositions that are taken for granted in
the context of evaluation. Here is Wittgenstein:
[T]he questions that we raise and our doubts depend
upon the fact that some propositions are exempt from
doubt, are as it were like hinges on which those turn.
3 My answer may not serve as a justifier if the blind man’s question
is posed in a skeptical spirit. But my concern is not with skeptical
doubts here. Pritchard grants the point just made in the text but, cor-
rectly in my view, denies that it can play “the remarkable anti-skep-
tical role” (2016, p.85) that Moore’s ‘I have two hands’ is meant to
play in Moore’s anti-skeptical effort.
4 As a reviewer correctly points out, the discussion of the blind man
in this section is not by itself sufficient to upend Wittgenstein’s claim
concerning groundlessness. That discussion continues in the follow-
ing sections. Another reviewer worries that my discussion fails to
take account of the fact that ‘hinges’ involve not specific, unchang-
ing propositional contents but rather particular commitments. This is
addressed in sections2.1–2.3 below. I thank both reviewers for their
stimulating challenges.
5 Pritchard (2016) insightfully discusses this skepticism-inspiring
sort of consideration under the label of the ‘insularity of reasons’
thesis, according to which the support perceptual evidence offers to
a given claim is indistinguishable in the ‘good case’, in which I actu-
ally see what I think I see, and the ‘bad case’, in which I seem to but
don’t actually see it (because, for example, I’m deceived by the evil
genius).
6 A referee doubts that Pritchard embraces the ‘direction of justifi-
cation’ claim just discussed, and doubts that Pritchard attributes it
to Wittgenstein. I should note that (1) that claim is ascribed here to
Wittgenstein, not Pritchard, and (2) the reason Wittgenstein offers for
thinking that optimal certainties are beyond rational doubt is precisely
that such certainties are always less susceptible to rational doubt than
any considerations that might lead us to doubt them. As Pritchard
puts it, in explicating Wittgenstein’s claim, Moorean certainties are
“immune to rational doubt [, f]or any rational basis for doubting the
Moorean certainty would be necessarily less certain than the opti-
mally certain Moorean certainty, and hence one would have more rea-
son to doubt the grounds offered for doubting the Moorean certainty
than to doubt the Moorean certainty itself.” That is, on Wittgenstein’s
view something that is less certain cannot successfully provide a rea-
son for doubting something that is more certain. Why not? Because
the direction of justification must go from the more to the less certain.
That thesis explains why Wittgenstein thinks that optimally certain
claims cannot be rationally doubted: we are more justified in doubting
the grounds offered for doubting the Moorean certainty than in doubt-
ing that certainty itself. If Wittgenstein rejects the direction of justifi-
cation thesis, his argument for thinking that optimal certainties cannot
be rationally doubted fails. My claim is that that thesis is false. As far
as Pritchard attributing the thesis to Wittgenstein goes, cf. Pritchard’s
explication, 2016, pp.64, 65, concerning “epistemic basis”, “rational
grounds”, etc. If Pritchard does not attribute the thesis to Wittgen-
stein, how are we to understand the arguments in which those phrases
occur?

H.Siegel
1 3
That is to say, it belongs to the logic of our scien-
tific investigations that certain things are in deed not
doubted.
But it isn’t that the situation is like this: We just can’t
investigate everything, and for that reason we are
forced to rest content with assumption. If I want the
door to turn, the hinges must stay put. (1969, §§341–
343, emphases in original)
Just as the hinges must stay put if the door is to turn, so
(according to Wittgenstein) our questions and doubts
depend upon hinge propositions that are beyond or exempt
from doubt. Inquiry depends upon those propositions being
beyond doubt—for if they are doubted, they drag everything
with it and plunge it into chaos. This entails, Wittgenstein
holds, that wholesale doubt is impossible, not just in fact
but in principle: without fixed hinges, the door of inquiry
can’t turn. Pritchard labels the view Wittgenstein here rejects
‘the universality of rational evaluation’ thesis, according to
which there are “no in principle constraints on the extent
of one’s rational evaluations”; “no in principle limitation
on rational evaluation” (2016, pp.55, emphasis in origi-
nal; 171). Instead, Wittgenstein holds, contrary to this ‘uni-
versality of rational evaluation’ thesis, that there is indeed
such a limitation: our rational evaluations can’t extend to
the hinge propositions themselves. They are beyond doubt,
beyond rational evaluation; they are groundless and can-
not be justified, even in principle, but they are nevertheless
both beyond rational doubt and the necessary foundation
upon which all inquiry and all rational evaluation depends.
As Pritchard puts it, “the very idea of a wholesale rational
evaluation is itself unintelligible, for it is in the very nature
of rational evaluations that they take place relative to hinge
commitments that are both groundless and indubitable”
(2016, p.66); “the very idea of a wholesale rational evalua-
tion is itself incoherent” (2016, p.70); “the very possibility
of rationally evaluating one’s beliefs presupposes arational
hinge commitments that are beyond rational evaluation”
(2016, p.174).
Is this right? Is Meditation 1, in which Descartes imagi-
nes that he is dreaming or being deceived by a powerful evil
genius intent on deceiving him, and thereby undertakes ‘a
wholesale rational evaluation’ of his beliefs, unintelligible
or incoherent? There doesn’t seem to be anything unintel-
ligible or incoherent about it. Descartes’ general doubts and
wholesale evaluations—as well as those by philosophers
from long ago and from more recent and contemporary fig-
ures as well—are not only perfectly coherent and intelligi-
ble, they are what many of us cut our philosophical teeth on,
and what we encourage our students to cut theirs on. If so,
then ‘universal’, non-local justification is possible, and such
justification can be sought for hinge propositions as much as
for non-hinge propositions.7
1.2.4 Indubitability
Moreover, why think that “all doubt takes place relative to
our arational hinge commitments, which are beyond doubt”
(2016, p.69)? It is, after all, quite counterintuitive, since we
do seem perfectly able to call particular hinge propositions
into question—that is, to subject them to principled doubt
and inquire into the strength of the cases that can be offered
for and against them. Indeed, we do so in philosophy all the
time. Is Berkeley’s denial of the existence of physical objects
unintelligible or incoherent? It seems not: his arguments are
sophisticated and clearly in need of attention if the hinge
proposition concerning the existence of physical objects is to
be rationally vindicated or maintained. If so, it is manifestly
not indubitable—not beyond doubt, rational or otherwise.
Similarly for ‘every human being has parents’ in this age of
gene editing, laboratory created designer macromolecules,
and other forms of genetic engineering. Is it really ‘incoher-
ent’ or ‘unintelligible’ to imagine, now or in the foresee-
able future, a human being created in the laboratory with
no parental egg, sperm, or DNA in its causal history? If
not, then why think that our hinge beliefs or commitments
are either certain, beyond challenge, or groundless?8 I won’t
enter into a serious defense of fallibilism here, but Witt-
genstein’s view seems far less plausible than the fallibilistic
alternative.
8 Pritchard introduces the notion of ‘über hinge’ propositions and
commitments to avoid this worry about the changeability of other,
non-über hinge propositions and commitments: the “entirely gen-
eral” über hinge commitment is simply “that one is not radically
and fundamentally mistaken in one’s beliefs”; and our other, “per-
sonal” hinge commitments “codify” the über one (2016, p. 95; cf.
pp.95–98). This is an ingenious move, and I’m happy to grant that
it helps Pritchard’s attempt to undercut skepticism along partly Witt-
gensteinian lines, although the idea of ‘codification’ in play here is
in need of further explication. Whether or not it helps establish the
claims that hinge propositions are groundless, local, indubitable, etc.
remains to be seen.
7 A reviewer correctly points out that Descartes had to hold some
things fixed (for example, that what he thought he was doing was
what he was actually doing; that he was capable of rational thought;
etc.) in order to carry out his thought experiment. But, as argued
below, this does not establish the ‘fixedness’ of hinges, since these
things, held fixed here, can in turn be ‘unfixed’ and subjected to
rational scrutiny. Another reviewer holds, bizarrely in my view, that
apparent debates and arguments about skeptical theses are necessarily
“fake” and “pretense”—Descartes (in Meditation 1), Hume, Unger,
and other skeptics weren’t in fact doing what they took themselves
to be doing, but only went “through the motions of doubt”, and did
“not express genuine doubts at all”. Apparently, doubt is necessarily
“fake” because universal rational evaluation is impossible, at least on
the Wittgensteinian view. But that view is itself untenable—so I’m
arguing, at any rate.

Hinges, Disagreements, andArguments: (Rationally) Believing Hinge Propositions andArguing…
1 3
1.2.5 Certainty
So far I have been treating as unproblematic Wittgenstein’s
various claims concerning certainty. But it must be said that
that word, in his hands, is multiply ambiguous. Consider my
belief that I have two hands. Is this belief of mine certain? It
is not, if ‘certain’ means ‘indubitable’—incapable of being
doubted—since I can readily doubt it, simply by putting
myself in a Cartesian frame of mind or finding myself in the
hospital bed imagined earlier. Does it mean ‘incorrigible’—
necessarily true, or incapable of being mistaken? My belief
is neither of these, since it is possible, both logically and
metaphysically, for it to be both false and mistaken.
Moreover, much of Wittgenstein’s discussion treats cer-
tainty as a matter of degree: beliefs can be more certain,
less certain, maximally certain, optimally certain, etc. This
understanding renders it a psychological (rather than epis-
temological or metaphysical) notion, indicating something
like strength of belief.9 But this psychological understanding
of it robs the notion of the ability to do the work Wittgen-
stein wants it to do: my believing something with maximal
strength or degree of conviction does not render it indu-
bitable, incorrigible, or incapable of either justification or
rejection.
So when Wittgenstein urges that belief presupposes
certainty—“Wittgenstein was also keen to emphasize the
point that belief presupposes certainty, to the extent that
widespread doubt is incompatible with belief of any kind”
(Pritchard 2016, p.112)—I cannot agree. For one thing, how
‘widespread’ must the doubt be before I become incapable
of believing? If I doubt everything the current U.S. Presi-
dent says about the state of international relations, the U.S.
economy, Nancy Pelosi, and all matters social/political, can I
not nevertheless believe either that I have two hands or that I
had scrambled eggs for breakfast? For another, can’t I enter-
tain even completely general and systematic doubt, as every
beginning philosophy student must, while maintaining my
belief about what I had for breakfast? (‘I believe I had the
eggs, even though I acknowledge that I would be mistaken
if I were deceived by the evil genius.’) Most importantly,
can I not maintain a thoroughgoing fallibilism, as Peirce,
Popper, and many others urge, according to which all my
beliefs—even the hinges—might be mistaken? It seems I
can do all these things. I don’t have to regard my beliefs that
I have two hands, or that I had those eggs, to be either indu-
bitable (because I can doubt them) or incorrigible (because
they are, if true, contingently true, and because they can be
mistaken—I may have confused this morning’s breakfast
with yesterday’s) in order to genuinely believe them.10
1.2.6 What About ‘Language Games’?
One plausible reading of Wittgenstein is that hinge proposi-
tions are groundless within a given language game or system
of presuppositions/judgment, but are evaluable/criticizable/
supportable from some other language game or perspective.
I am myself in some sympathy with this way of thinking
about hinges: in any given argumentative context, some-
thing must be taken as fixed by the arguers, at least at that
moment, in order for the argumentation to be productive. For
example, we can’t helpfully argue about the likely future of
the European Union in the light of Brexit if we don’t agree
on what ‘EU’ designates or refers to. Similarly, we can’t
productively argue about the prospects for the ‘denucleariza-
tion’ of the Korean Peninsula, in light of the recent summit
in Singapore, if we have different understandings of that
word (as the protagonists apparently do). But this sort of
relativity to a language game is innocuous, epistemologi-
cally, if we can move to a different language game and evalu-
ate the hinges in question from it. As Popper says,
I do admit that at any moment we are prisoners caught
in the framework of our theories; our expectations; our
past experiences; our language. But we are prisoners
in a Pickwickian sense: if we try, we can break out of
our frameworks at any time. Admittedly, we shall find
ourselves again in a framework, but it will be a better
and roomier one; and we can at any moment break out
of it again… The central point is that a critical discus-
sion and a comparison of the various frameworks is
always possible. (1970, p.56)
On this understanding of hinges as context- or language-
game bound, the epistemological air leaks out of the Witt-
gensteinian balloon: even if they are groundless within, they
are nevertheless perfectly evaluable without. That is, while
they may be groundless and/or certain within the bounds
of some particular language game, framework, or system
of presuppositions, they are perfectly defensible and chal-
lengeable once one moves into another, “better and roomier”
language game. In this case, their groundlessness and their
9 In addition to the many passages of On Certainty already cited, cf.
Pritchard (2016, e.g. pp. 64–66 and 85), in which all these various
understandings of ‘certain’ may be found.
10 A referee urges that according to Wittgenstein, ‘our certainty in
the hinges is revealed in our actions.’ But this doesn’t help, since we
often—indeed, always, if ‘certainty’ is understood non-psychologi-
cally—act without certainty. I head to the assigned classroom at time
t because I believe that the class is scheduled to begin at t + 5min-
utes. Does my action reveal my certainty? Certainly not—I’ve got the
time wrong too many times to be certain of any such thing. Actions
no more reveal certainty than do beliefs.

H.Siegel
1 3
certainty are equally deflated, and they are easily reconcil-
able with the fallibilistic alternative.11
1.2.7 Is Self‑Reflexivity aProblem?
Finally, what happens when we apply Wittgenstein’s view
to itself? If it is right, the case for it itself depends upon
some hinge proposition(s) or other, and those hinges are
groundless. They can’t be Wittgenstein’s examples of
hinges (‘the earth exists’, ‘every human being has parents’,
etc.), because nothing about the structure of justification is
entailed by them. Similarly, it can’t be Pritchard’s über hinge
commitment, namely “that one is not radically and funda-
mentally mistaken in one’s beliefs” (2016, p.95; cf. note 8),
both because nothing about the structure of justification is
entailed by it, either, and because it is manifestly not ground-
less or certain but rather eminently subject to philosophical
defense and critique, as Pritchard’s sophisticated discussion
exemplifies. If Wittgenstein is right that they are ground-
less, then it seems that philosophers who dispute Wittgen-
stein’s view have every right to judge that his case is without
merit precisely because it rests on hinges that are themselves
groundless, and in that sense unjustified.12 I won’t develop
this point further here, but it seems to me to be something
that should give Wittgensteinians pause.13,14
1.2.8 Why Go This Route?
So why does Pritchard accept the Wittgensteinian view and
reject non-locality, universality, the possible groundedness
of hinge propositions, or the seemingly straightforward argu-
ments against them offered in the previous sections? One
clear reason is that that rejection is integral to his systematic
case against the radical skeptic. That case, as I’ve already
granted, is skillfully developed and philosophically sophisti-
cated. But since our concern here is argumentation and most
argumentation does not concern skepticism, I hope that I can
set this reason aside here.
A second reason is that his diagnosis of skepticism shows,
he claims, that the problem results from faulty theorizing:
“the very idea of a universal rational evaluation (whether
positive or negative, Moorean or radically skeptical) is sim-
ply incoherent, and the product of faulty theorizing” (2016,
p.174).15 But the fault of this ‘faulty theorizing’ is never
clearly specified, beyond its inconsistency with the Wittgen-
steinian view Pritchard articulates and defends but that has
been problematized above. Indeed, it seems that its fault, if
genuine, lies in (1) contravening Wittgenstein’s alternative,
groundless and local-only view, and/or (2) not enabling us
to gain the upper hand against the skeptic. But the first of
these leaves unclear why we should buy that view, given the
problems delineated thus far, and the second seems clearly
enough to beg the question against the skeptic, since ‘reject
the possibility of universal evaluation, otherwise we’re stuck
with the skeptical problem’ appears to presuppose, rather
than establish, anti-skepticism.16
A third reason involves his positive anti-skeptical view:
By focusing on the über hinge commitment [“that one
is not radically and fundamentally mistaken in one’s
beliefs”; cf. note 8 above] we can see why rational
evaluation must be essentially local. For what possible
11 Thanks here to Al Neiman for helpful e-correspondence.
12 A referee wonders why this should be a problem for Wittgenstein.
The preceding sentence expresses the problem.
13 The connection here to the problem of epistemological relativism
is evident. On relativism generally, cf. Siegel (2004). Whether or not
Wittgenstein’s view is relativistic in an epistemically pernicious way
is a question widely discussed by Wittgenstein scholars; cf. Pritchard
(2009, 2011, 2016, pp. 109–110), and the papers in Coliva and Ped-
ersen (2017), esp. Carter (2017).
14 The main reason for thinking that Wittgenstein is right about all
this, at least as far as Pritchard’s analysis goes, is that his view ena-
bles the development of Pritchard’s original, sophisticated and plau-
sible undercutting of the skeptic’s challenge: his ‘biscopic proposal’
(2016, pp. 173–175 and passim), which promises to relieve us of
epistemic angst. Key to Pritchard’s, and also fellow hinge episte-
mologist Annalisa Coliva’s (2015) articulations and defenses of hinge
epistemology, is their taking Wittgenstein to be responding to radical
skepticism, according to which we don’t, or more strongly can’t, have
knowledge of the external world. As noted above, this is not my con-
cern here. There are many different versions or forms of skepticism,
the varieties of which needn’t occupy us at present. For a thorough
and penetrating discussion, cf. Pritchard (2016). I regret my inability
to consider Coliva’s challenging book here.
15 In addition to “faulty theorizing”, Pritchard also convicts his skep-
tically-minded or anti-Wittgensteinian opposition of engaging in or
being misled by “dubious philosophical theory” (p.3), “faulty philo-
sophical theory” (p.4, 68), “misguided philosophical theory” (p. 59),
“a faulty philosophical picture” (p. 67), and “defective philosophy”
(p.175). But is unclear why these are ‘faulty’, ‘dubious’, ‘misguided’,
or ‘defective’ other than that they are incompatible with his defense
of Wittgenstein, which is itself an integral part of his overall anti-
skeptical case, and in particular that they (mistakenly in Pritchard’s
view) reject the essential locality of rational evaluation, although, as
argued above, that thesis is itself problematic.
16 Pritchard, in his treatment of Barry Stroud’s thesis that “radi-
cal skeptical practices of epistemic appraisal” are a “purified ver-
sion” of our ordinary, “quotidian” epistemic practices, suggests that
Wittgenstein’s “account of the structure of reasons demonstrates”
that Stroud’s thesis cannot be right because the skeptic’s practices
“aspires to an impossible ideal” (2016, p. 149, emphasis added; cf.
p.69). But ‘demonstrates’ seems less appropriate than ‘presupposes’,
given the points just made in the text. Moreover, that the ideal is
impossible to realize, if it is, does not even suggest, let alone entail,
that the ideal is defective. Justice, for example, is a completely legiti-
mate ideal, one we ought to continue to strive to achieve even if it
can’t be perfectly realized. If the skeptic’s ideal of universal rational
evaluation is impossible to realize, that doesn’t show that it’s a flawed
ideal, but rather that our inability to realize it indicates a genuine lim-
itation on our ability to know.

Hinges, Disagreements, andArguments: (Rationally) Believing Hinge Propositions andArguing…
1 3
reason could we have for holding the über hinge com-
mitment? Whatever grounds we cited would already
presuppose the truth of this commitment after all.
Moreover, once we see that the other hinge commit-
ments we have—to personal hinge propositions and to
explicitly anti-skeptical hinge propositions—are sim-
ply a consequence of our über hinge commitment, then
it becomes clear that the extent to which our system
of rational support presupposes essentially groundless
commitments is quite considerable.
… since we are unable to have a rationally supported
belief in the über hinge proposition, it follows that we
are unable to have rationally supported beliefs in the
personal and anti-skeptical hinge propositions that
codify our über hinge commitment. We thus get the
Wittgensteinian conclusion: since all rational evalu-
ation necessarily takes place relative to groundless
hinge commitments, the very idea of a fully general
rational evaluation – that is, one that does not presup-
pose any hinge commitments – is incoherent, whether
that evaluation is positive (i.e., anti-skeptical) or nega-
tive (i.e., skeptical). The universality of rational evalu-
ation thesis is thus rejected. (ibid., pp.97–98)
This is a powerful articulation and defense of the Wittgen-
steinian ‘groundlessness’ doctrine, but it seems to me prob-
lematic because it constitutes just the kind of argument that
it urges is impossible:
1. Any possible reason for holding the über hinge commit-
ment presupposes the truth of that commitment.
2. Therefore, we can have no independent reason or ground
to hold the commitment.
3. Therefore, our commitment to the über hinge commit-
ment is groundless.
4. Therefore, all rational evaluation necessarily rests on the
groundless über hinge commitment (and perhaps other,
equally groundless personal and anti-skeptical hinge
commitments).
5. But, in contemplating the skeptical threat, we are engag-
ing in rational evaluation, which must necessarily rest on
groundless hinge commitments, and ultimately the über
hinge commitment.
6. Therefore, we should—indeed must, insofar as we want
to address skeptic-inspired epistemic angst—embrace
that commitment (and reject the universality of rational
evaluation thesis).
But this argument, if successful, seems itself to provide a
good reason for embracing the hinge commitment—in which
case the embrace of that commitment is not groundless after
all.17,18
To conclude this section: Wittgenstein’s epistemology,
and the arguments he offers in its support, seem to me want-
ing, for the reasons just rehearsed.19 But rather than continu-
ing to argue against it, I will next turn our attention to the
ramifications of his view for argumentation—and in particu-
lar, its ramifications for the prospects of arguing successfully
in cases of deep disagreement.
2 Hinge Epistemology andDeep
Disagreements
2.1 Can We Rationally Argue About Hinge
Propositions?
Can we argue about hinge propositions? As we’ve seen,
Wittgenstein says no: they are groundless, so positive argu-
ments for them are impossible; and they are beyond rational
challenge, so negative arguments against them are equally
impossible. If so, arguing about hinge propositions, ration-
ally or otherwise, is impossible. Is this right?
One problem for the Wittgensteinian view is that we do
in fact seem to argue about them. I’ve already pointed to
Bishop Berkeley, who, along with other idealists, famously
argued against the existence of physical objects, urging
instead a version of idealism according to which apparently
physical objects are really ‘mental’. I’m not myself a Berke-
leyan idealist, but it seems clearly mistaken to think that he
didn’t offer serious arguments against the existence of physi-
cal objects. Many others have joined Berkeley in defending
one version or another of idealism and arguing against that
particular hinge proposition. Indeed, teachers of courses in
the history of philosophy or metaphysics might well assign
the evaluation of Berkeley’s arguments as a paper topic or
17 Pritchard might object that the reason for embracing the über
hinge commitment (and the rest of the Wittgensteinian view) yielded
by this argument is pragmatic rather than epistemic: we should
embrace the view because if we don’t we cannot achieve our aims of
confronting the skeptical challenge and overcoming epistemic angst;
we should embrace it not because it is true but because it is the only
way to achieve our ends (cf. Pritchard 2005, pp.241 ff.). But since
embracing or “holding” a commitment is something we do (or not),
the argument seems clearly to provide a reason for doing that very
thing. For properly epistemic reasons to regard the Wittgensteinian
view as false, cf. sections A2a-g above.
18 Although I don’t think Pritchard’s case works, for the reasons
noted above, I am happy to acknowledge that his argument is excep-
tionally powerful, a terrific example of original philosophical theoriz-
ing at its best. In addition to the lengthy passage just cited, cf. his
summary of his case for the Wittgensteinian view, pp.173–175.
19 These and other complaints concerning Wittgenstein’s views are
further treated in Siegel (2008) and (2013).

H.Siegel
1 3
offer it as a question in an essay exam. Moreover, those argu-
ments can be and often are discussed and evaluated during
class meetings, and of course in scholarly writings as well.20
So there seems be a strong prima facie case for thinking that
that particular hinge proposition can be argued about and
subjected to critical scrutiny. The same can be said for other
hinge propositions, including, as we have seen, Pritchard’s
über hinge proposition “that one is not radically and funda-
mentally mistaken in one’s beliefs”—indeed, his (2016) is a
book-length treatment, chock full of arguments, concerning
that proposition itself, as is Descartes’ Meditations and the
forests’ worth of trees dedicated to the thousands of books
and papers concerning just such arguments. So, Wittgenstein
to the contrary notwithstanding, it does seem possible to
argue about hinge propositions.
If we can argue about hinge propositions, can such argu-
ments be rational? The Wittgensteinian tradition, as we
have seen, says ‘no’: they cannot be argued about; a fortiori
they cannot be rationally argued about. But unless we are
prepared to declare all the pro- and anti-idealist arguments
in the literature irrational—which we clearly should not be
prepared to do, since many of them are insightful, dialec-
tically appropriate, telling, and perfectly in keeping with
relevant disciplinary, logical and epistemological criteria
of argumentative quality and rigor—we must demur, and
conclude, Wittgenstein to the contrary notwithstanding, that
hinge propositions can indeed be subjected to critical evalu-
ation. That is, they can be (and are routinely) the subjects of
rational argumentation.
2.2 Can theBelief thatThere Are Physical Objects
Be Rational?: The Rationality of‘Hinge Beliefs’
By ‘hinge beliefs’ I mean simply beliefs concerning hinge
propositions. For example, ‘There are physical objects’ is
a hinge proposition; my belief that such objects exist is an
example of a hinge belief. Can such beliefs be rational? Here
we can be brief: if the preceding discussion is on track, hinge
beliefs can indeed be rational.
But here a new worry arises: are such ‘beliefs’ rightly
thought of as beliefs? I turn to this worry next.
2.3 Hinges, Beliefs, andCommitments
On Pritchard’s “nonbelief reading” of his account of Witt-
genstein’s epistemology, ‘hinge beliefs’ aren’t beliefs at all.
Rather, they are “visceral” commitments—since they are
“never the result of a rational process and are in their nature
unresponsive to rational considerations, then they are not
plausible candidates to be beliefs” (2016, p.90; cf. pp.84ff.;
ch. 4). And as we have seen, they are not eligible for rational
evaluation. Pritchard develops a highly sophisticated account
of the ways in which most of these commitments can change
with time and context, yet nevertheless ‘codify’ the über
hinge commitment, which cannot. Although hinge commit-
ments cannot, on this account, amount to knowledge-apt
beliefs, they can nevertheless constitute nearby-to-belief
propositional attitudes (2016, pp.90–103). As Pritchard
summarizes his nonbelief reading:
…our hinge commitments cannot be the result of any
rational process, but rather express prior visceral com-
mitments on our part, commitments that must be in
place in order to create the rational arena in which
rational evaluations function. This is why it is a mis-
take to even conceive of our hinge commitments as
being (knowledge-apt) beliefs, since a propositional
attitude that is in principle insensitive to rational con-
siderations cannot be a (knowledge-apt) belief. (2016,
p.174)
On this view, my ‘belief’ that the universe has existed for
more than 5minutes is not a belief at all. Neither is my
‘belief’ that my beliefs are not radically and fundamentally
mistaken (the über hinge commitment) a belief. I am ‘com-
mitted’ to these propositions, but my commitments to them
simply are not beliefs.
I respectfully disagree. I do believe that the universe has
existed for more than 5minutes, that there is an external
world, that my beliefs are not radically mistaken, that I’ll
soon be completing a draft of this paper, discussing it at
an upcoming conference and submitting it for publication
shortly thereafter, etc. These all—hinges and non-hinges
alike—seem clearly enough to be genuine beliefs, beliefs
that result from rational processes (or at least can be sus-
tained by such processes, since I can now consider whether
my hinge beliefs are worthy of sustained commitment), are
perfectly knowledge-apt, etc. The Wittgensteinian view is
not remotely plausible—it itself beggars belief. At least this
much is true: my belief that there is an external world is
indistinguishable, qua belief, from my belief that the sun
is now shining outside my office window as I type this
sentence.
Pritchard regards this indistinguishability as insufficient
to regard my hinge belief as a belief:
For sure, this commitment may feel like belief to the
person concerned, in that its phenomenology may be
identical to other, more mundane, beliefs that the sub-
ject holds. But the import of this point is moot once we
remember that the phenomenology of a propositional
attitude does not suffice to determine what proposi-
tional attitude is in play. The phenomenology of the
propositional attitude of wishful thinking may be, in
20 For a sampling of recent work, cf. Goldschmidt and Pearce (2017).

Hinges, Disagreements, andArguments: (Rationally) Believing Hinge Propositions andArguing…
1 3
certain cases, subjectively indistinguishable from the
phenomenology of the propositional attitude of believ-
ing, for example, but that does not make wishful think-
ing a kind of believing. (2016, p.102)
So consider the wishful thinker. She wishes that p be
true (say, that her horse will win the big race, perhaps
because she really needs the money), and, by engaging in
wishful thinking, comes to think that her horse will win.
Does she not believe that her horse will win? If not, it is
hard to see what wishful thinking is. That is, it is hard to
understand what wishful thinking is, if it doesn’t amount
to wishful believing. Arguing that some such ‘beliefs’—the
hinge ones—are ‘visceral’ and not candidates for knowl-
edge despite their phenomenological indistinguishability
from non-hinge beliefs, is insufficient to establish that they
are not beliefs after all. And it really does render it unclear
what propositional attitude is allegedly in play in the case
of hinge propositions—especially in light of the fact that
Wittgenstein himself regarded that propositional attitude as
that of belief: “At the foundation of well-founded belief lies
belief that is not founded” (1969, p.253, emphasis added);
“The difficulty is to realize the groundlessness of our believ-
ing” (1969, p.166, emphasis added; cf. pp.92, 243).21
2.4 Hinges andDeep Disagreements
So what shall we say of the role of hinge beliefs—or, if you
insist, hinge commitments—in cases of deep disagreement?
As many readers will know, argumentation theory’s con-
cern about deep disagreements, at least under that label,
stems from Robert Fogelin’s ‘The Logic of Deep Disagree-
ments’ (1985). I won’t rehearse Fogelin’s case here, but
rather simply cite his Wittgenstein-inspired conclusion:
Still, what shall we say about deep disagreements? We
can insist that not every deep disagreement is deep,
that even with deep disagreements, people can argue
well or badly. In the end, however, we should tell the
truth: there are disagreements, sometimes on impor-
tant issues, which by their nature, are not subject to
rational resolution. (Fogelin 1985, p.7)
Fogelin’s case for deep disagreements and their resistance
to rational resolution is explicitly based on Wittgenstein’s
discussion of hinges. And his conclusion, that disagreements
rooted in conflicting hinge propositions do not admit of
rational resolution, is manifestly in keeping with Wittgen-
stein’s view of hinges as not subject to rational evaluation.
So: does Wittgenstein’s view of hinges threaten the very pos-
sibility of the rational resolutions of disagreements involving
hinges? The short answer is: it does not. As argued above
and elsewhere, rational argumentation, even concerning
hinges, is perfectly possible; and a thoroughgoing fallibi-
lism, as applicable to hinges as to anything else, is the better
view (cf. Siegel 2013).
3 Conclusion
It should be obvious that nothing I’ve said here is aimed at
overcoming the skeptic or defeating the challenge posed by
radical skepticism. I hope that Pritchard’s ‘biscopic’ attempt
to rescue us from epistemic angst can succeed. But wor-
ries about skepticism, however central to epistemological
theorizing they might be, are of limited significance to argu-
mentation theory, because most argumentation does not (and
most arguments do not) involve skepticism. Reasons and evi-
dence aimed at non-skeptical theses, claims, and standpoints
can and sometimes do provide epistemic support for those
targets, and two central tasks of argumentation theory are
to explain how such support works and to provide criteria
for assessing its strength. If what I’ve argued above is right,
hinge propositions, like all others, are fully open to argu-
mentative support and challenge, and do not cause trouble
for the two epistemological tasks of argumentation theory
just mentioned. To the extent that argumentation theory’s
worry about deep disagreements rests on a Wittgensteinian
account of hinge beliefs, propositions, or commitments, such
disagreements similarly cause no trouble.22
Acknowledgements An earlier version was presented at the 9th ISSA
conference, Amsterdam, July 2018. I am grateful to the audience mem-
bers on that occasion for their critical reactions, and to John Biro, Al
Neiman, the editors of this special issue, two anonymous reviewers,
and especially Adam Carter for their challenging comments and helpful
suggestions on an earlier draft.
Compliance with Ethical Standards
Conflict of interest Harvey Siegel declares that he has no conflict of
interest.
22 A referee complains that Pritchard’s view precludes Fogelin-
style deep disagreements, in that Pritchard argues that his embrace
of Wittgenstein allows him to avoid what he calls ‘epistemological
incommensurability’, such that deep disagreements involving diver-
gent hinge propositions can be rationally resolved. (Cf. in particular
Pritchard 2009, 2011). But the complaint is misplaced. I do not claim
anything about Pritchard and deep disagreements, and am happy to
welcome him as an ally in the rejection of such disagreements.
21 This point, that it is difficult to grant Pritchard’s claim that ‘belief’
in hinge propositions does not amount to belief, is well articulated
in Coliva 2016. I should acknowledge that Wittgenstein is himself
ambivalent about the claim that the appropriate propositional attitude
to hinge propositions is belief: the passages just cited suggest that it
is belief, while others suggest, as Pritchard puts it, “that in the rel-
evant sense Wittgenstein does not think of our hinge commitments as
beliefs either” (2011, p.282).

H.Siegel
1 3
Research Involving Human Participants and/or Animals This article
does not contain any studies with human participants or animals per-
formed by any of the authors.
References
Carter JA (2017) Epistemic pluralism, epistemic relativism and “hinge”
epistemology. In: Coliva A, Pedersen NJLL (eds) Epistemic plu-
ralism. Series: Palgrave innovations in philosophy. Palgrave, Bas-
ingstoke, pp.229–249
Coliva A (2015) Extended rationality: a hinge epistemology. Palgrave
Macmillan, Basingstoke
Coliva A (2016) Review of Pritchard, Epistemic angst: radical skepti-
cism and the groundlessness of our believing. Notre Dame Philo-
sophical Reviews, June 29 2016
Coliva A, Pedersen NJLL (2017) Epistemic pluralism. Palgrave Mac-
millan, Basingstoke
Fogelin R (1985) The logic of deep disagreements. Informal Log
7(1):1–8. Reprinted as part of a special issue devoted to deep
disagreement. Informal Log 25(1) (2005)
Goldschmidt T, Pearce KL (2017) Idealism: new essays in metaphysics.
Oxford University Press, Oxford
Popper K (1970) Normal science and its dangers. In: Lakatos I, Mus-
grave A (eds) Criticism and the growth of knowledge. Cambridge
University Press, Cambridge, pp51–58
Pritchard D (2005) Epistemic luck. Oxford University Press, Oxford
Pritchard D (2009) Defusing epistemic relativism. Synthese
166(2):397–412
Pritchard D (2011) Epistemic relativism, epistemic incommensurabil-
ity, and Wittgensteinian epistemology. In: Hales SD (ed) A com-
panion to relativism. Blackwell, Malden, pp266–285
Pritchard D (2016) Epistemic angst: radical skepticism and the ground-
lessness of our believing. Princeton University Press, Princeton
Siegel H (2004) Relativism. In: Niiniluoto I, Sintonen M, Woleński J
(eds) Handbook of epistemology. Kluwer, Dordrecht, pp747–780
Siegel H (2008) Autonomy, critical thinking and the Wittgensteinian
legacy: reflections on Christopher Winch, Education, autonomy
and critical thinking. J Philos Educ 42(1):165–184
Siegel H (2013) Argumentation and the epistemology of disagreement.
Cogency 5(1):135–170
Wittgenstein L (1958) Philosophical investigations. Macmillan, New
Yor k
Wittgenstein L (1969) On certainty. Basil Blackwell, Oxford