Professional Documents
Culture Documents
Pluralisms in Truth and Logic, Jeremy Wyatt, Nikolaj J. L. L. Pedersen, Nathan Kellen (Ed.)
Pluralisms in Truth and Logic, Jeremy Wyatt, Nikolaj J. L. L. Pedersen, Nathan Kellen (Ed.)
and Logic
Edited by
Jeremy Wyatt
Nikolaj J. L. L. Pedersen
Nathan Kellen
Palgrave Innovations in Philosophy
Series Editors
Vincent Hendricks
University of Copenhagen
Copenhagen, Denmark
Duncan Pritchard
University of Edinburgh
Edinburgh, UK
Palgrave Innovations in Philosophy is a new series of monographs. Each
book in the series will constitute the ‘new wave’ of philosophy, both in
terms of its topic and the research profile of the author. The books will be
concerned with exciting new research topics of particular contemporary
interest, and will include topics at the intersection of Philosophy and
other research areas. They will be written by up-and-coming young phi-
losophers who have already established a strong research profile and who
are clearly going to be leading researchers of the future. Each monograph
in this series will provide an overview of the research area in question
while at the same time significantly advancing the debate on this topic
and giving the reader a sense of where this debate might be heading next.
The books in the series would be of interest to researchers and advanced
students within philosophy and its neighboring scientific environments.
Pluralisms in Truth
and Logic
Editors
Jeremy Wyatt Nikolaj J. L. L. Pedersen
Underwood International College Underwood International College
Yonsei University Yonsei University
Incheon, South Korea Incheon, South Korea
Nathan Kellen
University of Connecticut,
Storrs, CT, USA
This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Patrick Greenough, Sungil Han, Jinho Kang, Jiwon Kim, Junyeol Kim,
Seahwa Kim, Teresa Kouri Kissel, Kris McDaniel, Graham Priest, Agustín
Rayo, Greg Restall, Jisoo Seo, Stewart Shapiro, Gila Sher, Paul Simard
Smith, Erik Stei, Elena Tassoni, Pilar Terrés, Cory D. Wright, Crispin
Wright, Andy D. Yu, Luca Zanetti, and Elia Zardini. Special thanks go to
Filippo Ferrari, Michael P. Lynch, Sebastiano Moruzzi, and Joe Ulatowski.
Introduction 3
Nikolaj J. L. L. Pedersen, Jeremy Wyatt, and Nathan Kellen
vii
viii Contents
Logical Particularism277
Gillman Payette and Nicole Wyatt
Logical Nihilism301
Aaron J. Cotnoir
Index473
Notes on Contributors
xi
xii Notes on Contributors
2014) and The Metaphysics of Truth (OUP, 2018) and the editor of Truth: A
Contemporary Reader (Bloomsbury, under contract). His articles have appeared
in a number of journals, including the Journal of Philosophy, Australasian Journal
of Philosophy, and Pacific Philosophical Quarterly, amongst others.
Filippo Ferrari is Postdoctoral Research Fellow at the Institute of Philosophy,
University of Bonn. His research focuses primarily on two clusters of topics: the
normative aspects of enquiry and the debate about the nature of truth. He has
published his work in journals such as Mind, Synthese, Analysis, and Philosophical
Quarterly.
Rosanna Keefe is Professor of Philosophy and Head, Department of
Philosophy, University of Sheffield. Keefe specializes in philosophy of logic, phi-
losophy of language, and metaphysics. She is the author of Theories of Vagueness
(Cambridge, 2000) and numerous articles in journals such as Mind, Analysis,
Philosophical Studies, Australasian Journal of Philosophy, and Synthese.
Nathan Kellen works at the Department of Philosophy, University of
Connecticut. Kellen’s work is on truth, the philosophy of logic, philosophy of
mathematics, and ethics. Currently his main research project is an investigation
of truth pluralism and logical pluralism. He explores both of these views indi-
vidually but likewise examines how they might be connected.
Seahwa Kim is Professor of Philosophy and Dean, Scranton College, Ewha
Womans University. Kim specializes in metaphysics and the philosophy of
mathematics. Her articles have appeared in journals such as Australasian Journal
of Philosophy, Philosophical Studies, and Erkenntnis.
Teresa Kouri Kissel is Assistant Professor, Department of Philosophy, Old
Dominion University. Kouri Kissel specializes in philosophy of logic, philoso-
phy of mathematics, and mathematical and philosophical logic. Her dissertation
develops a new, neoCarnapian form of logical pluralism. Her articles have
appeared in Philosophia Mathematica, Erkenntnis, and Topoi.
Michael P. Lynch is Professor of Philosophy and Director, Humanities
Institute, University of Connecticut. Lynch’s work focuses on questions in meta-
physics, the philosophy of language, epistemology, and metaethics. He is author
of Truth in Context (MIT, 1998), True to Life (MIT, 2004), and Truth as One
and Many (OUP, 2009), as well as two books for popular audiences and a num-
ber of different articles in journals such as Philosophical Quarterly, Australasian
Journal of Philosophy, and Philosophical Studies.
Notes on Contributors xiii
a number of books, including The Taming of the True (OUP, 2002) and Changes
of Mind: An Essay on Rational Belief Revision (OUP, 2012). His articles have
appeared in many journals, including Mind, Philosophia Mathematica, Review of
Symbolic Logic, and Noûs.
Chase B. Wrenn is Professor of Philosophy, Department of Philosophy,
University of Alabama. Wrenn’s research focuses on truth, epistemology, and the
philosophy of mind and cognitive science. He is the author of Truth (Polity,
2014) and has had articles appear in journals including Australasian Journal of
Philosophy, Erkenntnis, Synthese, and The Philosophical Quarterly.
Jeremy Wyatt is Assistant Professor of Philosophy, Underwood International
College, Yonsei University. Wyatt’s main research interests are the philosophy of
language, metaphysics, and truth. His articles have appeared in Philosophical
Studies, Philosophical Quarterly, American Philosophical Quarterly, and Inquiry.
Andy D. Yu is JD student, University of Toronto. Yu completed a D.Phil. the-
sis (Fragmented Truth) at the University of Oxford. He works on philosophical
logic, the philosophy of language, metaphysics, and epistemology. He has pub-
lished in the Journal of Philosophy, Philosophical Quarterly, and Thought.
Elia Zardini is MidCareer FCT Fellow, LanCog Research Group, University of
Lisbon. Zardini specializes in logic and epistemology and has had articles appear
in many journals, including The Review of Symbolic Logic, Philosophical Studies,
Analysis, and Journal of Philosophical Logic. He is also the editor or coeditor of
Scepticism and Perceptual Justification (OUP, 2014), Substructural Approaches to
Paradox (special issue of Synthese, forthcoming), The Sorites Paradox (CUP, forth-
coming), and The A Priori: Its Significance, Grounds, and Extent (OUP,
forthcoming).
List of Figures
xv
Part I
Truth
Introduction
Nikolaj J. L. L. Pedersen, Jeremy Wyatt,
and Nathan Kellen
1 Pluralisms
The history of philosophy displays little consensus or convergence when
it comes to the nature of truth. Radically different views have been pro-
posed and developed. Some have taken truth to be correspondence with
reality, while others have taken it to be coherence with a maximally coher-
ent set of beliefs. Yet others have taken truth to be what it is useful to
believe, or what would be believed at the end of enquiry.1 While these
views differ very significantly in terms of their specific philosophical com-
mitments, they all share two fundamental assumptions: monism and sub-
stantivism. The views all assume that truth is to be accounted for in the
N. J. L. L. Pedersen (*)
Underwood International College, Yonsei University, Incheon, South Korea
e-mail: nikolaj@yonsei.ac.kr
J. Wyatt
Underwood International College, Yonsei University, Incheon, South Korea
N. Kellen
University of Connecticut, Storrs, CT, USA
same way across the full range of truth-apt discourse (monism) and that
truth is a substantive property or relation (substantivism).
The deflationist reaction to the traditional debate is to reject substantiv-
ism and, in some cases, to endorse monism. Truth, if it has a nature at all,
has a uniform nature across all truth-apt discourse, but there is not much
to say about it. The traditional debate went off-track exactly because
truth theorists thought that there was a whole lot to say about truth—
that somehow it had a deep or underlying nature that could be uncovered
through philosophical theorizing. Instead, according to many deflation-
ists, the (non-paradoxical) instances of the disquotational schema (“p” is
true if and only p) or the equivalence schema (it is true that p if and only
if p) exhaust what there is to say about truth.2
The pluralist reaction to the traditional debate is to reject monism and
endorse substantivism. Truth pluralists, encouraged by the seminal work
of Crispin Wright and Michael Lynch, appeal to more than one property
in their account of truth. Propositions from different domains of dis-
course are true in different ways. The truth of propositions concerning
the empirical world (e.g., 〈There are mountains〉) might be accounted
for in terms of correspondence, while the truth of legal propositions (e.g.,
〈Speeding is illegal〉) might be accounted for in terms of coherence with
the body of law.3 This amounts to a rejection of monism. By contrast,
truth pluralists have traditionally endorsed substantivism. They have
appealed to properties or relations that are substantive in nature (where
this means, at least, that they directly explain certain facts entirely in
virtue of characteristics pertaining to their natures).4
The history of logic, like the history of truth, displays little consensus.
Advocates of classical logic, intuitionistic logic, and relevant logic, for
instance, have argued back and forth about the merits and demerits of their
preferred systems. Again, as in the case of truth, this seems to suggest a shared
underlying assumption of monism: there is a uniquely correct logic, and
advocates of different systems are disagreeing about which one it is. The plu-
ralist reaction—notably advocated by Jc Beall and Greg Restall (2006)—is
to reject monism and maintain that there are several equally correct logics.
Speaking more generally, pluralist views are becoming increasingly
prominent in different areas of philosophy. Pluralism about truth has been
extensively developed, defended, and critically discussed. The same goes
Introduction 5
for pluralism about logic.5 Pluralism has also made inroads into ontology
where the idea that there are several ways of being has been defended, sup-
ported, and articulated in various ways. The work of Kris McDaniel is a
particularly rich source.6 In epistemology, a variety of pluralist theses can
likewise be found in the literature. The idea that there are several epistemi-
cally good-making features of belief can be found in different guises, as
pluralism about epistemic justification, warrant, desiderata, and value.
Prominent epistemologists such as Alvin Goldman, Tyler Burge, William
Alston, and Crispin Wright all endorse one of these forms of pluralism.7
These pluralist trends are philosophically significant. They go against a
one-size-fits-all conception of their relevant areas and invite a reconsid-
eration of the nature and character of some of the most fundamental
notions in core areas of philosophy—including truth, validity, being, and
justification. This volume takes as its focus two of kinds of pluralism:
pluralism about truth and pluralism about logic. It brings together 18
original, state-of-the-art essays. The essays are divided into three parts.
Part I is dedicated to truth pluralism, Part II to logical pluralism, and Part
III to the question as to what connections might exist between these two
kinds of pluralism.
Background
(mix-atom) π is beautiful.
The inference from 〈If Mt. Everest is extended in space, then Bob’s
drunk driving is illegal〉 and 〈Mr. Everest is extended in space〉 to 〈Bob’s
drunk driving is illegal〉 is a mixed inference. It is also a valid inference—
that is, necessarily, if the premises are true, then so is the conclusion. In
order to account for the validity of the inference, it would seem that there
must be some truth-relevant property that the premises and conclusion
all share which ensures that truth is preserved from premises to conclu-
sion. However, the pluralist seems to be unable to point to a property
that satisfies this constraint. For, as before, we can suppose that corre-
spondence to reality is the truth-relevant property for 〈Mr. Everest is
extended in space〉 and coherence with the body of law for 〈Bob’s drunk
driving is illegal〉. This means that one of the premises and the conclu-
sion have different truth-relevant properties. The problem of mixed infer-
ences challenges the pluralist to tell a story about the validity of mixed
inferences.10
Another fundamental problem confronting pluralists is what is some-
times called the “double-counting objection.” In essence, the objection is
that pluralists count two differences where only one is needed. They
endorse significant metaphysical differences regarding the nature of vari-
ous subject matters and, in addition, they endorse differences in the
nature of truth. However, in order to accommodate wide-ranging truth-
aptitude, differences need only be countenanced at one level—at the level
of the things themselves (numbers, trees, moral properties, laws, etc.) or
at the level of the content associated with different domains (expressivist
content vs. representational content). Drawing distinctions at the level of
truth, the objection goes, is superfluous.11
The Contributions
(SD) There are real differences in kind between the contents of our
beliefs and indicative statements.
Background
According to Beall and Restall, logic is plural in the sense that there are
several equally legitimate instances of what they call Generalized Tarski’s
Thesis (GTT):
Introduction 13
Beall and Restall argue that there are at least three equally legitimate ways
to construe casex in GTT: cases as (consistent and complete) possible
worlds, cases as (possibly incomplete) constructions, and cases as (possibly
inconsistent) situations. These three notions of case deliver different log-
ics—respectively classical logic, intuitionistic logic, and relevant logic.
To shed further light on the nature of Beall and Restall’s logical plural-
ism, let us highlight a number of key features of their view: legitimacy,
logical functionalism, logical generalism, logical relativism, meaning con-
stancy, and structural rules and properties.
legitimate only when restricted to certain domains or when used for certain
purposes. Rather, their view is that classical logic, intuitionistic logic, and
relevant logic are equally legitimate across the board. This idea can be
regarded as a form of logical generalism. The legitimacy of the logics admit-
ted by (GTT) and the three constraints is meant to be completely general.
Meaning Constancy Beall and Restall are pluralists about validity and
logical consequence. However, they maintain that the meaning of the
logical connectives is constant across the various logics that qualify as
legitimate. This view contrasts with a view often attributed to Carnap—
namely, that different logics have different connectives.17 According to
the latter view, the meaning of “and”, “not”, “or”, etc. changes from one
logic to another. By contrast, on Beall and Restall’s view, logical expres-
sions share the same—but incomplete—meaning across logics. However,
clauses that govern the connectives in different logics capture different
aspects of that shared meaning.
(Reflexivity) ϕ ⊨ ϕ
The Contributions
meaning across logics. She argues that this cannot be the case for intu-
itionistic negation and relevant negation. Thus, contrary to their own
claims, Beall and Restall’s logical pluralism ends up having a Carnapian
flavor.
Elia Zardini’s contribution “Generalised Tarski’s Thesis Hits
Substructure” offers an extended argument to the effect that Beall and
Restall’s logical pluralism is overly restrictive. He shows that (GTT)
implies that logical consequence has the structural features mentioned
above, that is, reflexivity, monotonicity, transitivity, contraction, and
commutativity. For instance, in every casex in which ϕ is true, ϕ is true—
thus (GTT) implies reflexivity. By the lights of Beall-Restall logical plu-
ralism, then, any system that fails to satisfy reflexivity cannot qualify as
logic properly so-called. Similarly, for any system that fails to satisfy
monotonicity, transitivity, contraction, or commutativity. In general,
Beall-Restall logical pluralism disqualifies any substructural logic from
qualifying as logic proper. Zardini argues that there are philosophical rea-
sons to doubt each of reflexivity, monotonicity, transitivity, contraction,
and commutativity. For instance, one might wish to abandon transitivity
in order to give an account of vagueness, and one might wish to abandon
contraction in order to address the semantic paradoxes. In light of these
considerations, Zardini deems Beall-Restall pluralism unacceptably
restrictive.
Gillman Payette and Nicole Wyatt’s contribution “Logical
Particularism” introduces and spells out a particularist view on logic.
According to logical particularism, there are no completely general
principles concerning validity. Rather, validity is a property of particu-
lar arguments or inferences. Since Payette and Wyatt think that differ-
ent, specific arguments call for different treatments vis-à-vis validity,
their logical particularism goes hand in hand with a particularist ver-
sion of logical pluralism. This particularist pluralism is at odds with the
kind of logical pluralism endorsed by Beall and Restall—which, as seen
above, involves a commitment to logical generalism. Logics on the
Beall-Restall view are all-purpose logics, which means that their rules
are completely general principles concerning validity. Payette and
Wyatt’s particularist pluralism is also at odds with domain-based forms
Introduction 17
Background
(TD) TD(ϕ) → ϕ
(TA) ϕ → TA(ϕ)
The combination of (TD) and (TA) looks superficially like the T-schema,
but it is not. The subscripts make all the difference in the world. They
indicate that the two conditionals involve different concepts—the con-
cepts that, on Scharp’s view, should replace the inconsistent, ordinary
concept of truth. Hence, the two conditionals cannot be combined into
a biconditional to yield the T-schema. Rather, according to Scharp, truth
theorists should rest content with (TD) and (TA) and the concepts of
descending and ascending truth that they respectively concern. It is in
this sense that Scharp puts forward a pluralistic account of truth—
whereas we might have thought that we could get along with a single
truth concept, Scharp’s contention is that we actually need a pair of
replacement concepts to adequately resolve the semantic paradoxes.
Insofar as Scharp’s views are motivated by the semantic paradoxes, they
represent yet another path into pluralist territory.27
The Contributions
she observes that the view clashes with the widely held view that logic is
topic neutral. Second, like others, she observes that (mod)—a natural candi-
date for addressing the issue of mixed inferences—threatens to undermine
domain-based logical pluralism. Third, building on the two first points,
Keefe offers the following collapse argument: mixed inferences can cut across
all domains. By (mod), the logic of this inference would be the intersection
of all logics that hold for some domain. However, since this logic is topic
neutral, this is what logic proper is—and hence, logical monism is correct.
Fourth, Keefe questions an implicit assumption behind domain-based logi-
cal pluralism, viz. that domain membership suffices to fix the logic of infer-
ences. She points to vagueness as a phenomenon that tells against this
assumption, as this phenomenon that cuts across domains and has tradition-
ally been thought to be of relevance to logic. Lastly, Keefe presents a positive
proposal: validity is assessed relative to context. This allows for logical plural-
ism because the rules or assumptions justified in different contexts may vali-
date different arguments. Keefe observes that (1) her context-based
framework can accommodate domain-based logical pluralism in the sense
that different context-domain pairs may validate different inferences, and (2)
the context-based approach allows for logical pluralism with respect to a
single domain since different contexts may validate different inferences per-
taining to the same domain. Furthermore, Keefe suggests that the context-
based approach is also better suited to deal with vagueness.
Kevin Scharp’s contribution “Aletheic and Logical Pluralism” explores
what he calls coordinated pluralism and compares it to his own replace-
ment theory. Coordinated pluralism is the combination of context-based
truth pluralism and context-based logical pluralism. Coordinated plural-
ism is motivated by considerations related to the semantic paradoxes. As
observed earlier, if truth is characterized by the unrestricted T-schema
and classical logic or certain other logics hold, you end up with paradox.
The double-barreled pluralism that coordinated pluralism offers is meant
to block paradox by letting the nature of truth and the nature of logic
vary across contexts in such a way that there is no context in which truth
and logic have natures that generate paradox. This is meant to happen in
virtue of their natures being coordinated in any given context. If, in a
given context, truth is strong, logic is weak and vice versa. This coordi-
nated pluralism seems like a promising approach to paradox. Might it be
Introduction 25
Notes
1. David 1994; Devitt 1984; Newman 2007; Rasmussen 2014; Russell
1912; Vision 2004; and Wittgenstein 1921 all endorse versions of the
correspondence theory. Coherence theorists include Blanshard 1939;
Joachim 1906; and Young 2001. (Neo-)pragmatists include James 1907,
1909; Peirce 1878; and Putnam 1981.
2. Deflationists of various stripes include Field 1986, 1994a, b; Grover
1992; Horwich 1998; Quine 1970; Ramsey 1927; Strawson 1950. For a
systematic treatment of deflationary metaphysics of truth, see Wyatt
2016.
3. We use angle brackets to represent propositions.
4. See Wright 1992, 2001; Lynch 2001, 2004b, 2009.
5. Works on pluralism about truth include Asay 2018; Beall 2000, 2013;
Cook 2011; Cotnoir 2009, 2013a, b; Edwards 2008, 2009, 2011,
2012a, b, 2013a, b, 2018; Engel 2013; David 2013; Dodd 2013; Kölbel
2008, 2013; Lynch 2000, 2001, 2004a, b, 2005a, b, 2006, 2008, 2009,
2013; Newhard 2013, 2017; Pedersen 2006, 2010, 2012a, b, 2014;
Pedersen and Edwards 2011; Pedersen and Wright 2010, 2012, 2013a,
b; Shapiro 2011; Sher 1998, 2005, 2013, 2016; Stewart-Wallace 2016;
Tappolet 1997, 2000, 2010; Williamson 1994; (C.D.) Wright 2005,
2010, 2012; Wright 1992, 1996a, b, 1998, 2001, 2013; Wyatt 2013;
Wyatt and Lynch 2016; and Yu 2017a, b. Works on logical pluralism
include Beall 2014; Beall and Restall 2000, 2001, 2006; Bueno and
Shalkowski 2009; Carnielli and Coniglio 2016; Ciprotti and Moretti
2009; Cook 2010, 2014; Eklund 2012; Field 2009; Goddu 2002;
Hjortland 2013; Humberstone 2009; Keefe 2014; Kouri 2016; Kouri
26 N. J. L. L. Pedersen et al.
and Shapiro forthcoming; Payette and Wyatt 2018; Priest 2001, 2008,
2014; Read 2006a, b; Restall 2001, 2002, 2012, 2014; Russell 2008,
2014; Shapiro 2014; Sher forthcoming; Terrés forthcoming; van
Benthem 2008; Varzi 2002; and Wyatt 2004.
6. MacDaniel’s work on ontological pluralism spans about a decade. His
book The Fragmentation of Being (2017) brings together much of his
earlier work on the topic. See also. Turner 2010, 2012; Eklund 2009.
7. Alston 2005; Burge 2003; Goldman 1988; Wright 2004. Pedersen 2017
formulates and defends pluralism about fundamental or non-derivative
epistemic goods.
8. This slogan is the title of Lynch 2009. Having advocated moderate plu-
ralism for nearly two decades in a wide range of works, Lynch is the most
prominent advocate of the view. The distinction between these two
forms of pluralism is due to Pedersen 2006.
9. We will use small caps to denote concepts.
10. Sher 2005 raises the problem of mixed atomics. For the problems of
mixed compounds and inferences, see respectively Tappolet 2000 and
Tappolet 1997.
11. Versions of the double-counting objection have been presented by Asay
2018; Blackburn 1998, 2013; Dodd 2013, Horwich 1996; Pettit 1996;
Quine 1960; and Sainsbury 1996.
12. Wright 1998. See also Wright 1992 (Chap. 1, II) and Lynch 2004b: 386.
13. Beall and Restall 2006.
14. Field 2009.
15. In attributing a logical relativist view to Beall and Restall, we follow
Shapiro 2014.
16. It may seem odd to attribute both logical generalism and logical relativ-
ism to Beall and Restall. For, isn’t the idea that logic is general in tension
with the idea that it is relative? No. Logical generalism, as we have char-
acterized it, concerns whether a given logic legitimately issues (in)valid-
ity verdicts across the board (in the technical sense of legitimacy tied to
Generalized Tarski’s Thesis and the constraints of necessity, formality,
and normativity). For logics that are general in this sense, there is a fur-
ther issue as to the semantic status of their (in)validity verdicts—in par-
ticular, whether those verdicts have absolute or relativized truth-values.
17. Carnap 1937, 1950.
18. For a comprehensive introduction to substructural logics, see Restall
2000.
19. Lynch 2009, Pedersen 2014. The version of logical pluralism presented
in Shapiro 2014 can also be regarded as a kind of domain-based logical
Introduction 27
pluralism, although the various domains that Shapiro considers are all
mathematical.
20. The realism/anti-realism debate is, of course, one of the major themes in
Dummett’s corpus of work. See e.g., the essays in Dummett 1978.
21. See Lynch 2009 for details. The set of intuitionistic validities is a proper
subset of the set of classical validities, and hence, (mod) delivers the ver-
dict that the logic of any mixed compound or inference is intuitionistic
logic. In cases where there is overlap between two logics but no subset
relation, the intersection is identical to neither of the domain-specific
logics. This applies in the case of intuitionistic logic and relevant logic.
22. Smith forthcoming supports logical pluralism via logical contextualism.
While Beall and Restall never themselves present their logical pluralism
as a form of logical contextualism, Caret 2017 offers a contextualist read-
ing of their view and argues that it can be used to block a certain funda-
mental objection.
23. Barnard and Ulatowski 2013, Ulatowski 2017. See Kölbel 2008, 2013
for another empirically-based form of truth pluralism.
24. For details concerning MacFarlane’s assessment-sensitive relativism, see
MacFarlane 2014. Wyatt and Lynch 2016 suggest that MacFarlane is
committed to the kind of truth pluralism described.
25. Scharp 2013.
26. We’re blurring the line between use and mention here, but we trust that
the idea is clear.
27. See Wyatt and Lynch 2016 for further discussion of the relations between
truth pluralism and Scharp’s replacement theory. See also Beall 2013 and
Cotnoir 2013b for paradox-based motivations for truth pluralism.
References
Alston, W. 2005. Beyond ‘Justification’: Dimensions of Epistemic Evaluation.
Ithaca: Cornell University Press.
Asay, J. 2018. Putting Pluralism in Its Place. Philosophy and Phenomenological
Research 96 (1): 175–191.
Barnard, R., and J. Ulatowski. 2013. Truth, Correspondence, and Gender.
Review of Philosophy and Psychology 4: 621–638.
Beall, J.C. 2000. On Mixed Inferences and Pluralism About Truth Predicates.
The Philosophical Quarterly 50: 380–382.
———. 2013. Deflated Truth Pluralism. In Pedersen & Wright 2013a,
323–338. New York: Oxford University Press.
28 N. J. L. L. Pedersen et al.
———. 2013b. Pluralism and Paradox. In Pedersen & Wright 2013a, 339–350.
New York: Oxford University Press.
David, M. 1994. Correspondence and Disquotation: An Essay on the Nature of
Truth. Oxford: Oxford University Press.
———. 2005. On Truth Is Good. Philosophical Books 46: 292–301.
———. 2013. Lynch’s Functionalist Theory of Truth. In Pedersen & Wright
2013a, 42–68. New York: Oxford University Press.
Devitt, M. 1984. Realism and Truth. Oxford: Blackwell.
Dodd, J. 2013. Deflationism Trumps Pluralism! In Pedersen & Wright 2013a,
298–322. New York: Oxford University Press.
Dummett, M. 1978. Truth and Other Enigmas. Cambridge, MA: Harvard
University Press.
Edwards, D. 2008. How to Solve the Problem of Mixed Conjunctions. Analysis
68: 143–149.
———. 2009. Truth Conditions and the Nature of Truth: Resolving Mixed
Conjunctions. Analysis 69: 684–688.
———. 2011. Simplifying Alethic Pluralism. The Southern Journal of Philosophy
49: 28–48.
———. 2012a. On Alethic Disjunctivism. Dialectica 66: 200–214.
———. 2012b. Alethic vs. Deflationary Functionalism. International Journal of
Philosophical Studies 20: 115–124.
———. 2013a. Truth, Winning, and Simple Determination Pluralism. In
Pedersen & Wright 2013a, 113–122. New York: Oxford University Press.
———. 2013b. Truth as a Substantive Property. Australasian Journal of
Philosophy 91: 279–294.
———. 2018. The Metaphysics of Truth. Oxford: Oxford University Press.
Eklund, M. 2009. Carnap and Ontological Pluralism. In Metametaphysics, ed.
D. Chalmers, D. Manley, and R. Wasserman, 130–156. Oxford: Oxford
University Press.
———. 2012. Multitude, Tolerance and Language-Transcendence. Synthese
187: 833–847.
Engel, P. 2013. Alethic Functionalism and the Norm of Belief. In Pedersen &
Wright 2013a, 69–86. New York: Oxford University Press.
———. 1986. The Deflationary Conception of Truth. In Fact, Science and
Morality, ed. G. MacDonald and C. Wright, 55–117. Oxford: Basil Blackwell.
———. 1994a. Deflationist Views of Meaning and Content. Mind 103:
249–285.
———. 1994b. Disquotational Truth and Factually Defective Discourse.
Philosophical Review 103: 405–452.
30 N. J. L. L. Pedersen et al.
Vision, G. 2004. Veritas: The Correspondence Theory and Its Critics. Cambridge,
MA: MIT Press.
Williamson, T. 1994. Critical Study of Truth and Objectivity. International
Journal of Philo-sophical Studies 30: 130–144.
Wittgenstein, L. 1921. Tractatus Logico-Philosophicus. In Annalen der
Naturphilosophie. English trans. D.F. Pears and B.F. McGuinness, 1961.
London: Routledge & Kegan Paul.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University Press.
———. 1996a. Précis of Truth and Objectivity. Philosophy and Phenomenological
Research 56: 863–868.
———. 1996b. Response to Commentators. Philosophy and Phenomenological
Research 56: 911–941.
———. 1998. Truth: A Traditional Debate Reviewed. Canadian Journal of
Philosophy 24 (suppl. vol.): 31–74.
———. 2001. Minimalism, Deflationism, Pragmatism, Pluralism. In The
Nature of Truth, ed. M.P. Lynch, 751–787. Cambridge, MA: MIT Press.
———. 2004. Warrant for Nothing (and Foundations for Free)? Proceedings of
the Aristotelian Society 78: 167–212.
Wright, C.D. 2005. On the Functionalization of Pluralist Approaches to Truth.
Synthese 145: 1–28.
———. 2010. Truth, Ramification, and the Pluralist’s Revenge. Australasian
Journal of Philosophy 88: 265–283.
———. 2012. Is Pluralism About Truth Inherently Unstable? Philosophical
Studies 159: 89–105.
Wright, C. 2013. A Plurality of Pluralisms? In Pedersen & Wright 2013a,
123–153. New York: Oxford University Press.
Wyatt, N. 2004. What Are Beall and Restall Pluralists About? Australasian
Journal of Philosophy 82: 409–420.
Wyatt, J. 2013. Domains, Plural Truth, and Mixed Atomic Propositions.
Philosophical Studies 166: 225–236.
———. 2016. The Many (Yet Few) Faces of Deflationism. The Philosophical
Quarterly 66: 362–382.
Wyatt, J., and M.P. Lynch. 2016. From One to Many: Recent Work on Truth.
American Philosophical Quarterly 53: 323–340.
Young, J.O. 2001. A Defence of the Coherence Theory of Truth. The Journal of
Philosophical Research 26: 89–101.
Yu, A. 2017a. Logic for Pluralists. Journal of Philosophy 114: 277–302.
———. 2017b. Can Alethic Pluralists Maintain Compositionality? The
Philosophical Quarterly 67: 625–632.
Truth: One or Many or Both?
Dorit Bar-On and Keith Simmons
As regards (1), one might worry that the pluralist will inherit all the
main problems of the various traditional substantivist theories of truth,
not just their advantages. In addition, however, we think that pluralism
faces distinctive difficulties, having to do, specifically, with supposing
truth to be many. In what follows, we’ll be focusing on moderate plural-
ism—a view that acknowledges, in addition to diverse truth properties, a
single property of truth (so that truth emerges as both one and many).
After presenting a recent, influential version of moderate pluralism (Sect.
2), we will articulate the main difficulties we see with this view (Sect. 3).
We believe that the difficulties we raise extend to other pluralist views,
but we do not address other pluralisms, except in passing.
Now, given the difficulties, it is natural to wonder whether there is a
metaphysically more conservative way to accommodate the core intu-
itions that originally motivated pluralism. In Sect. 4, we explore one such
way. To anticipate, we will suggest, first, that the difficulties raised in Sect.
3 speak strongly in favor of thinking of truth as one, but not also many.
At the same time, we agree with Wright and others that a certain plurality
must be recognized, if we are to allow for substantive metaphysical
debates between realists and anti-realists in various domains. However,
Truth: One or Many or Both? 37
2 Moderate Pluralism
Here we focus on moderate pluralism, the kind of pluralism that takes
truth to be both one and many. This is in contrast to strong pluralism,
where truth is many, but not one as well. It’s fair to say that most plural-
ists take moderate pluralism to be preferable, because it has ready answers
to certain objections that have been raised against pluralism. Here are
two:
(1) There appears to be some unity to the ways in which propositions can
be true—for example, it seems that any truth property would supply
a criterion for the correctness of beliefs, and it seems that any truth
property would be preserved by valid inference. So, there should be a
generic truth property to capture this unity.
38 D. Bar-On and K. Simmons
(2) Consider the conjunction <Speeding is illegal and there are moun-
tains>.2 This is an example of mixed discourse: the two conjuncts are
drawn from two different domains of discourse. The first conjunct is
from the legal domain, and we may suppose that it does not have the
correspondence property. Rather, it is superwarranted: it is warranted
at the present stage of inquiry, and would remain warranted without
defeat at every successive stage of inquiry. The second conjunct is an
atomic sentence from the geologic domain, and we may suppose that
it does have the correspondence property—it is true because it cor-
responds to geologic reality. But what about the conjunction as a
whole? It is true. However, we cannot say that it has the correspon-
dence property, because its first conjunct doesn’t. Perhaps we can say
that the conjunction as a whole is superwarranted—we might allow
that the second conjunct has the correspondence property and is
superwarranted. But attributing superwarrant to the conjunction
doesn’t adequately explain why it is true, because it doesn’t adequately
explain the truth of the second conjunct. Here, the moderate plural-
ist has an answer: the truth of the conjunction is explained in terms
of generic truth.
Once we have truth as one and many, we need some account of how
the many relate to the one. Moderate pluralists offer a variety of accounts
of this relation—for example, one account takes the relation to be real-
ization, and another account takes it to be manifestation.3 Either way, we
start with certain core principles which characterize the concept of truth.
According to Michael Lynch, for example, there are at least three core
principles, or “core truisms”:
Objectivity: The belief that p is true if, and only if, with respect to the
belief that p, things are as they are believed to be. (Lynch
2009, p. 8).
Norm of Belief: It is prima facie correct to believe that p if, and only if,
the proposition that p is true. (op. cit., p. 10).
End of Inquiry: Other things being equal, true beliefs are a worthy goal
of inquiry. (op. cit., p. 12).
Truth: One or Many or Both? 39
Most pluralists are monists about the concept of truth, and these core tru-
isms are taken to pin down this unique concept. Then, at the level of
properties, we have, according to the moderate pluralist, the generic
property of truth, call it truth-as-such, and the various properties which
relate to it—perhaps by realizing this generic property, perhaps by mani-
festing it.4
To focus the discussion, we consider Lynch’s manifestation pluralism—
though we think that what we have to say carries over to other versions of
pluralism. According to manifestation pluralism, truth-as-such is the
property that satisfies the core principles as a matter of conceptual necessity.
Now consider the manifestation relation: What is it for, say, superwarrant
to manifest truth-as-such? It is for superwarrant to have among its fea-
tures those that truth-as-such has essentially. So, suppose that a truth
about traffic laws has the superwarrant property. For the superwarrant
property to manifest truth-as-such is for superwarrant to satisfy the core
truisms, since truth-as-such satisfies these principles essentially. In legal
discourse, say, superwarrant satisfies the core truisms, and so here super-
warrant manifests truth-as-such.
Clearly, truth-as-such manifests itself—the features that truth-as-such
has essentially are among the features that truth-as-such has. So, for
example, the proposition that speeding is illegal has two truth-manifesting
properties: truth-as-such and superwarrant. Following Lynch, call a
proposition that has two truth-manifesting properties an unplain truth,
and a proposition that is only true-as-such a plain truth.
To see manifestation pluralism in action, consider the mixed conjunc-
tion <Speeding is illegal and there are mountains>. This conjunction as a
whole does not possess superwarrant or correspondence. Rather it is
plainly true: it is true-as-such, and has no other truth property. The plain
truth of the whole conjunction supervenes on the unplain truth of its
conjuncts. Each separate conjunct is what Lynch calls strongly grounded—
the truth-as-such of each conjunct is manifested by a further truth prop-
erty. <Speeding is illegal> is true-as-such because the proposition is
superwarranted; <there are mountains> is true-as-such because the prop-
osition corresponds to the facts. In contrast, the truth-as-such of the
whole conjunction is not manifested by any further truth property of the
conjunction itself. The conjunction itself is a plain truth that is weakly
grounded in the unplain truth of its components.
40 D. Bar-On and K. Simmons
there will also be a necessary connection between F and the truth of <<p>
is true>. That is, any world which contains F is a world in which <<p> is
true> is true, and any world which does not contain F is a world in which
<<p> is true> is false. It seems natural to take this as saying that there’s a
correspondence between <<p> is true> and F. But then it follows that
<<p> is true> is not plainly true—like <p>, it is both true-as-such and has
the correspondence property.
At this point, the manifestation pluralist might appeal to the depen-
dence asymmetry between <p> and <<p> is true>: <<p> is true> depends
for its truth on the truth of <p>, but not vice versa. The truth attribution
<<p> is true> is weakly grounded; but <p> itself is strongly grounded.
Now this might support the thought that the correspondence between
<<p> is true> and F is dependent on the correspondence between <p>
and F. <<p> is true> corresponds to F only because (i) <p> and <<p> is
true> are necessarily equivalent, and (ii) <p> corresponds to F. But even
if this is so, the correspondence between <<p> is true> and F remains.
Even if this correspondence is in some way derivative, still <<p> is true>
has the correspondence property—it is not plainly true. Again, we can
take the inheritance view as reflecting the derivative way in which <<p>
is true> has the correspondence property: it inherits the property from its
attributee <p>. And again we can state the obvious: If <<p> is true>
inherits the correspondence property, it has the property.
The manifestation pluralist might claim that there’s a relevant differ-
ence between <<p> is true> and <p>. <<grass is green> is true> has a
proposition as its subject; <grass is green> has grass as its subject. Does this
make a difference to the correspondence relation? This depends on one’s
view of the equivalence between <<p> is true> and <p>. We might hold,
with Frege and many others, that the necessary equivalence between
<<p> is true> and <p> is an indication of the special, transparent way in
which the truth predicate works when applied to individual propositions.
The predicate ‘true’ is not an ordinary property-ascribing predicate. If we
accept that the truth predicate is transparent in this way, it is very hard to
see how <p> and <<p> is true> could differ in whatever truth properties
they do have—in particular, if one corresponds to F, so does the other.
If, on the other hand, one thought of ‘true’ as an ordinary property-
ascribing predicate, then, as a correspondence theorist, one should think
44 D. Bar-On and K. Simmons
(1) A truth attribution shares its truth properties, whatever they may be,
with its attributee.
One further remark. The manifestation pluralist might take a leaf out
of the deflationist’s book, and regard a generalization such as (G) as
equivalent to an infinite conjunction of conditionals: If George says that
there are mountains then there are mountains, and if George says that
grass is green then grass is green, and … . Similarly, a generalization such
as ‘Something George said yesterday is true’ would be regarded as an
infinite disjunction. Then the problem of dealing with generalized truth
attributions reduces to the problem of dealing with truth-functional
compounds. The manifestation pluralist will say, as we saw above, that
truth-functional compounds are plainly true, while their atomic compo-
nents are unplainly true. This is so whether the compounds are mixed (as
with <Speeding is illegal and there are mountains>) or unmixed. Lynch
writes: “Compound propositions, mixed or not, are true because they are
plainly true”.8 The plain truth of a conjunction is a matter quite indepen-
dent of the discourses from which the conjuncts are drawn. But now
problems emerge for truth-functional compounds, in parallel with the
problems for truth attributions.
Suppose that at some stage of inquiry, both <p> and <q> are separately
warranted, and would remain warranted without defeat at every succes-
sive stage of inquiry. Then it seems hard to deny that <p&q> is warranted
at the given stage and at every successive stage: that is, if <p> and <q> are
separately superwarranted, then <p&q> is superwarranted. And if we
accept a sense in which the truth of <p&q> depends on the truth of <p>
and <q> separately, then this asymmetry is accommodated naturally by
the inheritance view, which preserves the superwarrant of <p&q>.
Similarly with supercoherence. If adding each of <p> and <q> separately
to F makes F more coherent, and will continue to do so through all sub-
sequent improvements, then so will adding <p&q> to F. Suppose <p>
corresponds to fact F, and <q> to fact F*. Then the worlds which contain
both facts F and F* will be exactly those worlds in which <p&q> is true—
we have a correspondence between <p&q> and the facts. In each case, we
have an unmixed conjunction which is superwarranted, or supercoherent
or corresponds—they are not plainly true.
The manifestation pluralist’s claim is that any true conjunction is
plainly true. The plain truth of a conjunction is a matter supposedly inde-
pendent of the mixed or unmixed character of the conjunction. But now
46 D. Bar-On and K. Simmons
we have seen that there are conjunctions that are not plainly true. So, the
manifestation pluralist’s argument for the plain truth of conjunctions has
broken down—we’ve now lost whatever motivation there was for suppos-
ing that, say, <Speeding is illegal and there are mountains> is plainly true.
And, as we saw earlier, the conjunction cannot possess the correspon-
dence property, given the first conjunct, and even if we allow that the
conjunction is superwarranted, this fails to explain the truth of the sec-
ond conjunct. The conjunction is true, but the manifestation pluralist
has no adequate account of its truth.
The point here connects to the case of truth attributions. Suppose
George only ever says two things: “Speeding is illegal” and “There are
mountains”. Now consider again:
4 Wherein Plurality?
Difficulties with moderate pluralism, some independently motivated
commitments, and an attempt to accommodate the main motivations for
pluralism, lead us to propose the following set of desiderata for a view of
truth:
Under the envisaged scenario, there remains only one sort of truth: that
which is defined by the platitudes-satisfying role. It is just that what truth
involves in one area – what realizes the appropriate role – may be different
from what it involves in another. The difference … will be explained by
reference to the different subject-matters: the different truth-conditions, and
the different truth-makers, in each discourse.15
Now, to gain traction against the truth pluralist, who insists on locat-
ing the relevant plurality in kinds of truth as opposed to kinds of meaning,
it is important to note that the notion of truth-conditions (as well as that
of subject-matter, and even content) is invoked in discussions of truth in
two ways that can—and, we submit, should—be separated.
A Davidsonian truth-conditional theory of meaning aims to yield as
theorems meaning-specifying biconditionals, such as
(W) “Wasser ist nass” is true if, and only if, water is wet.
(W′) “Wasser is nass” is true if, and only if, H2O is wet.
Our aim in what follows will be to take a stab at clarifying the distinc-
tion we have in mind, and to explain how this distinction bears on what
we take to be the best way to accommodate the motivations behind
pluralism.19
Begin with truth-conditions as they figure in meaning-giving bicondi-
tionals. Truth-conditions thus understood feature in Davidson’s seminal
“Truth and Meaning”, where he proposes that a theory of meaning for a
natural language should take the form of a theory of truth for that lan-
guage.20 Meaning-giving biconditionals are designed to capture the
logical place occupied by individual sentences in the whole (potentially
infinite) network of sentences of a language. Meaning-giving bicondi-
tionals are relatively neutral, metaphysically speaking. This relative neu-
trality is well-captured by Davidson himself, when he says:
Truth: One or Many or Both? 51
Our question is: What are we here adding to the claim that we have
something to do with things being illegal, but nothing to contribute with
respect to whether something is a mountain? What difference is there in
what ‘accounts for the circumstance’ that the two different propositions
are true that is not simply a matter of the difference in the constitution of
the relevant facts? Of course, we can ascend to the ‘formal mode’, and
instead of talking about what makes for the existence of mountains ask
what accounts for the truth of <There are mountains>. If we want to gen-
eralize over the whole domain, we may need to use the truth predicate,
viz. “For all p, if p is a mountain-statement, then p is true iff …”. And,
depending on how the condition is filled in, we may be able to say, for
example, that mountain statements are true in a mind-independent way.
But our point is that the possibility of characterizing the differences in
the formal mode in no way betrays commitment to a new, additional dif-
ference—one that requires postulating differences in ways of being true,
or the possession of divergent truth properties by statements in different
areas of discourse. Indeed, when Pedersen and Lynch expound the
‘semantic difference’ between the truth of <Speeding is illegal> and
<There are mountains>, they themselves immediately resort to talk in the
material mode: “Mountains are mind-independent entities while laws are
social – and so, mind-dependent – constructs”.29
Unlike others who have worried about double counting, our objection
is not motivated by a pluralist view of propositional content or a defla-
tionist view of truth.30 If we are right, there is a way of making sense of
Truth: One or Many or Both? 57
disputes between realists and anti-realists that neither goes via a dis-
tinction at the level of propositions nor depends on deflating all truth.31
But it does not depend on invoking different kinds of truth, either. We
thus endorse alethic monism: there is only one way for true sentences,
propositions, beliefs, and so on, to be true. However, there may be
multiple kinds of worldly conditions that make them true. The rele-
vant plurality can be captured in the material mode; it doesn’t require
any semantic or alethic ascent. Of course, given the equivalence of <p>
and <<p> is true>, one can advert to a formal mode and speak of the
truth of ‘x is red’ being a different sort of thing from the truth of ‘x is
divisible by 2’—indeed, sometimes putting things in terms of truth
may be unavoidable. It is the additional move, to a plurality of truth
properties, each appropriate to a different domain of discourse, that we
here oppose. This move, we maintain, is not forced on us by taking
seriously debates between realists and anti-realists. Alethic plurality
contributes no explanatory power; all we need is a plurality of kinds of
worldly conditions.32
* * *
that make a statement true will vary. But then it stands to reason that it
will not in general be obvious what the ‘realizers’ of truth in various areas
are, in contrast with what constitutes winning a game. Access to what in
the world makes true statements true requires a metaphysical investiga-
tion. When it comes to games, since they are invented, and in that sense
‘of our own making’, knowing what game is being played guarantees
knowing what constitutes winning it. (That is part of what is instituted
when the game is designed.) Not so for what renders statements true in a
given domain of discourse. In general, engaging in an area of discourse
does not bring into existence the relevant worldly conditions. And, what-
ever disagreements we may have cannot be settled—as they can in the
winning case—simply through reflection on the rules of the discourse.
Where disputes arise, they are metaphysical ones.34
Notes
1. Wright (2013: 124)
2. Here and throughout, ‘<p>’ is a name of the proposition that p.
3. See, for example, Lynch (2013).
4. See, for example, op. cit.
5. Lynch 2013.
6. The equivalence here is stronger than necessary equivalence. As
Jeremy Wyatt has pointed out to us, it’s arguable that A and B can be
necessarily equivalent without both being superwarranted for a subject
S. Consider the necessarily equivalent propositions <There’s water in the
glass> and <There’s H2O in the glass>. Suppose S grasps all the relevant
concepts. It is possible that <There’s water in the glass> is superwarranted
for S, but <There’s H2O in the glass> is not, since S may have warrant to
believe that there’s water in the glass, but not that there’s H2O in the glass.
But if S grasps the concept of truth, <p> and <<p> is true> will be
conceptually equivalent for S, and if one is superwarranted for S, so is
the other.
7. Lynch offers this definition of propositional coherence (with respect to
moral propositions) in Lynch 2009, p.171.
8. Lynch 2013.
9. Bar-On and Simmons (2007), Bar-On et al. (2004).
10. Bar-On and Simmons (2007).
Truth: One or Many or Both? 59
32. Asay (2016) also argues that all the plurality we need is to be found in
the world, not in a plurality of truth properties. But Asay’s plurality is a
plurality of truthmakers rather than truth conditions. And Asay is a
primitivist about the concept of truth and a deflationist about the prop-
erty; we don’t make these commitments.
33. See Wright (2013: VII), who follows Edwards (2011, 2013).
34. Our thanks to Jeremy Wyatt for many helpful comments, and to the
participants in the Conference on Pluralism about Logic and Truth,
University of Connecticut at Storrs, April 2015.
References
Asay, J. 2016. Putting Pluralism in Its Place. Philosophy and Phenomenological
Research 96 (1): 175–191.
Bar-On, D. 1996. Anti-Realism and Speaker Knowledge. Synthese 106:
139–166.
Bar-On, D., and K. Simmons. 2007. The Use of Force Against Deflationism:
Assertion and Truth. In Truth and Speech Acts, ed. D. Greimann and
G. Siegwart, 61–90. New York/London: Routledge.
Bar-On, D., W. Lycan, and C. Horisk 2004. Deflationism, Meaning and Truth-
Conditions, reprinted with Postscript in Deflationary Truth, ed.
B.P. Armour-Garb and J.C Beall, 321–352. Chicago: Open Court Readings
in Philosophy.
Davidson, D. 1980. Essays on Actions and Events. Oxford: Oxford University
Press.
———. 1984. Inquiries into Truth & Interpretation. Oxford: Oxford University
Press.
———. 1990. The Structure and Content of Truth. Journal of Philosophy 87:
279–328.
———. 1996. The Folly of Trying to Define Truth. Journal of Philosophy 93:
263–278.
Dodd, J. 2013. Deflationism Trumps Pluralism. In Truth Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C. Wright, 298–322. Oxford:
Oxford University Press.
Edwards, D. 2011. Simplifying Alethic Pluralism. Southern Journal of Philosophy
49: 28–48.
———. 2013. Truth, Winning, and Simple Determination Pluralism. In Truth
Pluralism: Current Debates, ed. Nikolaj J.L.L. Pedersen and C. Wright,
113–122. Oxford: Oxford University Press.
Truth: One or Many or Both? 61
Horwich, P. 2006. A World Without Isms. In Truth and Realism, ed. P. Greenough
and M.P. Lynch, 188–202. Oxford: Oxford University Press.
Lynch, M. 2013. Three Questions for Truth Pluralism. In Truth Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C. Wright, 21–41. Oxford:
Oxford University Press.
Pedersen, Nikolaj J.L.L., and M. Lynch. 2018. Truth Pluralism. In The Oxford
Handbook of Truth, ed. M. Glanzberg. Oxford: Oxford University Press.
Pettit, P. 1996. Realism and Truth: A Comment on Crispin Wright’s Truth and
Objectivity. Philosophy and Phenomenological Research 56: 883–890.
Strawson, P. 1992. Analysis and Metaphysics. Oxford: Oxford University Press.
Wright, C. 1996. Précis to Truth and Objectivity and Response to Commentators.
Philosophy and Phenomenological Research 56: 863–868, 911–941.
———. 2003. Saving the Differences. Cambridge, MA: Harvard University
Press.
———. 2013. A Plurality of Pluralisms? In Truth Pluralism: Current Debates,
ed. Nikolaj J.L.L. Pedersen and C. Wright, 123–153. Oxford: Oxford
University Press.
Truth Pluralism, Quasi-Realism,
and the Problem of Double-Counting
Michael P. Lynch
1 Unity and Diversity
Philosophical tradition has long decreed that truth comes in only one
kind. To talk about different kinds of truth is like saying that the Sphinx
has a different kind of eye—it is just another way of saying that it has
none at all.1 But philosophical tradition also decrees that traditions are
there to be opposed. Truth pluralism—or the idea, roughly, that there are
different kinds of truth—stands with the opposition in this case.
Since its introduction onto the contemporary scene by Crispin Wright
over two decades ago, truth pluralism has been connected to, and partly
motivated by, two major explanatory projects. One project involves
accommodating the intuitions that drive both realism and anti-realism
(Wright 1992). By adopting truth pluralism, the suggestion goes, we may
say that in some domains, statements or beliefs are true in a realist way,
while in other domains, truth is understood in ways traditionally cham-
pioned by “anti-realists.” Thus, for example, some statements are true
M. P. Lynch (*)
Department of Philosophy, University of Connecticut, Storrs, CT, USA
e-mail: mplynch@uconn.edu
semantic diversity: there are real differences in kind between the con-
tents of our beliefs and indicative statements.
cognitive unity: all such statements and beliefs are subject to a single
type of cognitive normative assessment or correctness.
…they strike at the level of the proposition: they mark distinctions of sub-
ject matter, and perhaps eventually distinctions of objectivity or the possi-
bility of cognitively fault-free disagreement. But why add to a distinction
of content, another, mirroring distinction, one only applying to kinds of
truth or conceptions of truth?” (Blackburn 2013, p. 265)
to disentangle them. As we’ll see, only one version of the point is compat-
ible with Blackburn’s own quasi-realism; but it faces its own problems—
problems that encourage the reintroduction of the very distinctions it
was aimed against. As such, I conclude that truth pluralism need not fear
being audited for double-counting—or at least no more than the quasi-
realist. Indeed, the quasi-realist, I will argue, has excellent motivations for
adopting—or readopting, as we shall see—truth pluralism.
2 Subject-Matter First
The first way of construing the double-counting objection can be found
in the Blackburn quote given just above. It is the idea that semantic diver-
sity can be reduced, or solely explained as, diversity in subject-matter.
Quine first made this point:
There are philosophers who stoutly maintain that “true” said of logical or
mathematical laws and “true” said of weather predictions or suspect confes-
sions are two usages of an ambiguous term “true”…Why not [instead] view
“true” as unambiguous but very general, and recognize the difference
between true logical laws and true confessions as a difference merely
between logical laws and confessions? (Quine 1960, p. 131)9
The point is that Ramsey and Wittgenstein do not need to work with a
sorted notion of truth—robust, upright, hard truth versus some soft and
effeminate imitation. They need to work with a sorted notion of a proposi-
tion, or if we prefer it a sorted notion of truth-aptitude. (Blackburn 1998b,
pp. 166–167)
Like Quine before him, Blackburn takes issue with the truth pluralist’s
view that appealing to different kinds of truth will help explain the differ-
ences. And, like Quine, he is motivated to take this view at least in part
because of his allegiance to a version of deflationism about truth. In par-
ticular, Blackburn is attracted to Horwich’s minimalism.
According to Horwich (1998), the concept of truth functions as a logi-
cal device for generalization: it allows us to overcome our merely medical
limitations and make certain generalizations, such as “every proposition
is either true or false.” We grasp that concept by being a priori disposed
to grasp instances of the equivalence schema:
The problem is not with a subjective source of value in itself, but with peo-
ple’s inability to come to terms with it, and their consequent need for a pic-
ture in which values imprint themselves on a pure passive, receptive witness…
Truth Pluralism, Quasi-Realism, and the Problem… 71
This passage (and the detailed discussion of the idea by Blackburn that
follows) is striking. The idea is precisely that of the truth pluralist and
motivated in a similar way: namely, by pointing out that those who
embrace a “subjective source” for value can respond to the obvious objec-
tions by demonstrating that moral judgments are apt for a distinctive,
but nonetheless real, kind of truth.
To sum up so far: The double-counting objection was initially framed
as the simpler alternative to truth pluralism. The thought seemed to be:
well, we can accomplish all of our theoretical goals—explaining both
semantic diversity and cognitive unity—just by appealing to something
we already have for “free”: different kinds of propositions. We don’t need
(“in addition”) kinds of truth. And that all goes swimmingly, save for the
fact that the most obvious way to press the objection—Quine’s way—
self-consciously fails (or ignores) that very explanatory task. The Quinean
route ends up either getting unity at the price of diversity (by enforcing a
blanket realism), or diversity at the price of unity (by enforcing a distinc-
tion between the truth-conditional and the non-truth-conditional).
Consequently, we arrive at the quasi-realist’s actual suggestion: truth
explains unity, kinds of propositions explain diversity. But this slogan is
already seeming less simple, and indeed, less distinct, than it did at first
blush. It is starting to sound like truth pluralism in disguise.
There are propositions properly theorized about in one way, and ones prop-
erly theorized about in another. The focus of theory is the nature of the
commitment voiced by one adhering to the proposition, and the different
functional roles in peoples’ lives (or forms of life, or language games) that
these different commitments occupy. (Blackburn 1998b, p. 167)
believe or assert that p (as opposed to merely wondering whether p), then
you become liable to certain pushbacks from your interlocutors if it turns
out that not-p. Moreover, you incur certain obligations to provide them
with evidence that p should they demand it, to stand by certain easily
recognized implications of p, and so on. In short, what distinguishes one
sort of commitment/attitude/speech-act from another is the type of sta-
tus and responsibility accorded to the person who engages in that com-
mitment/attitude/speech-act. They are not distinguished in terms of
truth, reference, or direction of fit (whether word-to-world or
world-to-word).
Price, in particular, has urged that a view of this sort is the most intui-
tive extension of the quasi-realist view (see, e.g., 2011, 2013). And he has
been explicit that it is designed to capture what I’ve called semantic diver-
sity and cognitive unity. Moreover, he has urged that the view be adopted
globally—that we become global expressivists, as he has put it. Of course,
in order to take the view to help us with our present problem of distin-
guishing kinds of propositions without appeal to different kinds of truth,
we’d need to be able to show at least two things.
First, as Brandom (1998) in particular has tried to do, we would have
to show that we could build a real theory of content out of this theory of
the types of attitudes/commitments/speech-acts. Thus, for example, such
a theory must have the resources to distinguish between contents that
concern our own informational states and attitudes, and those that are
wholly independent of them. Whether it can do so remains an open
question, so let’s put this aside.
Second, the Pricean pluralist must be able to distinguish ethical and
non-ethical propositions while still maintaining the idea that, for exam-
ple, ethical propositions can still be believed and asserted, albeit in a differ-
ent way than propositions about the physical world are believed and
asserted. It is, after all, the appeal to this “different way” of believing and
asserting that makes the commitment-first approach different from the
truth-conditional approach. The point here, to clarify, is not that the
global, Pricean quasi-realist must simply distinguish between different
attitudes (like belief and desire) that are operative in different domains of
discourse; this is something that all players can grant. Rather, they must
distinguish between different kinds of one type of attitude—for example,
belief (Dreier 2004; Ridge 2006).
74 M. P. Lynch
Represent: A belief with the content that x is F is true if and only if the
object represented by the concept x has the property repre-
sented by the concept F.
Thus, one might think, if there are different types of representation, then
we get not only different kinds of belief but also different kinds of truth,
namely:
Represent-I: An i-belief with the content that x is F is true if and only
if the object i-represented by the concept x has the prop-
erty i-represented by the concept F.
Represent-E: An e-belief with the content that x is F is true if and only
if the object e-represented by the concept x has the prop-
erty e-represented by the concept F.
How might the Pricean quasi-realist resist this charge? One way would be
to just accept that Represent-I and Represent-E result in two distinct
concepts of truth—concepts which “true” is simply ambiguous between.
Alternatively, they can insist on their minimalism about truth. Truth is
one thing, they can say, and representation is another. And all there is to
say about truth as such is still simple, and requires no appeal to the dif-
ferent kinds of representation.
But this point also won’t distinguish the Pricean quasi-realist in any sig-
nificant way from the truth pluralist. As noted earlier, truth pluralism, in its
principal formulations, isn’t an ambiguity view. It can indeed be described
as the view that there are different kinds of truth—in the same sense in
which the functionalist in the philosophy of mind can say that there are
76 M. P. Lynch
different kinds of pain. In Wright’s view and in others’, there is more than
one property that can determine, realize, or manifest the property of truth
as such. And it is difficult to see how Represent-I and Represent-E aren’t
just saying exactly that—namely, what determines the truth of i-beliefs is
i-representation and what determines the truth of e-beliefs is e-representa-
tion. But even if this is somehow resisted, the fact remains: while we started
with the thought that there was a strong contrast between truth pluralists
and quasi-realists, we have ended with the thought we might as well call
representational pluralism. The bump in the carpet has been moved.
Unsurprisingly, perhaps, Price himself suggests that the real difference
between truth pluralism and representation pluralism lies in how propo-
nents of each view tend to conceive of their project. Truth pluralists like
Wright typically see themselves as revealing the relational properties that
determine whether a belief or statement’s content is true. They are doing
a bit of metaphysics in service of their semantics. It is this that Price
wishes most to resist. For Price, representation-talk of either kind is just
that: talk. It serves a purpose, but it drops out at what he calls the “upper
level” (2013, 155ff). In saying that, for example, some assertions or atti-
tudes are more e-representational than others, we are ultimately not mak-
ing a distinction between different kinds of relational properties. Instead,
we are making claims about the “talk not the ontology” (2013, 158); we
are merely describing the roles such assertions play in our greater theoriz-
ing. This is the point of his insisting that his view represents a global
expressivism or quasi-realism.
Ultimately, then, the Pricean quasi-realist explains the differences
between kinds of thought and talk only internally to our thought and talk
themselves. Interestingly, Blackburn remains skeptical; and his skepti-
cism is worth quoting in full:
can, as suggested above, simply stick with the idea that truth is a func-
tional property. To define truth functionally in the sense I am interested
in is to define it by way of its connections to other related concepts.
These connections are embodied in certain common truisms that have
played a central role in the historical discussions over truth: the equiva-
lence schema; the slogan we called “Represent”; the idea that what makes
a proposition correct to believe is that it is true; that what valid infer-
ences preserve is truth; that truth is a worthy goal of inquiry; and the
like. The idea is that these truisms, or ones very much like them, jointly
pick out the truth-role. Beliefs are true just when they are correct, when
they are the sort of beliefs we aim to have during inquiry, when the
world is as they portray it as being. That is the job description of truth,
as it were.
It seems to me that this is just the sort of view that both Blackburn’s
quasi-realist and Price’s expressivist need. It allows us to say, against tradi-
tional theorists, that truth as such may be a very thin, functional prop-
erty. But seeing truth as a functional property in this sense is entirely
consistent with there being other properties, distinct from truth, which
satisfy this description and so realize that thin property—which, in short,
play the truth-role. And that is a helpful thought, because it opens to
door to appealing to these other properties to explain what grounds the
norm in different domains or discourses. It allows us new tools for
addressing what plays the truth-role for different kinds of content and,
therefore, giving us an easy explanation for what makes those kinds dif-
ferent kinds in the first place. They are different kinds because they are
subject to being correct in different ways.
Of course, the ultimate suitability of this suggestion rides in part on
whether we can make sense of a property—distinct from correspon-
dence—that could play the truth-role for domains like the ethical. One
possibility, as already noted, is to say that truth is sometimes realized by
an extremely deflated property. Another is to follow Blackburn’s earlier
lead: to look at the “materials at hand,” and to see whether we can con-
struct such a property from properties already in play in the moral
realm—properties, for example, like coherence. This, in fact, was pre-
cisely the property that Blackburn put to such effective use in his earlier
(1984) work.
Truth Pluralism, Quasi-Realism, and the Problem… 79
The basic recipe might be sketched as follows. Step one: look to what
warrants moral beliefs. According to a leading theory of moral epistemol-
ogy inherited from Rawls, a moral belief is warranted to the degree to
which the moral framework to which it belongs is in a state of wide
reflective equilibrium. We can, if we like, see this as a form of coherence
theory: S’s moral judgment that p is warranted to the degree that it
coheres with the rest of S’s moral and non-moral judgments. The term
“coherence” can be taken here to pick out a family of epistemic desider-
ata. It is generally thought that a framework is coherent insofar as, and to
the degree to which, its members display relations of mutually explana-
tory support, it is complete, and it is consistent. Call these coherence-
making features. Such features themselves come in degrees: members of a
framework can be more or less consistent, more or less mutually explana-
tory, and so on. A framework of judgments increases in coherence to the
degree to which it exemplifies these features, on balance, to a greater
degree. “On balance” because the features are not themselves isolated in
their coherence increasing power. A framework would not be more coher-
ent on balance, for example, simply by increasing its size (completeness)
by including consistent but explanatorily unconnected judgments.
Intuitively, by increasing its explanatorily isolated judgments, the coher-
ence of the framework would on balance remain static or decrease.
Step two: use these definitions to make sense of what it would be for a
moral framework to improve in coherence. Framework F is more coher-
ent at t2 than at t1 when, on balance, it has at t2 either more of the
coherence-making features or some of those features to a greater degree.
So, if completeness, consistency, and explanatory connectedness are
coherence-making features, adding a consistent and explanatorily con-
nected judgment to the system will increase that system’s coherence along
those dimensions. Consequently, we can say that P coheres with moral
framework F if, and only if, including P in F would, on balance, make F
more coherent.
Step three is to use this notion of improvement to build a property that
could be said to realize truth (as opposed to warrant). This is not a trivial
task. But neither is it hopeless. One suggestion, building on work by
Wright and others, would be to say that if warranted moral judgments are
those that are coherent, then true moral judgments are those that are, as
we might say, supercoherent:
80 M. P. Lynch
Notes
1. The analogy is Nietzsche (1968, p. 540).
2. Recent examples include (Bar-On and Chrisman 2009; Blackburn
1998a; Ridge 2006, 2009; Schroeder 2008, 2010).
3. See (Lynch 2013a).
4. See for example, (Wright 1998b) and (Lynch 2009, 2013b); in contrast,
see (Kölbel 2008) (Cotnoir 2009).
5. The functionalist way of putting the point was first expressed by Pettit
(1996) and Lynch (1998); Wright (2013) now interprets his view in a
similar way.
6. Expressions of the pluralist position (see Wright 1992; Lynch 2001) have
sometimes obscured this point by presenting the position as relativizing
or indexing truth realizing properties to domains. As David (2013) has
correctly noted, these flourishes are unessential to the truth pluralist’s
point. See also Lynch 2013b.
7. A recent sampling includes: Cotnoir 2009, 2013a, b; Edwards 2008,
2018; Horton and Poston 2012; Jarvis 2012; Pedersen 2006, 2012a, b;
C. Wright 2013; and C. D. Wright 2005, 2010, 2012.
8. One might also see Jamin Asay as making a similar point (Asay 2018).
9. Quine, interestingly, does not say who these philosophers were. Clearly,
though, truth pluralism was, as it were, in the air at this time.
10. Here I refer only to traditional correspondence theorists (see, e.g., Russell
1966 and Fumerton 2002). Such a view should be distinguished from a
position—which is intended to be a form of truth pluralism—that says that
correspondence itself comes in different kinds (see, e.g., Sher 2004, 2005).
11. Of course Quine himself didn’t think there are moral or modal proper-
ties, and Blackburn would agree. These are not the sorts of things we
expect to quantify over in our final, most rigorous theory of the world.
12. Wright makes a similar point in his 1998 response to Blackburn (Wright
1998a).
13. Many people have contributed to my thoughts on these matters over the
years, including notably, Nikolaj J. L. L. Pedersen, Jeremy Wyatt, and
Nathan Kellen—the editors of this volume—and Crispin Wright, Simon
Blackburn, Huw Price, Mark Timmons, Mike Ridge, Mark Chrisman,
Dorit-Bar On, and Cory Wright.
While working on this paper, I benefitted from participation in the
Pluralisms Global Research Network (National Research Foundation of
Korea grant no. 2013S1A2A2035514). This support is also gratefully
acknowledged.
82 M. P. Lynch
References
Asay, J. 2018. Putting Pluralism in Its Place. Philosophy and Phenomenological
Research 96 (1): 175–191. https://doi.org/10.1111/phpr.12303.
Bar-On, D., and M. Chrisman. 2009. Ethical Neo-expressivism. In Oxford
Studies in Metaethics, ed. R. Shafer-Landau, vol. 4, 132–165. Oxford: Oxford
University Press.
Beall, Jc. 2013. Deflated Truth Pluralism. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 323–338. New York:
Oxford University Press.
Blackburn, S. 1984. Spreading the Word. Oxford: Oxford University Press.
———. 1998a. Ruling Passions. Oxford: Oxford University Press.
———. 1998b. Wittgenstein, Wright, Rorty and Minimalism. Mind 107
(425): 157–181.
———. 2013. Deflationism, Pluralism, Expressivism, Pragmatism. In Truth
and Pluralism: Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright,
263–277. New York: Oxford University Press.
Brandom, R. 1998. Making It Explicit: Reasoning, Representing, and Discursive
Commitment. Cambridge, MA: Harvard University Press.
Chrisman, M. 2008. Expressivism, Inferentialism, and Saving the Debate.
Philosophy and Phenomenological Research 77 (2): 334–358.
Cotnoir, A. 2009. Generic Truth and Mixed Conjunctions: Some Alternatives.
Analysis 69 (3): 473–479.
Cotnoir, A.J. 2013a. Validity for Strong Pluralists. Philosophy and
Phenomenological Research 86 (3): 563–579.
———. 2013b. Pluralism & Paradox. In Truth and Pluralism: Current Debates,
ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 339–350. New York: Oxford
University Press.
David, M. 2013. Lynch’s Functionalist Theory of Truth. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 42–68.
New York: Oxford University Press.
Dodd, J. 2013. Deflationism Trumps Pluralism. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 298–322. New York:
Oxford University Press.
Dreier, J. 2004. Meta-Ethics and The Problem of Creeping Minimalism.
Philosophical Perspectives 18 (1): 23–44.
Dworkin, R. 1996. Objectivity and Truth: You’d Better Believe It. Philosophy &
Public Affairs 25 (2): 87–139.
Truth Pluralism, Quasi-Realism, and the Problem… 83
———. 2009. The Truth in Ecumenical Expressivism. In Reasons for Action, ed.
D. Sobel and S. Wall, 219–242. New York: Cambridge University Press.
Russell, B. 1966. On the Nature of Truth and Falsehood. In Philosophical Essays,
ed. B. Russell, 147–159. London: George Allen & Unwin.
Sainsbury, R.M. 1996. Review: Crispin Wright: Truth and Objectivity.
Philosophy and Phenomenological Research 56 (4): 899–904.
Schroeder, M. 2008. Being For. Oxford: Oxford University Press.
———. 2010. Noncognitivism in Ethics. London: Routledge.
Sher, G. 2004. In Search of a Substantive Theory of Truth. Journal of Philosophy
101 (1): 5–36.
———. 2005. Functional Pluralism. Philosophical Books 46 (4): 311–330.
Tappolet, C. 1997. Mixed Inferences: A Problem for Pluralism About Truth
Predicates. Analysis 57 (3): 209–210. https://doi.org/10.1111/1467-8284.
00077.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University
Press.
———. 1998a. Comrades Against Quietism: Reply to Simon Blackburn on
Truth and Objectivity. Mind 107 (425): 183–203.
———. 1998b. Truth: A Traditional Debate Reviewed. Canadian Journal of
Philosophy 28 (Supplement): 31–74.
———. 2013. Plurality of Pluralisms. In Truth and Pluralism: Current Debates,
ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 123–156. New York: Oxford
University Press.
Wright, C.D. 2005. On the Functionalization of Pluralist Approaches to Truth.
Synthese 145: 1–28.
———. 2009. Truth, Ramsification, and the Pluralist’s Revenge. Australasian
Journal of Philosophy 88 (2): 265–283.
———. 2012. Is Pluralism about Truth Inherently Unstable? Philosophical
Studies 159 (1): 89–105.
The Metaphysics of Domains
Douglas Edwards
1 Introduction
Pluralist theories of various sorts commit themselves to the notion of a
domain. Most prominently, truth pluralists suggest that there are differ-
ent stories to tell about how sentences1 get to be true depending on the
domain to which the sentence belongs.2 Indeed, some commitment to
the idea is ingrained in the very motivations for the view. The truth plu-
ralist’s dissatisfaction with truth monism is that one theory of truth does
not seem to fit all the different kinds of thought and talk that we ordinar-
ily take to be truth-apt, so we need to open ourselves up to the idea that,
rather than having one global theory of truth, we have a number of dif-
ferent theories of truth which apply locally, for different kinds of thought
and talk.
There is also a practical reason why truth pluralists commit themselves
to domains. Suppose for a moment that we hold that there are many
distinct truth properties, and we have no division of sentences into differ-
ent domains. On this picture, all sentences are capable of possessing all
D. Edwards (*)
Department of Philosophy, Utica College, Utica, NY, USA
the different truth properties. Now, the different truth properties are dis-
tinct properties, so it is possible for a sentence to possess one of the truth
properties but lack another. If we take falsity to be lack of truth, then we
have a sentence which both has a truth property and lacks a truth prop-
erty, which gives us a sentence that is both true and false. This compli-
cates matters and makes life difficult for pluralist proposals that do not
try to sort sentences into separate domains. Given that, as noted above,
truth pluralists have independent reason to hold that sentences are sorted
into domains, it seems natural to adopt them in response to this sort of
problem.
However, some pluralists are sceptical of the notion of a domain, and
Lynch (2013a) ends up rejecting domains, choosing instead to start with
specific sentences (or propositions, for Lynch), as opposed to sentences of
a particular kind. This, to my mind, is a mistake. As just noted, a key
component of pluralist views is that different kinds of subject matters are
apt for different sorts of treatment. If this idea is given up, then a central
motivation for the view is lost. As a result, I think that pluralists should
take the notion of a domain seriously as a central aspect of the view.
In this chapter, I give an account of what domains are and respond to
some problems of domain individuation. I begin by discussing the
semantic and metaphysical aspects of a domain, and then show how the
two aspects combine to yield an account of domains in general. Following
this, I will discuss two problems of domain individuation: the problem of
mixed atomics, and the problem of mixed compounds.
To say that we can analyse moral concepts, like the concept of being right,
is to say that we can specify which property the property of being right is
by reference to platitudes about rightness: that is, by reference to descrip-
tions of the inferential and judgemental dispositions of those who have
mastery of the term ‘rightness’. (Smith 1994: 39)
1989: 13, 21–9; 1990: 4–5), and the like. What these platitudes about
substance force us to admit, at the very least, is that there are limits on the
kind of content a set of requirements can have if they are to be moral
requirements at all, as opposed to requirements of some other kind (Dreier
1990). (Smith 1994: 40)
Option (i) holds that there are properties ‘out there’, so to speak, and
this can explain why we have predicates of a certain kind: they are respon-
sive, in some way, to existent properties. Option (ii) holds that there are
90 D. Edwards
properties that the relevant predicates pick out, but these properties exist
because they are projections of the predicates, as opposed to being already
existing things that our predicates respond to. Option (iii) is the error-
theoretical option, where, perhaps due to the specification of the predi-
cate kind, or alternatively because the world is uncooperative, there are
no properties that correspond to the predicates of a certain kind. An error
theorist about morality, for example, holds that there are no properties
that correspond to our moral predicates.5
Our focus here will be on (i) and (ii). We can illustrate the differences
between them using the following principle about properties:
(P) The object referred to by ‘a’ falls under the predicate ‘is F’ iff the
object referred to by ‘a’ has the property referred to by ‘is F’.
think that all metallic objects share some significant feature in common.
Despite the fact that both properties are classes, there are significant
metaphysical differences between them, leading to the former’s status as
an ‘abundant’ property, and the latter’s status as a ‘sparse’ property. In
particular, sparse properties ground genuine similarities between their
bearers, and have a causal-explanatory role, whereas abundant properties
do not ground genuine similarities between their bearers, and do not
have a causal-explanatory role.
The two ways of looking at properties broadly correspond to the two
ways we considered the predicate-property relation above, with sparse
properties being those mentioned in option (i) and abundant properties
being those mentioned in option (ii). To relate this to our discussion of
predicates, I am going to use the terminology of ‘responsive’ predicates
and ‘generative’ predicates. Broadly speaking, responsive predicates
respond to sparse properties, and generative predicates generate abun-
dant properties.14
The basic idea for our purposes here is that we can secure reference
to a mathematical object—the number 5, say—by noticing that its
associated singular term appears in at least one true sentence. Thus,
the truth of the sentence ‘the number 5 is prime’ is sufficient to secure
the existence of a referent for the singular term ‘the number 5’: the
number 5.
We can use this idea to construct the notion of an abundant object,
which pairs with the notion of an abundant property. If we hold that an
object a is abundant, then it cannot be that the singular term ‘a’ refers to
a because a exists, as we are explicitly denying that a has existence prior to
‘a’ referring to a. As a consequence, we need some grounds to establish in
virtue of what it is that ‘a’ refers to a. Hale and Wright offer an account
here in terms of truth: ‘a’ refers to a if ‘a’ appears in a true sentence.
Singular terms can thus be responsive or generative: responsive singular
terms respond to sparse objects, and generative singular terms generate
abundant objects.
94 D. Edwards
5 Summarizing Domains
I have suggested that a domain has both a semantic and a metaphysical
component, which are closely related. The semantic component of a
domain is understood in terms of singular terms and predicates. The
domain to which a singular term or predicate belongs is determined by
the kind of singular term or predicate that the relevant singular term or
predicate is identified to be. This is done by identifying the functional
role of a singular term or predicate kind to demarcate different kinds of
singular terms or predicates. The semantic aspect of a domain will thus be
understood as a singular term and predicate kind. For example, the
semantic aspect of the moral domain will be the singular terms and predi-
cates identified as being moral singular terms and predicates in virtue of
their playing the functional role associated with moral singular terms and
predicates.
The metaphysical aspect of a domain will be composed of the objects
and properties that the singular terms and predicates in the semantic
aspect of the domain refer to. Thus, the metaphysical aspect of the
moral domain, for example, will be the objects and properties referred
to by moral singular terms and predicates. This is not to say that all
there is to objects and properties of different kinds is that they are
referred to by different kinds of singular terms and predicates. This is
because, as we have seen, the relationships between singular terms and
objects, and predicates and properties, varies: in some cases, the nature
of the objects and properties is dependent on the singular terms and
predicates (the abundant model), whereas in other cases, the singular
terms and predicates are dependent on the objects and properties (the
sparse model).
This account of domains is intended to be available to theorists of vari-
ous sorts, not just truth pluralists. In the remainder of the chapter, I will
focus on a couple of problems of domain individuation that have been
posed for truth pluralism, though I think the problems remain general
problems for those who wish to individuate different kinds of subject
matter.
The Metaphysics of Domains 95
A statement using [thick] terms can be analyzed into something like “this
act has such-and-such a character, and acts of that character one ought not
to do.” It is essential to this account that the specific or “thick” character of
these terms is given in the descriptive element. The value part is expressed,
under analysis, by the all-purpose prescriptive term ought. (Williams 1985:
144)
some sort of resistance in the face of danger, for example. This may sug-
gest to some that there is some difference in kind between thick predi-
cates, like courageousness, and thin predicates, such as goodness, as one
imposes descriptive content whereas the other does not. However, this is
a mistake. Even thin properties like being morally good impose some
descriptive content on their bearers. For example, if an action is deemed
to be morally good, then this at the very least implies that the action was
carried out by an intentional agent, as opposed to simply being the result
of happenstance.19 The movement of the branch of a tree in the breeze,
for example, would not be something that could be morally good, on the
grounds that it does not meet the requirements for being a bearer of the
property of being morally good. This suggests that thin predicates like
moral goodness do impose some descriptive content on their bearers,
even if it is more minimal than the descriptive content imposed by thicker
predicates.
If we take this line, then we can say that all moral predicates will have
a degree of thickness, with the thinner predicates being predicates like
moral goodness, and the thicker predicates being predicates like coura-
geousness. There will also be predicates thicker than courageousness
which impose very specific descriptive features on their bearers, such as
the property of being a good father.20 However, we would say that, despite
the variations in degrees of thickness, these are all moral predicates. The
fact that some moral predicates imply some descriptive content does not
mean that they are not moral predicates, as predicates of pretty much any
property imply some constraints on what can bear it. The status of these
predicates as moral predicates will be determined by the functional role
that these predicates play in the cultivation and evaluation of character,
and the way we decide what we ought to do.
Moreover, this explanation generalizes to thick predicates of any kind,
and is not just intended for moral predicates. The key idea is that we look
to the function of a predicate to determine the domain to which the sen-
tence in which it occurs belongs. For example, we might take the predi-
cates ‘is sublime’ and ‘is beautiful’ to be thick predicates, as, even though
they seem to be aesthetic predicates, they have different descriptive con-
tent. However, the fact that they have different descriptive content does
not affect their status as aesthetic predicates due to the function they have
as ways to evaluate of pieces of art, for example. Once again, the fact that
100 D. Edwards
some predicates also have some descriptive content does not affect their
key function.
I have suggested that the examples of mixed atomics divide into two
main classes: (i) those where object and property are from different
domains; (ii) those where the property seems mixed. I argued that neither
of these classes poses a problem for the account of domains given above.
8 Summary
As noted earlier, the notion of a domain has been both a key and contro-
versial aspect of pluralist theories. Problems of domain individuation,
such as the problem of mixed atomics and the problem of mixed com-
pounds, along with the general difficulties of ‘dividing up’ language into
different domains have presented a significant challenge to the role of the
The Metaphysics of Domains 103
Notes
1. I will be using sentences, as opposed to propositions, as the main examples
of truth-bearers. See Edwards (2018: Chap. 1) for more on this choice.
2. For examples of domain-based ontological pluralism, see Cotnoir and
Edwards (2015) and Edwards (2018). See Lynch (2009) for an example
of domain-based logical pluralism.
3. Note that Wyatt (2013) recommends talking of ‘topics’ and ‘domains’ as
separate things, with Wyatt’s ‘topics’ loosely corresponding to my ‘meta-
physical aspects’, and Wyatt’s ‘domains’ loosely corresponding to my
‘semantic aspects’. I choose to use the word ‘domain’ for both, because I
do not think that these aspects can be separated enough to warrant them
being called different things, as opposed to parts of the same thing. I
hope it will become clear why below.
4. See Edwards (2018) for further development of this idea in relation to
social and institutional predicates.
5. Examples of error theory in morality are Mackie (1977) and Joyce
(2001).
6. See Haslanger (2012: 89–98) for an extended discussion of coolness.
7. For more on this distinction in relation to truth, see Edwards (2013).
8. The name is due to Armstrong (1978).
9. Note that this is not the most abundant view of properties available. As
Lewis (1983) notes, if we take the view that properties are classes (class
nominalism), then properties will be more abundant than on predicate
nominalism, as there will be classes to which there is no predicate attached.
10. See Edwards (2014: Chap. 5) for more on this view.
11. See, for example, Armstrong (1978). See also Edwards (2014: Chap. 2).
12. This perhaps requires that universals are taken to be the immanent uni-
versals favoured by Armstrong (1978), as opposed to abstract universals,
and I will assume that here.
13. See Edwards (2014: Chap. 6) for more on this idea.
14. Note that this terminology still applies if we are thinking about proper-
ties as classes, and classes as mind-independent. This is because, even if
104 D. Edwards
References
Armstrong, D.M. 1978. Universals and Scientific Realism. Cambridge:
Cambridge University Press.
Cotnoir, A.J. 2009. Generic Truth and Mixed Conjunctions: Some Alternatives.
Analysis 69 (2): 473–479.
Cotnoir, A.J., and D. Edwards. 2015. From Truth Pluralism to Ontological
Pluralism and Back. Journal of Philosophy 112 (3): 113–140.
David, M. 2013. Lynch’s Functionalist Theory of Truth. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 42–68.
New York: Oxford University Press.
Edwards, D. 2008. How to Solve the Problem of Mixed Conjunctions. Analysis
68 (2): 143–149.
———. 2009. Truth-Conditions and the Nature of Truth: Re-Solving Mixed
Conjunctions. Analysis 69 (4): 684–688.
———. 2013. Truth as a Substantive Property. Australasian Journal of Philosophy
91 (2): 279–294.
———. 2014. Properties. Cambridge: Polity Press.
———. 2018. The Metaphysics of Truth. Oxford: Oxford University Press.
Eklund, M. 2011. What Are Thick Concepts? Canadian Journal of Philosophy 41
(1): 25–49.
Gamester, W. Forthcoming. Logic, Logical Form, and the Disunity of Truth.
Analysis.
Hale, B. 1994. Singular Terms. In The Philosophy of Michael Dummett, ed.
B. McGuinness and G. Oliveri, 17–44. Boston: Kluwer Academic Publishers.
Hale, B., and C.J.G. Wright. 2005. Logicism in the 21st Century. In The Oxford
Handbook of Philosophy of Mathematics and Logic, ed. Stewart Shapiro,
166–202. Oxford: Oxford University Press.
Haslanger, S. 2012. Resisting Reality: Social Construction and Social Critique.
New York: Oxford University Press.
Joyce, R. 2001. The Myth of Morality. Cambridge: Cambridge University Press.
Kim, S., and Nikolaj J.L.L. Pedersen. 2018. Strong Truth Pluralism. In Pluralisms
in Truth and Logic, ed. Nikolaj J.L.L. Pedersen, Jeremy Wyatt, and Nathan
Kellen. Cham: Palgrave Macmillan.
Lewis, D. 1983. New Work for a Theory of Universals. Australasian Journal of
Philosophy 61 (4): 343–377.
Lynch, M.P. 2009. Truth as One and Many. Oxford: Oxford University Press.
106 D. Edwards
———. 2013a. Three Questions About Truth. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 21–41. New York:
Oxford University Press.
———. 2013b. Expressivism and Plural Truth. Philosophical Studies 163 (2):
385–401.
Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin.
Pedersen, Nikolaj J.L.L., and M.P. Lynch. 2018. Truth Pluralism. In The Oxford
Handbook of Truth, ed. M. Glanzberg. Oxford: Oxford University Press.
Pedersen, Nikolaj J.L.L., and C.D. Wright 2013. Pluralist Theories of Truth. In
The Stanford Encyclopedia of Philosophy, ed. E.N. Zalta, http://plato.stanford.
edu/archives/spr2013/entries/truth-pluralist/.
Sher, G. 2005. Functional Pluralism. Philosophical Books 46: 311–330.
Smith, M. 1994. The Moral Problem. Malden: Wiley-Blackwell.
———. 2013. On the Nature and Significance of the Distinction Between
Thick and Thin Ethical Concepts. In Thick Concepts, ed. S. Kirchin, 97–120.
Oxford: Oxford University Press.
Stewart-Wallace, A. 2016. An Old Solution to the Problem of Mixed Atomics.
Acta Analytica 31 (4): 363–372.
Tappolet, C. 2000. Truth, Pluralism and Many-Valued Logics. The Philosophical
Quarterly 50 (200): 382–383.
Williams, B. 1985. Ethics and the Limits of Philosophy. London: Routledge.
Wright, C.J.G. 1983. Frege’s Conception of Numbers as Objects. Vol. 2. Aberdeen:
Aberdeen University Press.
Wyatt, J. 2013. Domains, Plural Truth, and Mixed Atomic Propositions.
Philosophical Studies 166 (1): 225–236.
Strong Truth Pluralism
Seahwa Kim and Nikolaj J. L. L. Pedersen
Earlier versions—or parts of this chapter—have been presented by Nikolaj J. L. L. Pedersen at the
University of St. Andrews (January 2009); University College Dublin (October 2010); University
of Tokyo (October 2012); Sungkyunkwan University (December 2012); Northern Institute of
Philosophy (July 2013); Truth & Pluralism (Pacific APA 2014); University of Barcelona
(LOGOS, July 2014); Yonsei University (December 2014); University of Toronto (April 2015);
University of Connecticut (April 2015); Nanyang Technical University (August 2015); and
Lingnan University (February 2016). An earlier version of the chapter was presented by both
authors at the 1st Pluralisms Global Research Network Workshop at Yonsei University (January
2014). Thanks to the following people for helpful discussion: Dorit Bar-On, Jc Beall, Mandel
Cabrera, Ben Caplan, Colin Caret, Roy Cook, Aaron Cotnoir, Doug Edwards, Filippo Ferrari,
Tim Fuller, Sungil Han, Joe Hwang, Lina Jansson, Jinho Kang, Junyeol Kim, Sungsu Kim, Max
Kölbel, Michael Lynch, Adam Murray, Franklin Perkins, Graham Priest, Gurpreet Rattan, Sven
Rosenkranz, Stewart Shapiro, Gila Sher, Keith Simmons, Cory Wright, Crispin Wright, Jeremy
Wyatt, Byeong-Uk Yi, Andy Yu, and Elia Zardini. Research for this chapter was supported by
grant no. 2013S1A2A2035514 (Pedersen and Kim) and grant no. 2016S1A2A2911800
(Pedersen) from the National Research Foundation of Korea. We gratefully acknowledge this
support.
S. Kim
Scranton College, Ewha Womans University, Seoul, South Korea
N. J. L. L. Pedersen (*)
Underwood International College, Yonsei University, Incheon, South Korea
e-mail: nikolaj@yonsei.ac.kr
1 Truth Pluralism
A survey of traditional theories of truth reveals significant differences in
the details of the various theories. Some account for truth in terms of
correspondence to reality, others in terms of coherence. Yet others favor a
pragmatist story put in terms of what is believed at the end of enquiry or
what it is useful to believe. Interestingly, despite their differences tradi-
tional theories share a very fundamental assumption: monism about
truth. Advocates of correspondence, coherence, and pragmatist theories
have traditionally approached the truth debate as a one-player game. One
theory—and one theory only—can tell us what the nature of truth is.1
Recently, the fundamental assumption of monism has been chal-
lenged. Truth pluralists hold that there are different ways of being true.
Propositions about riverbanks might be true in virtue of corresponding
to reality while propositions about the law might be true in virtue of
cohering with the body of law. According to pluralists, monists are thus
wrong in maintaining that exactly one of the traditional theories correctly
identifies a single way of being true, applicable across all truth-apt
domains of discourse. Instead two or more accounts are correct, assum-
ing that their scope is restricted to specific domains.
Truth pluralism has attracted a great deal of attention in the last few
decades, with Crispin Wright and Michael P. Lynch being the view’s two
most prominent advocates. Wright and Lynch do not stand alone, how-
ever. A steadily increasing number of truth theorists endorse, develop, or
defend some form of pluralism.2
Different versions of pluralism have emerged. Strong pluralists give up
on the idea of truth-as-such by denying that there is a single truth prop-
erty applicable across all truth-apt domains of discourse. Truth is many,
not one. Moderate pluralists, on the other hand, hold on to the idea of
truth-as-such. There is a single generic truth property that applies across
all truth-apt discourse. Truth is one. However, at the same time truth is
many because propositions belonging to different domains may possess
the generic truth property in virtue of having distinct properties such as
correspondence or coherence.
This chapter presents and develops a form of strong pluralism. We
spell this idea out in a strongly reductionist fashion (to be given a slightly
more refined formulation in section more refined formulation in Sect. 3):
Strong Truth Pluralism 109
(TR) There is a plurality of properties T1, …, Tn such that (i) for every
proposition Φ, Φ’s being true is identical to Φ’s being Ti
(1 ≤ i ≤ n), and (ii) there is no property among T1, …, Tn that
satisfies (i) for every Φ.
Some discourse concerns more than one domain. Consider, for example,
the following compound proposition:
( mix - Ù ) Mt.Everest is extended in space and Bob’s drunk driving is illegal.
This is a mixed conjunction. It pertains to more than one domain, the
empirical domain and the legal domain.
The problem of mixed compounds emerges when we consider the
issue of what to say about the truth of mixed compounds.5 Clearly (mix-
∧) is true—its conjuncts, after all, are both true. However, it is less clear
how exactly the strong pluralist is going to account for this. According to
our brand of strong pluralism, each conjunct belongs to a specific domain
and within that domain truth is (token-)identical to a certain property.6
Let us suppose that the truth property for the first conjunct is correspon-
dence, and that it is coherence with the body of law for the second.
However, what truth property does the conjunction itself have? Given
our favored form of pluralism this amounts to the question of which
property is identical to truth of the conjunction. Let us consider two
arguably implausible options.
One option is to say that the conjunction is correspondence-true. (We
sometimes use the label “F-true” or “F-truth” to signify that F is the
truth-reducing property for a given proposition.) This is implausible
because it completely neglects the contribution made by the second con-
junct—a conjunct that is coherence-true. Another option is to say that the
conjunction is coherence-true. However, this is implausible for the same
kind of reason: it would completely neglect the contribution made by
one of the conjuncts.
A third option is to say that the conjunction possesses some property
distinct from both correspondence and coherence. We favor a response of
this kind. However, we recognize that anyone who does so faces two
tasks:
1. Something must be said about the nature of the third property. What
property is it? Also, while it is distinct from the truth properties of
both of its conjuncts, does its nature or instances somehow depend on
them, and if so, how?
Strong Truth Pluralism 111
2. Since the strong pluralist rejects the idea that there is a truth property
that applies across all truth-apt domains, an assurance must be
provided that the property possessed by the compound is not such a
property.
Propositions and Domains
Despite the central role that domains play within the standard pluralist
framework not much systematic work has been done on their nature. We
hope to do at least a little better by outlining a systematic treatment of
domains.7
We take the range of propositions that pluralists are interested in (and
typically deal with) to be the following:
Prop:
( D Ú Ù - pure ) If D (F j ) = ¼ = D (F j + m ) = Di , then
D (F j ∨…∨ F j+m ) = D (~ (F j ∨…∨ F j+m ))
= Di
Strong Truth Pluralism 113
What properties are in the plurality of truth properties T1, …, Tn? There
is a number of base-level truth properties Tj, …, Tj + m. These include the
“standard” pluralist properties such as correspondence, coherence,
114 S. Kim and N. J. L. L. Pedersen
(TA) For atomic Φ, Φ’s being true is identical to Φ’s being Tk, where Tk
is the base-level truth property that reduces truth for atomic prop-
ositions in Φ’s domain.
Hyperintensionality
(p1) is true while (p2) is false, and (p ∧ p1) is false while (p ∧ p2) is true.
This shows that, although p and p ∧ p are intensionally equivalent, they
cannot be the same proposition. For, what would happen if they were?
This would clash with our observations concerning (p1), (p2), (p ∧ p1),
and (p ∧ p2)—observations that are non-negotiable commitments of our
view.
In light of these considerations we cannot adopt any account that
treats p and p ∧ p as having the same content or as being the same propo-
sition. Thus, for example, we cannot adopt a standard possible-worlds
Strong Truth Pluralism 117
4 Truth Grounding
On the picture presented above, correspondence, coherence, superwar-
rant, and so on are truth-reducing properties only for atomic proposi-
tions. T∨ and T∧ are truth-reducing properties only for compounds of a
particular logical form, that is, disjunctions and conjunctions,
respectively.
We now turn to the issue of how compound truth is grounded. Our
account of truth grounding is complementary to our account of truth
reduction. Both accounts are needed for a comprehensive metaphysics of
truth. The issue of truth reduction for a given compound is the issue of
specifying a property to which the truth of that compound is identical.
The issue of truth grounding for a given compound is the issue of map-
ping how, or in what ways, the truth of the compound depends on the
semantic status of its constituents. Given the characterization of T∨, if p
∨ q is T∨, this is so because at least one disjunct has its truth-reducing
property. But this is to say that the instantiation of the truth-reducing
property of a disjunct suffices to ground the instantiation of T∨, that is,
disjunction-truth. Given the characterization of T∧, if p ∧ q is T∧, this is
118 S. Kim and N. J. L. L. Pedersen
“[p]” is read as the fact that p while “[p] ← [q]” is read as [p] is grounded
by [q] (or [q] grounds [p]). Upper-case Greek letters denote (possibly
empty) sets of facts. Strong asymmetry says that it is not the case that [p]
is among the grounds of [q], provided that [q] grounds [p]. Strong irre-
flexitivity says that no fact is among its own grounds, and transitivity says
that if [q], Γ ground [p] and ∆ grounds [q], then Γ, ∆ ground [p].11
These general observations about truth grounding and grounding’s
features can be used to shed light on the grounds of truth in relation to
specific compounds.
Consider the following five compounds (parentheses used to indicate
scope):
Given (T∨*) we can say something about the metaphysical ground of the
truth of (∨1) and (∨2). As earlier, suppose that correspondence and
coherence with the body of law are the truth-reducing properties for
Strong Truth Pluralism 119
the second conjunct, 〈Bob’s drunk driving is illegal〉’s cohering with the
body of law.
Applying transitivity to (∨3) we get that the truth of 〈(Mt. Everest is
extended in space and Bob’s drunk driving is illegal) or 2 + 2 = 5〉 is
grounded in 〈Mt. Everest is extended in space〉’s corresponding to reality
and 〈Bob’s drunk driving is illegal〉’s cohering with the body of law.
The truth of (∧2), like the truth of any other conjunction, is identical
to its being a conjunction with conjuncts that all possess their truth-
reducing property. (∧2) has two conjuncts, each of which is a disjunc-
tion. Their truth reduces to being a disjunction with at least one disjunct
that has its truth-reducing property. For 〈Mt. Everest is extended in space
or 2 + 2 = 5〉 there is exactly one such disjunct—the first one. This dis-
junct has its truth-reducing property, that is, 〈Mt. Everest is extended in
space〉 corresponds to reality. For 〈Bob’s drunk driving is illegal or
3 + 8 = 12〉 there is also precisely one disjunct that has its truth-reducing
property—again, the first one. This disjunct has its truth-reducing prop-
erty, that is, 〈Bob’s drunk driving is illegal〉 coheres with the body of law.
How about grounding? We have the following instances of grounding:
(∧2)’s being a conjunction with conjuncts that all have their truth-
reducing property is grounded jointly in 〈Mt. Everest is extended in
space or 2 + 2 = 5〉’s being a disjunction with at least one disjunct that has
its truth-reducing property and 〈Bob’s drunk driving is illegal or
3 + 8 = 12〉’s being a disjunction with at least one disjunct that has its
truth-reducing property. 〈Mt. Everest is extended in space or 2 + 2 = 5〉’s
being a disjunction with at least one disjunct that has its truth-reducing
property is grounded in 〈Mt. Everest is extended in space〉’s correspond-
ing to reality. 〈Bob’s drunk driving is illegal or 3 + 8 = 12〉’s being a dis-
junction with at least one disjunct that has its truth-reducing property is
grounded in 〈Bob’s drunk driving is illegal〉’s cohering with the body of
law. Applying transitivity we get that (∧2)’s being a conjunction with
conjuncts that all have their truth-reducing property is grounded in 〈Mt.
Everest is extended in space〉’s corresponding to reality and 〈Bob’s drunk
driving is illegal〉’s cohering with the body of law.
The resulting picture of truth grounding is this: the truth of any given
atomic is identical to its truth-reducing property (correspondence,
coherence, etc.). The truth of any compound is identical to its
Strong Truth Pluralism 121
1. Something must be said about the nature of the truth property pos-
sessed by mixed compounds. In particular, while it is distinct from the
truth properties possessed by its constituents, does its nature or
instances somehow dependent on them—and if so, how?
2. Since the strong pluralist rejects the idea that there is a truth property
that applies across all truth-apt domains, we need some assurance that
the truth property possessed by neither disjunctions nor conjunctions
is not a property of this kind.
Turn to the second issue, the issue of giving an assurance that the
truth-reducing properties possessed by mixed disjunctions and mixed
conjunctions do not apply across the board. This is straightforward. We
just need to make two observations. The first observation is that the
truth-reducing property for disjunctions, whether pure or mixed, is the
property of being a disjunction with at least one disjunct that has its
truth-reducing property. This is a property that only disjunctions can
have, and so, the truth-reducing property for mixed disjunctions is not a
generic truth property. The second observation is that the truth-reducing
property for conjunctions, whether pure or mixed, is the property of
being a conjunction with conjuncts that all have their truth-reducing
property. This is a property that only conjunctions can have, and so, the
truth-reducing property for mixed conjunctions is not a generic truth
property.
6 Conclusion
We set ourselves two aims in this chapter. The first was to present and
develop a version of strong truth pluralism. The second was to give a
strongly pluralist solution to the problem of mixed compounds. We
have achieved both aims. In Sects. 3 and 4 we introduced and devel-
oped a version of strong pluralism, an alethic incarnation of the (token-
token) identity theory. The metaphysics of the view was presented in
some detail, with a specification of its range of truth-reducing proper-
ties and an account of truth grounding. We introduced compound-
specific truth-reducing properties—identifying the truth of disjunctions
with being a disjunction with at least one disjunct that has its truth-
reducing property and the truth of conjunctions with the being a con-
junction with conjuncts that all have their truth-reducing property.
The nature of these properties put us in a position to offer a straightfor-
ward solution to the problem of mixed compounds. We thus conclude
that whatever insurmountable problems or challenges might seem to
confront strong pluralism, the problem of mixed compounds is not
one of them.
124 S. Kim and N. J. L. L. Pedersen
Notes
1. Correspondence theorists include David 1994, Devitt 1984, Newman
2007, Rasmussen 2014, Russell 1912, Vision 2004, and Wittgenstein
1921. Coherence theorists include Bradley 1914, Rescher 1973, Walker
1989, and Young 2001. Pragmatists or neo-pragmatists include James
1907, 1909, Peirce 1878, and Putnam 1981.
2. For example, Beall 2000, 2013; Cook 2011; Cotnoir 2009, 2013a, b;
Cotnoir and Edwards 2015; Edwards 2008, 2009, 2011, 2012, 2013,
2018; Gamester forthcoming; Kölbel 2008, 2013; Pedersen 2006, 2010,
2012a, b, 2014, ms-a, ms-b, ms-c; Pedersen and Edwards 2011; Pedersen
and Wright 2013a, b, 2016; Wyatt 2013; Yu 2017. Gila Sher and Terence
Horgan and various collaborators have developed a pluralist version of
the correspondence theory that incorporates different ways of corre-
sponding. See Sher 2005, 2013, 2016 and Horgan 2001; Barnard and
Horgan 2006, 2013. Works by Lynch and Wright include Lynch 2001,
2004, 2006, 2009, and 2013 and Wright 1992, 1994, 1995, 1996,
1998, 2001, and 2013.
3. Versions of strong pluralism are presented in Cotnoir 2009, 2013a;
Gamester forthcoming; Pedersen 2006, ms-a, ms-b, ms-c, Pedersen and
Lynch 2018 (Sect. 20.6, due to Pedersen).
4. See, for example, Lynch 2009. For surveys that provide a systematic pre-
sentation of objections to pluralism, see Pedersen 2012b, Pedersen and
Lynch 2018, Pedersen and Wright 2016.
5. The problem of mixed compounds is raised by Tappolet 2000, Sainsbury
1996, and Williamson 1994. Various pluralist options are discussed by
Cook 2011, Cotnoir 2009, Edwards 2008, 2009, 2018, Gamester forth-
coming, Lynch 2004, 2006, 2009, 2013, Pedersen 2012 b, Pedersen and
Lynch 2018, Sher 2005, 2013, and Wright 2013.
6. It is quite tedious always to use formulations that make our endorse-
ment of a token-token version of the identity theory explicit and dis-
tinguish it from its type-type counterparts. Sometimes we use
formulations that are compatible with the type-type identity theory.
However, in those cases it should be borne in mind that our reduction-
ism kicks in at the level of tokens. One reason to opt for token identity
is that it seems to integrate quite naturally with our account of truth
grounding, that is the other component of our proposed metaphysics
of truth. Grounding is usually regarded as a relation that obtains
Strong Truth Pluralism 125
References
Barnard, R., and T. Horgan. 2006. Truth as Mediated Correspondence. The
Monist 89: 31–50.
———. 2013. The Synthetic Unity of Truth. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 180–196. Oxford:
Oxford University Press.
Strong Truth Pluralism 127
Peirce, C.S. 1878/1992. How to Make Our Ideas Clear. Reprinted in e Essential
Peirce, ed. N. Houser and C. Kloesel, 124–141, Vol. 1. Bloomington: Indiana
University Press.
Putnam, H. 1981. Reason, Truth and History. Cambridge: Cambridge University
Press.
Rasmussen, J. 2014. Defending the Correspondence Theory of Truth. Cambridge:
Cambridge University Press.
Rescher, N. 1973. The Coherence Theory of Truth. Oxford: Oxford University
Press.
Russell, B. 1912. Problems of Philosophy. Reprinted by Oxford University Press,
1971.
Rosen, G. 2010. Metaphysical Dependence: Grounding and Reduction. In
Modality: Metaphysics, Logic, and Epistemology, ed. B. Hale and A. Hoffmann,
109–136. Oxford: Oxford University Press.
Sainsbury, M. 1996. Crispin Wright: Truth and Objectivity. Philosophy and
Phenomenological Research 56: 899–904.
Sher, G. 2005. Functional Pluralism. Philosophical Books 46 (4): 311–330.
———. 2013. Forms of Correspondence: The Intricate Route from Thought to
Reality. In Pedersen & Wright 2013a, 157–179.
———. 2016. Epistemic Friction: An Essay on Knowledge, Truth, and Logic.
Oxford: Oxford University Press.
Tappolet, C. 2000. Truth Pluralism and Many-Valued Logic: A Reply to Beall.
The Philosophical Quarterly 50: 382–384.
Vision, G. 2004. Veritas: The Correspondence Theory and Its Critics. Cambridge,
MA: MIT Press.
Walker, R.C.S. 1989. The Coherence Theory of Truth: Realism, Anti-realism,
Idealism. London: Routledge.
Williamson, T. 1994. A Critical Study of Truth and Objectivity. International
Journal of Philosophical Studies 30: 130–144.
Wittgenstein, L. 1921. Tractatus Logico-Philosophicus. In Annalen der
Naturphilosophie. English translation by D. F. Pears & B. F. McGuinness
(1961). London: Routledge & Kegan Paul.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University
Press.
———. 1994. Realism, Pure and Simple: A Reply to Timothy Williamson.
International Journal of Philosophical Studies 2: 327–341.
———. 1995. Truth in Ethics. Ratio 8: 209–226.
130 S. Kim and N. J. L. L. Pedersen
1 Introduction
Truth pluralism is the view that there are many ways of being true.
Further, these ways of being true vary with the domain of discourse.
While propositions about the physical world, for example
may be true in virtue of some correspondence with the world, this does
not seem plausible for other domains of discourse. Consider a proposi-
tion about morality, for example
N. Kellen (*)
University of Connecticut, Storrs, CT, USA
A principal reason for adopting truth pluralism is that the view provides a
framework for understanding the intuitive appeal of respectively realism
and antirealism with respect to different domains. The intuitive appeal
stems in part from the observation that both traditional realist accounts of
truth, such as the correspondence theory, and traditional antirealist
accounts, such as the coherence theory, face a similar pattern of failure.
Theories that seem plausible in some domains fail to seem as plausible in
others. (Pedersen and Lynch 2018, p. 1)
These two features are at the core of nearly every extant account of truth
pluralism. Truth pluralism, properly understood, is a combination of
realism and anti-realism into a single, coherent framework. It is this
motivation which underwrites the more obvious aspect of truth plural-
ism—that there are many ways of being true which vary by domain. It is
in this way that truth pluralism is the spiritual successor to the realism/
anti-realism debates which dominated analytic philosophy for decades
after Michael Dummett first introduced his semantic anti-realism in
(Dummett 1959). Truth pluralists are those who recognise that both
camps got something right—that there were intuitions from both realism
and anti-realism that should be captured by our theory of truth.
In what follows, I will take this understanding of truth pluralism as a
combination of realism and anti-realism and expand upon it by consider-
ing the extent to which various truth pluralisms are realist or anti-realist,
in virtue of their methodological commitments. After drawing two dis-
tinctions which will allow us to categorise various pluralist views as realist
or anti-realist I will introduce another form of truth pluralism, method-
ological truth pluralism, which does not privilege either its realist or anti-
realist aspects.
there tends to be a strong presumption in its favour such that any devia-
tion from it is characterised as a type of anti-realism (and of course plural-
ism is a deviation from global realism).
On the other hand, there is some reason to consider truth pluralism as
a form of realism. Anti-realism is generally characterised in terms of its
epistemic constraint: if a domain is anti-realist then all of its truths are
knowable in principle. But if a truth pluralism contains at least one realist
domain, then there will be some unknowable truths, violating the anti-
realist constraint, making it a form of realism.
While each of these positions have merit, I want to suggest a more
disciplined way of determining how a given truth pluralist theory should
be characterised. To do so, we must examine the methodological accounts
of the theories in question. As I’ve already noted, truth pluralists are plu-
ralists precisely because they wish to account for competing realist and
anti-realist intuitions. But some truth pluralists privilege one set of intu-
itions over the other, thus giving realist or anti-realist a privileged posi-
tion in their theory. Consider, for example, Crispin Wright’s
minimalism:
…although the notion of truth as correspondence to the facts might fit our
domain of discourse about the material world, a different notion of truth –
perhaps one with less metaphysical baggage, constructed out of coherence,
or justification or warrant – may fit the domains in which the correspon-
dence notion looks problematic. (Edwards 2011, pp. 31–2)
The principle here is the same as in the case of Wright and anti-realism.
The realist account of truth (correspondence) is the default, and one from
136 N. Kellen
…it is not enough for the fabric of our moral thought to be woven tightly –
to be durably coherent – it must also be nailed down, or grounded on a
firmer floor. (Lynch 2009, p. 173)
property had by all the true propositions, while moderate truth pluralists
(e.g. Lynch 2009) believe that there is a generic truth property which all
the true propositions have, in virtue of having one of the other, varying
truth properties. While much of the literature up until this point has
been dominated by moderate truth pluralisms, strong truth pluralism
has recent defenders, including Pedersen and Kim (2018), Cotnoir
(2013) and Ferrari, Morruzzi, Pedersen (ms.). We can also create a table
for this distinction10:
Strong truth pluralism Moderate truth pluralism
Wright (1992), Cotnoir (2013) Lynch (2009), Wright (2013)
Pedersen and Kim (2018) (Edwards 2011)
(Ferrari, Morruzzi, Pedersen, ms.)
Notice that this way of dividing truth pluralist views cuts across the
fundamentality distinctions I introduced. The strong/moderate distinc-
tion does not track the realism/anti-realism distinction whatsoever. While
Lynch (2009) and Wright (1992) differ according to the strong/moderate
distinction, it is not because they fall on opposite sides of the realist/anti-
realist debate, but rather because one advocates for a generic truth prop-
erty and another does not.11 If we are concerned with distinguish types of
truth pluralism based on their core commitments, I suggest that we do so
in virtue of my fundamentality criteria.
It may very well be that our language is so diverse that there is no rea-
son to hold a privileged attitude towards realism or anti-realism with
respect to arbitrary domains. Perhaps realism and anti-realism, contra
(Wright 1992; Edwards 2011; Lynch 2009) are on a par. Neither realism
nor anti-realism are the default, and neither require more evidence to
establish themselves as the metaphysics of a domain than the other;
instead, what they require is simply different evidence.
Consider again the assembly line metaphor. Absent some knowledge
of the general features of the boxes such that they are more likely to be
say, realist than anti-realist, it makes little sense for the worker to assume
that the next box coming down the line is realist. This of course does not
mean that she ought to assume that it is anti-realist instead. What she
ought to do is examine each box as it comes down the line and determine
how to categorise it, based on its own particular features.
So too for the truth pluralist, one might argue. Both realists and anti-
realists have claimed methodological superiority for their views. Realists
often note that realism better comports with common sense intuition,
while anti-realists claim that anti-realism is more metaphysically and
epistemically respectable. Perhaps what the pluralist ought to do is to step
back from each of these claims and instead take a methodologically neu-
tral stance, and examine each theory’s case with respect to each individual
domain of discourse. Call such an approach to truth pluralism method-
ological pluralism about truth.
Methodological pluralism about truth may have many virtues. It does
not claim—nor need to provide support for such a claim—that our lan-
guage and conceptual frameworks are structured in such a way that either
realism or anti-realism are more likely to hold in a particular domain. The
methodological pluralist about truth remains neutral (or silent) on that
issue. Given the difficulty of establishing such conclusions, and the long-
standing debates over their success, this should count as a point in favour
of the methodological pluralist.
Further, the methodological pluralist about truth is, in a sense, even
more pluralist than its rival views. Pluralism was developed because
monists failed to see that their theories of truth are implausible globally
but plausible locally. The truth pluralist who treats realism or anti-realism
as fundamental may commit the same sin as the monist, although on a
Methodological Pluralism About Truth 141
4 Conclusion
I have suggested that truth pluralism is best understood as the spiritual
successor to the realism and anti-realism debate, insofar as it attempts to
combine the best aspects of each theory. This raises a natural question:
how ought we understand this new, mixed view? To answer this, I sug-
gested that we again look at the core of truth pluralism: its realist and
anti-realist motivations, and how it treats these views in theory construc-
tion. I suggested that some truth pluralists hold one of the two to be
methodologically fundamental, that is, as the default stance from which we
move. Others go further, in not only giving a privileged methodological
status to one theory but to build the secondary account out of the first,
thereby making the primary aspect theoretically fundamental. I then
showed how these two distinctions can be used to categorise various types
of truth pluralism in a way that cuts across the standard way of under-
standing the literature.
142 N. Kellen
Notes
1. Barring the truth of some fully reductive moral naturalism, that is.
2. This is similar to what (Beall 2013, p. 324) calls “language-relative truth
pluralism”.
3. The only exception I am aware of is Beall (2013), if one accepts that his
deflated truth pluralism is a pluralism about truth as opposed to a plural-
ism about truth-predicates which may be unrelated to the actual philo-
sophically robust concept of truth. Note also that Beall’s pluralism does
not accept DOMAIN-VARIABILITY, which again sets it far apart from
other views in the literature.
4. This includes the most developed accounts of truth pluralism in Wright
(1992) and Lynch (2009), as well as, for example Cotnoir and Edwards
(2015), Edwards (2011), Pedersen (2006). Cotnoir (2013), who does
not explicitly call for dividing domains by realism/anti-realism but pro-
vides a semantic framework for truth pluralisms which have certain
domains which maintain classical logic and others which have paracom-
plete logics, a conclusion commonly held to follow from adopting real-
ism and anti-realism respectively. Another potential outlier is Gamester
(2017), who does divide up his truth pluralism into realist and anti-
realist parts, but the anti-realist parts are motivated by expressivism
Methodological Pluralism About Truth 143
References
Beall, J. 2013. Deflated Truth Pluralism. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 323–338. Oxford:
Oxford University Press.
Burgess, A., and J. Burgess. 2014. Truth. Princeton: Princeton University Press.
Cotnoir, A.J. 2013. Validity for Strong Pluralists. Philosophy and Phenomenological
Research 86 (3): 563–579.
Cotnoir, A.J., and D. Edwards. 2015. From Truth Pluralism to Ontological
Pluralism and Back. Journal of Philosophy 112 (3): 113–140.
Dummett, M. 1959. Truth. Proceedings of the Aristotelian Society 59 (1):
141–162.
Edwards, D. 2011. Simplifying Alethic Pluralism. The Southern Journal of
Philosophy 49 (1): 28–48.
Field, H. 1994. Deflationist Views of Meaning and Content. Mind 103 (411):
249–285.
Gamester, W. 2017. The Diversity of Truth: A Case Study in Pluralistic
Metasemantics. PhD Dissertation, University of Leeds.
Künne, W. 2005. Conceptions of Truth. Oxford: Clarendon Press.
Lynch, M.P. 2009. Truth as One and Many. Oxford: Oxford University Press.
Pedersen, Nikolaj J.L.L. 2006. What Can the Problem of Mixed Inferences
Teach Us About Alethic Pluralism? Monist 89 (1): 102–117.
Pedersen, Nikolaj J.L.L., and S. Kim. 2018. Strong Truth Pluralism. In Pluralisms
in Truth and Logic, ed. J. Wyatt, Nikolaj J.L.L. Pedersen, and N. Kellen,
107–130. London: Palgrave Macmillan.
Pedersen, Nikolaj J.L.L., and M.P. Lynch. 2018. Truth Pluralism. In The Oxford
Handbook of Truth, ed. M. Glanzberg. Oxford: Oxford University Press.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University
Press.
———. 1998. Truth: A Traditional Debate Reviewed. Canadian Journal of
Philosophy 28 (sup1): 31–74.
Wright, Crispin. 2013. A Plurality of Pluralisms. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 123–153. Oxford:
Oxford University Press.
Normative Alethic Pluralism
Filippo Ferrari
1 Introduction
Some philosophers have argued that truth is a norm of judgement.1 This
thesis has been given in a variety of formulations—that true judgements
are the correct ones; that it is better to judge truly than to judge falsely;
and that the truth is what judges ought to pursue in enquiry. I will assume
that truth somehow functions as a norm of judgement, and I will be
focusing on two core questions concerning the judgement-truth
norm—namely:
(i) what are the normative relationships between truth and judgement?
(ii) do these relationships vary or are they constant?
F. Ferrari (*)
Universität Bonn, Bonn, Germany
2 The Truth-Norm
Advocates of the thesis that truth is a norm disagree about how the norm
should be conceived, and many formulations of it have been given. The
following list provides a sample of formulations that can be found in the
recent debate (my emphases):
Horwich We ought to want our beliefs to be true (and therefore not-
want to have any false ones).4
James The true is the name of whatever proves itself to be good in
the way of belief.5
Loewer True belief is valuable and, other things being equal, it is a
rational doxastic policy to seek true beliefs.6
Lynch End of Inquiry: True propositions are those we should aim
to believe when engaging in inquiry.
Norm of Belief: True propositions are those that are correct
to believe.7
Normative Alethic Pluralism 147
While philosophers in the list take the truth norm to apply to belief,
I take judgements to be the things that truth is primarily normative
over. Following Shah and Velleman, I consider a judgement to be a
cognitive mental act of affirming a proposition. It is a mental act
because “it involves occurrently presenting a proposition, or putting it
forward in the mind. It is a cognitive act because it involves presenting
the proposition as true”.11 In this respect, judgements differ from
beliefs, which are cognitive mental attitudes. The precise relationship
between judgement and belief is a complex issue which won’t concern
us in this chapter.12 Somewhat simplistically, I take it that the end
product of a judgement is typically a belief whose content is a proposi-
tion. Since the examples I will consider involve an evaluative process, I
take them to be fairly typical cases of beliefs formed through an act of
judging. Thus, I use the expression ‘norms governing judgement’ as a
catchall expression to indicate norms governing the formation, main-
tenance, and relinquishing of beliefs formed by means of an act of
judging.
Two things about this list and the current debate on the judgement-
truth norm deserve some discussion. The first concerns the variability in
the use of the normative vocabulary involved in the formulations above.
Some formulations employ terms like ‘should’ or ‘ought’—thus using
deontic terms; others involve notions such as ‘valuable’ and ‘good’—thus
using an axiological vocabulary; others still talk in terms of truth’s being
the aim of beliefs—using a teleological vocabulary; and, finally, others
formulate the truth-norm in terms of correctness or fittingness—taking
these notions to be normatively independent of any axiological or deon-
tic element. I introduce the term criterial to indicate this latter category
of normative notions. Mixed formulations—that is, formulations mixing
different normative vocabularies—are also possible, as Lynch’s way of
148 F. Ferrari
cashing out the normative role of truth both in terms of aim and in terms
of correctness seems to suggest.
Abstracting from the many issues concerning the precise formulation
and interpretation of the truth norm, we can capture the highlighted
variability in the ways in which the norm is conceived by distinguishing
the following normative dimensions13:
Jill’s contrary view that people would deem appropriate for Julie to have
is one that falls somewhere in the middle of the two extremes of the fun-
damental moral case and the basic taste case. Thus, we may typically
expect Julie to be somehow surprised by Jill’s contrary judgement regard-
ing Gould’s 1955 execution of Goldberg Variations. This is because she
would think that Jill’s judgement is somehow off-colour. Thus, a certain
degree of criticism would be regarded as appropriate. In particular, it
would be natural for Julie to consider Jill’s aesthetic sensibility on this
occasion to be not as good as her own. In this respect, contrary to the
basic taste case, disagreement about refined aesthetics seems to give rise to
an attribution of fault. Although it seems appropriate for Julie to con-
tinue to regard Jill as well-informed as she is about Glenn Gould’s musi-
cal production, by being committed to her own judgement about Gould
and thus to her own scale of aesthetic value, she is committed to assessing
Jill’s opinion as inferior—indeed, one that it would be better not to have.
As a result, it might happen that in some cases the conversation continues
by each subject trying to persuade the opposite party to change her mind.
However, in contrast with the fundamental moral case described above,
the degree of intensity of the reactive attitude that would be appropriate
for Julie to have in response to Jill’s contrary view is significantly lower
than in the fundamental moral case. We could certainly understand, if
not justify, a high degree of heat in disputes about fundamental moral
issues, but not in disputes about refined aesthetics. Regardless of how
much aesthetically off-colour I consider your contrary view, granted that
we are equally knowledgeable about the subject matter in question I
would still feel some pressure to regard you as permitted to hold to the
view you endorse.25
Arguing for both (IND 1) and (IND 2) would get us full normative inde-
pendence, which means the strongest available form of NAP—call it
strong NAP. Otherwise we could argue for only one direction of norma-
tive independence, and we would get two possible moderate forms of
NAP—the first endorses (IND 1) without endorsing (IND 2), while the
second endorses (IND 2) without endorsing (IND 1).32 Since the rele-
vant direction of normative independence that we need in order to deal
with the variability illustrated in the examples above is that expressed by
(IND 1), in what follows I will argue for (IND 1) only, and I will remain
silent on (IND 2). Moreover, to provide a full defence of the relevant
moderate version of NAP we need an argument to show that (IND 2)
fails—that is, to show that we have at least one of the following directions
of normative dependence:
158 F. Ferrari
(DEP 1) deontic⇒axiological
(DEP 2) deontic⇒criterial
(DEP 3) axiological⇒criterial
Something may be bad without its badness being a matter of anyone’s hav-
ing done anything they ought not have done, and without its being the
case that there is anyone who ought to change it; some prospective state of
affairs or object may be good without its being the case that there is anyone
who ought to produce it or bring it about.38
(IND 3) c
riterial fault⇏axiological fault; criterial
fault⇏deontic fault; axiological fault⇏deontic fault.
11 Conclusions
I have discussed what I take to be an interesting variability in the norma-
tive significance of disagreement and I have suggested that such variabil-
ity requires some flexibility in the normative function that truth exerts in
different areas of discourse. Because a monistic view on truth’s normative
function—what I have called normative alethic monism—forces us to
adopt a Procrustean attitude towards the normative significance of dis-
agreement, it is inadequate as a general model of the normativity of truth.
I have argued for this by presenting a variation of what is known in the
truth pluralism literature as the scope problem.
I have outlined a pluralist framework for understanding the normativ-
ity of truth—normative alethic pluralism—that promises to score better
than NAM in addressing the variability in the normative significance of
disagreement discussed above. The key point of my proposal is to under-
stand variation in normative fault by looking at which dimensions of
truth’s normative profile operate in a given domain of discourse. Once we
adopt this pluralistic stance towards the normativity of truth, we obtain
164 F. Ferrari
Notes
1. For example: Dummett 1959; Gibbard 2005; Horwich 2013; Shah and
Velleman 2005; Wedgwood 2007; Wright 1992.
2. Since the label can be misleading in one important respect, I should
clarify that with it I mean a pluralist account of the normative function
that truth plays in relation to judgements. It is not part of the proposal
to claim that this pluralist account requires or entails a pluralistic account
of the nature of truth—although, the two views, taken together, gives a
highly coherent and neat package.
3. See Ferrari (2016b) for an account of the normativity of truth—espe-
cially of what I call the axiological dimension of the normativity of truth
(see below, section “The Truth-Norm”)—which is compatible with the
minimalist conception of truth advocated by Horwich (in, e.g., Horwich
1998). For further discussion of value and Horwich’s minimalism, see
Ferrari 2018.
4. Horwich 2013: 17.
5. James 1975: 42.
6. Loewer 1993: 266.
7. Lynch 2013: 24.
8. McHugh 2014: 177.
9. Wedgwood 2007.
10. Williams 1973: 136.
11. Shah and Velleman 2005: 503.
12. On this, see Chrisman 2016 and Sosa 2015, part III.
13. I focus on judgement rather than belief (as all the authors mentioned
above do) to avoid intricate issues concerning the doxastic voluntarism
versus non-voluntarism debate.
14. Singularity resembles Williamson’s simple account, but without the con-
stitutivist element. See Williamson 2000: 240.
15. See Williams 2012 for an account of normative silence.
16. MacFarlane 2014: Chap. 6.
Normative Alethic Pluralism 165
30. The cogency of this option depends on whether we can make sense of a
purely non-normative notion of truth. This is a debated issue among
philosophers working on truth: see, for instance, Dummett 1959;
Wright 1992; Lynch 2009; Wrenn 2015. For some replies, see, e.g.,
Horwich 1998; Ferrari 2016b; Ferrari and Moruzzi (2018).
31. ‘⇏‘should be read as: ‘does not enforce…’.
32. It is helpful to point out that the views that I call ‘strong NAP’ and ‘mod-
erate NAP’ are quite different from the views that are often called ‘strong
alethic pluralism’ and ‘moderate alethic pluralism’ in the alethic plural-
ism debate. Two remarks are especially relevant on this: first, that strong
NAP doesn’t entail strong alethic pluralism and, second, that moderate
NAP doesn’t entail moderate alethic pluralism.
33. See Thomson 2008 for a discussion of the various kinds of normativity
in relation to judgements.
34. See McHugh 2014 and McHugh and Way 2015.
35. There is a similar, and familiar, contrast in the normative ethics debate:
one might think that right actions are those that maximise utility but ask
what is good about doing what is right.
36. McHugh (2014: 177) suggests that we should understand correctness in
terms of fittingness: ‘the attitude of belief sets truth as the standard that
a proposition must meet in order for it to be a fit object of that attitude
[…] For an attitude to be fitting is for it to have a normative property.
But it is not fitting because you ought to hold it, or because you may
hold it, or because it would be good if you held it. Fittingness, I main-
tain, is distinct from these other normative properties’.
37. See, for instance, David 2005; Hazlett 2013; Kvanvig 2003; Lynch
2005; McHugh 2012.
38. McHugh 2012: 10.
39. Chisholm 1963: 3.
40. Driver 1992: 286.
41. Turri first applied the category of the suberogatory to the case of the
normativity of assertions, but differently, and with different aims; see
Turri 2013.
42. Ferrari and Moruzzi (2018).
43. This paper has enormously benefitted from discussions with Elke
Brendel, Matthew Chrisman, Massimo Dell’Utri, Douglas Edwards,
Matti Eklund, Andreas Fjellstad, Patrick Greenough, Thomas
Grundmann, Paul Horwich, Nathan Kellen, Michael Lynch, Giacomo
Melis, Anne Meylan, Moritz Müller, Carol Rovane, Andrea Sereni, Erik
Normative Alethic Pluralism 167
Stei, Elena Tassoni, Joe Ulatowski, Giorgio Volpe, Jack Woods, Chase
Wrenn, Cory Wright, Jeremy Wyatt, Luca Zanetti, Dan Zeman. Special
thanks are due to Carrie Ichikawa Jenkins, Sebastiano Moruzzi, Nikolaj
J. L. L. Pedersen, Eva Picardi and Crispin Wright. Moreover, I would
like to acknowledge the generous support of the Deutsche
Forschungsgemeinschaft (DFG—BR 1978/3–1) for sponsoring my
postdoctoral fellowship at the University of Bonn.
While working on this paper, I benefitted from participation in the
Pluralisms Global Research Network (National Research Foundation of
Korea grant no. 2013S1A2A2035514). This support is also gratefully
acknowledged.
References
Chisholm, R. 1963. Supererogation and Offense: A Conceptual Scheme for.
Ethics. Ratio 5: 1–14.
Chrisman, M. 2016. Epistemic Normativity and Cognitive Agency. Noûs.
Doihttps://doi.org/10.1111/nous.12184.
David, M. 2005. On Truth Is Good. Philosophical Books 46 (4): 292–301.
Driver, J. 1992. The Suberogatory. Australasian Journal of Philosophy 70 (3):
286–295.
Dummett, M. 1959. Truth. Proceeding of the Aristotelian Society 59: 141–162.
Ferrari, F. 2016a. Disagreement About Taste and Alethic Suberogation. The
Philosophical Quarterly 66 (264): 516–535.
———. 2016b. The Value of Minimalist Truth. Synthese. Doi https://doi.
org/10.1007/s11229-016-1207-9.
———. 2018. The Value of Minimalist Truth. Synthese 195 (3): 1103–1125.
https://doi.org/10.1007/s11229-016-1207-9.
Ferrari, F., and S. Moruzzi. 2018. Ecumenical Alethic Pluralism. Canadian
Journal of Philosophy. https://doi.org/10.1080/00455091.2018.1493880.
Ferrari, F., and C. Wright. 2017. Talking with Vultures. Mind 126 (503):
911–936.
Gibbard, A. 2005. Truth and Correct Belief. Philosophical Issues 15: 338–350.
Hazlett, A. 2013. A Luxury of the Understanding: On the Value of True Belief.
Oxford: Oxford University Press.
Horwich, P. 1998. Truth. 2nd ed. Oxford: Oxford University Press.
———. 2013. Belief-Truth Norms. In The Aim of Belief, ed. T. Chan, 17–31.
Oxford: Oxford University Press.
168 F. Ferrari
James, W. 1975. Pragmatism and the Meaning of Truth. Cambridge MA: Harvard
University Press.
Kvanvig, J. 2003. The Value of Knowledge and the Pursuit of Understanding.
Cambridge: Cambridge University Press.
Loewer, B. 1993. The Value of Truth. Philosophical Issues 4: 265–280.
Lynch, M. 2005. True to Life. Cambridge MA: The MIT Press.
———. 2009. Truth as One and Many. Oxford: Oxford University Press.
———. 2013. Three Questions for Truth Pluralism. In Truth and Pluralism, ed.
Nikolaj J.L.L. Pedersen and C.D. Wright, 21–41. Oxford: Oxford University
Press.
MacFarlane, J. 2014. Assessment Sensitivity: Relative Truth and Its Applications.
Oxford: Oxford University Press.
McHugh, C. 2012. The Truth Norm of Belief. Pacific Philosophical Quarterly 93:
8–30.
———. 2014. Fitting Belief. Proceedings of the Aristotelian Society 114 (2):
167–187.
McHugh, C., and J. Way. 2015. Fittingness First. Ethics 126 (3): 575–606.
Shah, N., and D. Velleman. 2005. Doxastic Deliberation. Philosophical Review
114 (4): 497–534.
Sosa, E. 2015. Judgement and Agency. Oxford: Oxford University Press.
Strawson, P. 1962. Freedom and Resentment. Proceedings of the British Academy
48: 1–25.
Thomson, J.J. 2008. Normativity. La Salle: Open Court.
Turri, J. 2013. Knowledge and Suberogatory Assertion. Philosophical Studies 3:
1–11.
Wedgwood, R. 2007. The Nature of Normativity. Oxford: Oxford University
Press.
Williams, B. 1973. Problems of the Self. Cambridge: Cambridge University Press.
Williams, R. 2012. Indeterminacy and Normative Silence. Analysis 72 (2):
217–225.
Williamson, T. 2000. Knowledge and Its Limits. Oxford: Oxford University Press.
Wrenn, C. 2015. Truth. Cambridge: Polity Press.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University
Press.
———. 1998. Truth: A Traditional Debate Reviewed. Canadian Journal of
Philosophy Supplementary 28: 31–74.
Wyatt, J. 2018. Absolutely Tasty: An Examination of Predicates of Personal
Taste and Faultless Disagreement. Inquiry 61 (3): 252–280.
Truth in English and Elsewhere:
An Empirically-Informed Functionalism
Jeremy Wyatt
We should reconcile ourselves with the fact that we are confronted, not with
one concept, but with several different concepts[.] [W]e should try to make
these concepts as clear as possible (by means of definition, or of an axiomatic
procedure, or in some other way); to avoid further confusions, we should
agree to use different terms for different concepts; and then we may proceed to
a quiet and systematic study of all concepts involved, which will exhibit their
main properties and mutual relations. (Tarski 1944, p. 355)
This chapter explores the future of functionalism about truth. I’ll aim to
defend functionalism while also explaining why and how functionalists
should rely on empirical evidence regarding ordinary thought about
truth. I’ll begin by sketching the functionalist framework and outlining
some of its signature virtues. Next, I’ll raise what I take to be the most
serious problem with functionalism—that although it entails empirical
claims regarding ordinary thought about truth, its main proponent
Michael Lynch has offered no empirical evidence in support of the view.
J. Wyatt (*)
Underwood International College, Yonsei University, Incheon, South Korea
I’ll then venture to patch this lacuna by discussing some extant empiri-
cal data on ordinary thought about truth. These data, while inconclusive
and in need of supplementation, support two underinvestigated pluralist
views of ordinary thought about truth. Once we integrate these pluralist
approaches with the functionalist framework, we get a much more
nuanced view that is appropriately sensitive to relevant empirical
findings.
To close my defense, I’ll take a look at a second, pressing objection to
functionalism that has been advanced by Cory Wright (2010) and explain
why it doesn’t point to any serious flaws in the view. My overall conclu-
sion, then, will be that functionalism is a distinctive and promising
approach to the study of truth that can withstand philosophical scrutiny
and cleanly integrate empirical inquiry into our ongoing search for the
nature of truth.
[A] property might have various features not reflected in our concept of
that property. To choose a well worn example, heat (the property of a phys-
ical object’s being more or less hot) is revealed by physics to be an average
kinetic energy of constituent molecules, even though our ordinary concept
of heat involves no such component (that’s not the way we ordinarily iden-
tify heat).
2 Alethic Functionalism
The Framework
disposition to believe p1,…,pn, provided that one has the concepts that
one needs to understand these propositions and has also reflected on the
propositions long enough to understand them.
To illustrate, Lynch takes FT to include the propositions expressed by
the following sentences5:
(Objectivity) A belief is true iff with respect to that belief, things are as
they are believed to be.
(Norm of Belief ) It is prima facie correct to hold a belief iff that belief is
true.
(End of Inquiry) Other things being equal, holding true beliefs is a wor-
thy goal of inquiry.
(PT) A belief exemplifies truth iff with respect to that belief, things are as
they are believed to be and it is prima facie correct to hold a belief iff
that belief exemplifies truth and other things being equal, holding
beliefs that exemplify truth is a worthy goal of inquiry.
(RT) There is a y (a belief exemplifies y iff with respect to that belief, things
are as they are believed to be and it is prima facie correct to hold a
belief iff that belief exemplifies y and other things being equal, holding
beliefs that exemplify y is a worthy goal of inquiry).
Truth in English and Elsewhere: An Empirically-Informed… 173
The final step is to determine whether any actual properties witness RT.
If we discover that exactly one property does so, then we say that the folk
theory FT is uniquely realized. In this case, we can simply call the realizing
property truth. Accordingly, call this possibility monism. However, the
functionalist also leaves room for two further possibilities. It may be that
many properties witness RT; in this case, we say that FT is multiply real-
ized, or that we’ve confirmed a certain sort of pluralism about truth. The
third possibility is that upon investigation, we find that no property wit-
nesses RT. In this case, we say that FT is unrealized, that is, that no actual
property is represented by the folk concept truth. Call this result defla-
tionism.6 Figure 1 summarizes the Conceptual and Metaphysical Stages
of functionalism.7
One attractive feature of functionalism is that it incorporates
Ramsification, which is a precise and well-understood method.8 It also
comports with Alston’s insight by allowing for the possibility that the
representation provided by the folk concept truth is partially accurate,
even as there is more to say about the metaphysics of truth than what is
articulated in this concept. If FT is uniquely or multiply realized, then the
properties that realize it may have additional features that aren’t men-
tioned in FT. On the other hand, if FT is unrealized, then that, too, will
probably come as a surprise to many possessors of the folk concept
truth, insofar as they presumably take some property in the world to
answer to their thought and talk about truth. This, then, would also be a
sense in which the truth about the nature of truth diverges from the rep-
resentation afforded by truth.
We’ve seen, then, that functionalism enjoys some significant virtues, and
this is all to the good. However, I want now to describe what I take to be
functionalism’s most serious shortcoming and to then explain how we
should remedy it.
The functionalist’s main ambition at the Conceptual Stage is to offer
an account of the folk concept truth. Their driving idea is that this con-
cept is underwritten by the folk theory FT of truth, and Lynch articulates
a view as to which propositions comprise FT. We should now ask: what
evidence do we have that FT in fact consists of the propositions expressed
by (Objectivity), (Norm of Belief ), and (End of Inquiry)?
Lynch offers two lines of evidence in support of his view on FT. One is
that he suspects that the propositions at issue are among our most funda-
mental (implicit) beliefs about truth.10 The other is that certain influen-
tial philosophers such as Aristotle, James, and Peirce have taken these
propositions to be central truths about truth.11
The basic problem is that these lines of evidence are inadequate in
principle. Lynch’s view about FT is empirical—it is a view about how
ordinary thinkers are, in fact, disposed to think about truth. However, we
have little reason to believe that philosophers enjoy special insight into
this issue. When we receive our professional training, we no doubt acquire
a range of new abilities, but we don’t tend to spend much time in gradu-
ate school systematically examining how untrained, native speakers think
Truth in English and Elsewhere: An Empirically-Informed… 175
As we’ll see, our current data point toward a notable degree of complexity
in ordinary thought about truth. Accordingly, it will be useful to clearly
articulate a pluralist view of ordinary thought about truth that has two
main features. First, it should at least be supported by many of our data.
Second, such a view should be able to serve as a guiding hypothesis in
future studies of ordinary thought about truth.
176 J. Wyatt
The basic thesis behind what we’ll call conceptual pluralism about truth
is that there is more than one actual, folk truth concept. Notice, then,
that conceptual pluralism isn’t the claim that there could be thinkers who
use a truth concept that differs from the one used by, say, contemporary
English speakers. Rather, it is a claim about the actual world to the effect
that it contains thinkers some of whom use a different truth concept than
do others.
When we evaluate conceptual pluralism empirically, the following
basic procedure should serve us well.12 We begin by examining ordinary
speakers’ usage of ‘true’ and ‘truth’, as well as words in other languages
that can be directly translated using ‘true’ or ‘truth’. Call any such word a
piece of alethic vocabulary. If we find significant differences in speakers’
usage of alethic vocabulary, then we should go on to determine whether
some speakers associate different concepts with these words than others
do. Any concepts that we discover we’ll provisionally call truth
concepts.13
We can distinguish between two kinds of conceptual pluralism. The
first is what we’ll call interlinguistic conceptual pluralism:
(Inter) There are at least two actual linguistic communities L1 and L2 such
that some L1-members use a truth concept T1, whereas some L2-
members use a distinct truth concept T2.
T1 T2 T2 T3
L1 L2
(Intra) There is at least one actual linguistic community L such that some
L-members use a truth concept T1, whereas other L-members use a
distinct truth concept T2.
Bruno case: Bruno has just finished painting his house. Bruno painted his
house the same color as the sky on a clear summer day. Bruno claims
his house is blue.
Truth in English and Elsewhere: An Empirically-Informed… 179
Donna case: Donna is traveling in Germany, but does not speak German.
She watches as a sailor asks for der Stadtplan and is handed what looks
like a map of the city. Donna asks for der Stadtplan in a shop and is
sold a city map. Donna still speaks no German, but believes that ask-
ing for der Stadtplan is a good way to obtain a city map from a German
shopkeeper.
(WC1) Few (say, < 50% of ) native English speakers are disposed to believe
upon reflection that (CT) is viciously circular.
Barnard and Ulatowski (2017, Tables 3 and 5) found that 86% of their
philosophers and 85.8% of their non-philosophers either agreed or
strongly agreed with (3). It’s thus plausible that most of these subjects
took (3) to be free from vicious circularity.
These findings are also admittedly limited in the present context, given
that (3)—in contrast to (CT)—doesn’t explicitly mention correspon-
dence and wasn’t explicitly presented as a (partial) definition of truth.
However, like Bourget and Chalmers’ findings, they also provide weak
support for (WC1). In light of the limitations of the existing evidence
that bears on (WC1), it would be especially helpful for future studies to
directly examine the degree to which English speakers take (CT) to be
viciously circular.
When exploring English and Akan speakers’ views on truth and corre-
spondence, Wiredu (1987, p. 29) also considers a direct translation of
(CT) into Akan:
In (CTA), ‘true’ and ‘fact’ are translated using the same Akan expression
‘ete saa’, which means it is so. Thus, as Wiredu points out, (CTA) can also
be faithfully translated into English as ‘That a proposition is so amounts
to its coinciding with what is so’. This observation leads Wiredu to offer
the following insightful conjecture:
Truth in English and Elsewhere: An Empirically-Informed… 183
This has the beauty of a tautology, but it teaches little wisdom. It seems
to me unlikely that thinking in this language, one could be easily
tempted into correspondence formulations of this sort. Indeed, in this
language, it is pretty clear that the problem of truth must be the problem
of clarifying the idea of something being so. (Ibid. Cp. 1985, p. 47;
2004, pp. 48–9)
We can thus render Wiredu’s second conjecture along the following lines:
(WC2) Most (> 50% of ) native Akan speakers are disposed to believe
upon reflection that (CTA) is viciously circular.
(a) Monism: All of the folk theories FT1,…,FTn are uniquely realized by
the same property P;
(b) Pluralism1: All of the folk theories are uniquely realized, but some are
realized by a different property than others;
(c) Pluralism2: Some, but not all, of the folk theories are multiply
realized;
(d) Pluralism3: All of the folk theories are multiply realized; and
(e) Deflationism: All of the folk theories are unrealized.
Options (a) and (e) are the most familiar, insofar as they respectively
represent monism and a pure sort of deflationism. We can helpfully think
of options (b)–(d) as representing increasingly complex grades of plural-
ism about the properties that realize the various folk theories. These
grades of pluralism mark the basic degrees to which multiple realization
can creep into our metaphysics of truth when we try to navigate the
divide between monism and deflationism.25
What I would emphasize at this stage is that before we endorse any of
these metaphysics of truth, we must do a great deal of further
metaphysical—and empirical—work. My hope is that the updated func-
tionalist framework will serve as a useful instrument for this work.
For the sake of argument, I will suppose with Wright that the functional-
ist takes all of PT’s conjuncts to be true.29 We then pose Wright’s problem:
to know that PT is true, mustn’t the functionalist already know the nature
of the property that ‘truth’ denotes, that is, truth? But isn’t their view that
our knowledge of truth’s nature is based on our prior knowledge of PT? If
so, then it seems that to come to know that PT is true, they must already
know that PT is true, which is impossible.
The functionalist should dispel this impression of circularity by point-
ing out that they can know that PT is true without already knowing the
nature of truth. This follows from a more general principle. Assuming
that A is a competent user of the word ‘true:’30
(4) If A knows that p and they know that the content of sentence ‘S’ is
that p, then A knows that ‘S’ is true.
To see the plausibility of (4), suppose that a high school student Van
looks at the whiteboard in his classroom and sees the sentence ‘The chem-
istry exam will begin at 1:30 p.m. today’. Van knows that the content of
188 J. Wyatt
this sentence is that the chemistry exam will begin at 1:30 p.m. on that
day. If he also knows that the exam will in fact begin at 1:30 p.m. on that
day, then he knows that the sentence on the whiteboard is true. It’s sim-
ply not the case that to know that the sentence is true, Van must know
what the nature of truth is. It’s enough that he’s a competent user of the
word ‘true’ and thus understands how this word applies to English
sentences.
Let’s now think in the same vein about PT. Suppose that PT is the con-
junction ‘S1 and S2 and S3’ and that the functionalist maintains that PT is
true. We ask them: how do you know that PT is true? To explain this, the
functionalist needs only point out (i) that the contents of its conjuncts
are respectively that S1, that S2, and that S3 and (ii) that they know that
that S1, that S2, and that S3. From (i), (ii), and (4), it follows that they
know that PT is true. Consequently, it’s simply not the case that to acquire
knowledge of PT, the functionalist must have antecedent metaphysical
knowledge of truth’s nature. Like Van, what they need is an understand-
ing of how ‘true’ applies to English sentences. They acquire this under-
standing simply in virtue of being a competent, mature English
speaker—and thus prior to embarking on their analyses of truth and
truth.31
6 Conclusions
In this chapter, I hope to have conveyed why alethic functionalism prom-
ises to be a powerful tool for theorists studying the nature of truth.
Functionalism delivers an insightful treatment of the concept-property
distinction in truth theory and in doing so, sets our sights on the con-
spicuous value of empirically studying ordinary thought about truth.
Despite considerable advances by experimental truth theorists, we’ve
managed only to break ground on this avenue of inquiry. This means that
we’ll need to expend a great deal of effort to see it to fruition—and in
doing so, we should direct a sizable portion of our effort to the evaluation
of conceptual pluralism. What awaits us as we pursue this line of inquiry
is a sharper understanding of ordinary thought about truth, the nature of
truth, and the interconnections between these topics. Whether the final
Truth in English and Elsewhere: An Empirically-Informed… 189
Notes
1. In what follows, I’ll use italics to denote properties and relations and
small caps to denote concepts.
2. Cf. Lynch (2009, p. 7).
3. For further illuminating reflections on this distinction, see Asay (2013,
Chap. 1; 2018); Bar-On and Simmons (2007); Eklund (2017, § 2); and
Lynch (2005, 2009, Chap. 1).
4. For a canonical account of Ramsification, see Lewis (1970).
5. (2009, Chap. 1). I’ve altered the wording of the sentences, though this
won’t influence the concern that I’ll develop for functionalism in § 2.
In making these alterations, I’ve been guided by the assiduous study of
Lynch’s functionalism by Marian David (2013).
6. More specifically, this metaphysics of truth amounts to what Künne
(2003, pp. 3–4, Chap. 2) calls “nihilism,” or what I prefer to call pure
deflationism. Not all deflationists (e.g. Horwich 1998) are pure defla-
tionists in this sense, but some (e.g. Quine (1948, 1970) and Strawson
(1949, 1950)) have endorsed or would endorse the view. According to
the taxonomy here, deflationists such as Horwich who commit to the
existence of a single property truth should be classified as monists. Thus
to fully flesh things out, we’d need to distinguish between deflationary
monists and substantivist monists (and between deflationary and sub-
stantivist pluralists). Doing so requires a good bit of effort; for details, see
Wyatt (2016).
7. As I’m using the expressions ‘pluralism about truth’ and ‘truth pluralism’,
one may endorse what is often called ‘alethic pluralism’ without thereby
endorsing truth pluralism. I take truth pluralism to be the view that
more than one actual property realizes FT. In other words, truth plural-
ism is the view that more than one actual property fits the job descrip-
tion for truth—full stop, and hence not merely relative to some domain
or other (on this distinction, see David (2013) and Edwards (2011,
pp. 34–7)). Extant alethic pluralists—strong and moderate alike—thus
stop short of endorsing truth pluralism. Strong alethic pluralists (e.g.
190 J. Wyatt
Cotnoir (2013) and Kim and Pedersen, this volume) take no actual
property to realize FT simpliciter and moderate alethic pluralists (e.g.
Edwards (2011) and Lynch (2009)) take exactly one such property to do
so. This, I would stress, isn’t a criticism of these views, only an observa-
tion about their structure.
It’s also worth noting that truth pluralism, in the present sense, is
structurally similar to Beall and Restall’s logical pluralism, insofar as the
properties that would realize FT would do so with respect to all domains.
8. In his most recent work (2009, 2013), Lynch has elected to not rely
explicitly on Ramsification, thereby departing from the strategy in his
earlier, pathbreaking work on functionalism (2000, 2001, 2004b, 2005).
However, I would point out that in his recent work, Lynch still speaks of
properties playing the ‘truth role’. The truth role is supplied by PT and
playing it looks to be nothing over and above witnessing RT, so it seems
that Lynch is still loyal at heart to Ramsification, even though he chooses
not to wear the badge.
9. Lynch (2009, p. 84; 2013, p. 27) and Wright (2005, n. 14) also allude
to this point (see also Devlin (2003)). However, other authors are a bit
too quick in assimilating functionalism to pluralism. I have in mind here
the otherwise illuminating critical discussions by Caputo (2012); Horton
and Poston (2012); Newhard (2013, 2014, 2017) and Wright (2005, §§
4.1, 4.2). What I would point out is that the criticisms advanced by
these authors affect the functionalist only if they commit to some sort of
alethic pluralism—a move that Lynch does make, but which is neverthe-
less entirely optional.
10. Lynch (2004a, Chap. 1; 2009, p. 8).
11. Lynch (2009, pp. 10–14; 2013, p. 24).
12. Cp. Barnard and Ulatowski (2013, 2017); Fisher, et al. (2017); Kölbel
(2008); and Mizumoto (ms). See especially Ulatowski (2017, pp. viii, ix,
2, 8–9).
13. I say ‘provisionally’ because it is at this point that we should look for
underlying similarities among the concepts that speakers deploy when
they use alethic vocabulary. If we detect such similarities, we should
then—and only then—advance a general account of what makes a con-
cept a truth concept. To do otherwise would amount to gratuitous the-
ory building in the absence of sufficient data.
14. Moltmann’s investigations (2015, 2018) into truth predicates, and
related kinds of predicate, in various natural languages also look to bear
significantly on intralinguistic pluralism. The same goes for the work of
Truth in English and Elsewhere: An Empirically-Informed… 191
Dzobo (1992, pp. 79–83) and Wiredu (1985, 1987, 2004) on the
Ghanaian languages Ewe and Akan.
15. 2013, p. 621.
16. Ibid. p. 631.
17. Ibid. p. 633.
18. Ibid. Figs. 2 and 3.
19. I include the parenthetical qualification to flag a further sort of variance
that Barnard and Ulatowski (ibid. Fig. 1) found among their subjects.
They found that their subjects (male and female alike) were more likely
overall to agree with (1) when presented with the Bruno case than when
presented with a case involving a simple arithmetical calculation. We
might call this kind of variance topic-sensitivity. What I would point out
is that this topic-sensitivity is independent of intralinguistic pluralism,
insofar as the former may be present across the community of English
speakers. For further investigation of topic-sensitivity, see Ulatowski
(2017).
20. That said, Mizumoto (ms) has gathered interesting data pertaining to
Japanese and English that do look to support interlinguistic pluralism. I
should also note the fascinating collection of papers surveyed by Maffie
(2001). By contrast, Matthewson and Glougie (forthcoming) investigate
interesting cross-linguistic uniformities in the use of alethic vocabulary.
21. I take (WC1) to be a conjecture that Wiredu would be willing to make,
although he doesn’t explicitly advance it. He comes extremely close to
doing so at 2004, pp. 47–9. Cp. 1985, pp. 47–8, 49–50; 1987, p. 28.
22. Bourget and Chalmers (2014, § 2) note that in principle, anyone was
allowed to take their survey, though the target group about which they
mainly report are the professional philosophers from the mentioned
departments.
23. Bedu-Addo argues that the best Akan expression to use is ‘nokware’, a
view with which Kwame looks to be sympathetic.
24. My thinking here has been influenced by some suggestive remarks due to
Wright (2005, pp. 18–21).
25. The grades of pluralism in (b)–(d) are notably different from the main-
stream pluralist truth theories that have been developed thus far. One
major difference is that the pluralisms in (b)–(d) make no reference to
the notion of a ‘domain’, which figures prominently in mainstream plu-
ralist theories. Rather, these pluralist views integrate Ramsification with
conceptual pluralism, an approach that hasn’t been attempted by main-
stream alethic pluralists.
192 J. Wyatt
References
Alston, W. 2002. Truth: Concept and Property. In What Is Truth? ed. Schantz,
11–26. New York/Berlin: de Gruyter.
Asay, J. 2013. The Primitivist Theory of Truth. Cambridge: Cambridge University
Press.
———. 2018. truth: A Concept Unlike Any Other. Synthese Special Issue
Truth: Concept Meets Property, Wyatt, ed. https://doi.org/10.1007/
s11229-017-1661-z.
Barnard, R., and J. Ulatowski. 2013. Truth, Correspondence, and Gender.
Review of Philosophy and Psychology 4 (4): 621–638.
———. 2017. The Objectivity of Truth, a Core Truism? Synthese Special Issue
Truth: Concept Meets Property. https://doi.org/10.1007/s11229-017-1605-7.
Bar-On, D., and K. Simmons. 2007. The Use of Force Against Deflationism:
Assertion and Truth. In Truth and Speech Acts: Studies in the Philosophy of
Language, ed. Greimann and Siegwart, 61–89. London: Routledge.
Bedu-Addo, J.T. 1985. Wiredu on Truth as Opinion and the Akan Language. In
Philosophy in Africa: Trends and Perspectives, ed. Bodunrin, 68–90. Ile-Ife:
University of Ife Press.
Bourget, D., and D. Chalmers. 2014. What Do Philosophers Believe?
Philosophical Studies 170: 465–500.
Caputo, S. 2012. Three Dilemmas for Alethic Functionalism. The Philosophical
Quarterly 62 (249): 853–861.
Cotnoir, A. 2013. Validity for Strong Pluralists. Philosophy and Phenomenological
Research 83: 563–579.
David, M. 2013. Lynch’s Functionalist Theory of Truth. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 42–68.
New York: Oxford University Press.
Devlin, J. 2003. An Argument for an Error Theory of Truth. Philosophical
Perspectives 17 (1): 51–82.
194 J. Wyatt
Dzobo, N.K. 1992. Knowledge and Truth: Ewe and Akan Conceptions. In
Person and Community: Ghanaian Philosophical Studies I, ed. Wiredu and
Gyeke, 73–84. Washington, DC: The Council for Research in Values and
Philosophy.
Edwards, D. 2011. Simplifying Alethic Pluralism. Southern Journal of Philosophy
49 (1): 28–48.
Eklund, M. 2017. What Is Deflationism About Truth? Synthese Special Issue
Truth: Concept Meets Property, Wyatt, ed. https://doi.org/10.1007/
s11229-017-1557-y.
Fisher, M., J. Knobe, B. Strickland, and F. Keil. 2017. The Influence of Social
Interaction on Intuitions of Objectivity and Subjectivity. Cognitive Science
41: 1119–1134.
Horton, M., and T. Poston. 2012. Functionalism About Truth and the
Metaphysics of Reduction. Acta Analytica 27 (1): 13–27.
Horwich, P. 1998. Truth. 2nd ed. Oxford: Oxford University Press.
Jackson, F. 1998. From Metaphysics to Ethics: A Defence of Conceptual Analysis.
Oxford: Oxford University Press.
Kölbel, M. 2008. “True” as Ambiguous. Philosophy and Phenomenological
Research 77 (2): 359–384.
Künne, W. 2003. Conceptions of Truth. Oxford: Oxford University Press.
Kwame, S. 2010. Nokware. In The Oxford Encyclopedia of African Thought, ed.
F. Abiola Irele and B. Jeyifo. New York: Oxford University Press. Accessed at
http://www.oxfordaasc.com/article/opr/t301/e275.
Lewis, D. 1966. An Argument for the Identity Theory. The Journal of Philosophy
63 (1): 17–25.
———. 1970. How to Define Theoretical Terms. The Journal of Philosophy 67:
427–446.
———. 1972. Psychophysical and Theoretical Identifications. Australasian
Journal of Philosophy 50: 249–258.
———. 1994. Reduction of Mind. In Companion to the Philosophy of Mind, ed.
Guttenplan, 412–431. Malden: Blackwell.
Lynch, M. 2000. Alethic Pluralism and the Functionalist Theory of Truth. Acta
Analytica 24: 195–214.
———. 2001. A Functionalist Theory of Truth. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 42–68. New York:
Oxford University Press.
———. 2004a. True to Life. Cambridge, MA: MIT Press.
———. 2004b. Truth and Multiple Realizability. Australasian Journal of
Philosophy 82: 384–408.
Truth in English and Elsewhere: An Empirically-Informed… 195
———. 2005. Alethic Functionalism and Our Folk Theory of Truth. Synthese
145: 29–43.
———. 2009. Truth As One and Many. Oxford: Oxford University Press.
———. 2013. Three Questions for Truth Pluralism. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 21–41.
New York: Oxford University Press.
Maffie, J. 2001. Truth from the Perspective of Comparative World Philosophy.
Social Epistemology 15 (4): 263–273.
Matthewson, L., and J. Glougie. Forthcoming. Justification and Truth: Evidence
from Languages of the World. In Epistemology for the Rest of the World, ed.
Stich, Mizumoto, and McCready. Oxford: Oxford University Press.
Mizumoto, M. (ms) A Prolegomenon to the Cross-Linguistic Study of Truth.
Moltmann, F. 2015. ‘Truth Predicates’ in Natural Language. In Unifying the
Philosophy of Truth, ed. Achourioti et al., 57–83. Dordrecht: Springer.
———. 2018. Truth Predicates, Truth Bearers, and Their Variants. Synthese
Special Issue Truth: Concept Meets Property. https://doi.org/10.1007/
s11229-018-1814-8.
Newhard, J. 2013. Four Objections to Alethic Functionalism. Journal of
Philosophical Research 38: 69–87.
———. 2014. Alethic Functionalism, Manifestation, and the Nature of Truth.
Acta Analytica 29 (3): 349–361.
———. 2017. Plain Truth and the Incoherence of Alethic Functionalism.
Synthese 194 (5): 1591–1611.
Quine, W.V.O. 1948. On What There Is. Review of Metaphysics 2 (5): 21–36.
Reprinted in From a Logical Point of View, 2nd Revised ed. Quine, 1–19.
New York: Harper and Row.
———. 1970. Philosophy of Logic. 2nd ed. Englewood Cliffs: Prentice Hall.
Strawson, P.F. 1949. Truth. Analysis 9 (6): 83–97.
———. 1950. Truth. Proceedings of the Aristotelian Society Supplementary 24:
129–156.
Tarski, A. 1944. The Semantic Conception of Truth: And the Foundations of
Semantics. Philosophy and Phenomenological Research 4 (3): 341–376.
Ulatowski, J. 2017. Commonsense Pluralism About Truth: An Empirical Defence.
Cham: Palgrave Macmillan.
Wiredu, K. 1985. The Concept of Truth in the Akan Language. In Philosophy in
Africa: Trends and Perspectives, ed. Bodunrin, 43–54. Ile-Ife: University of Ife
Press.
———. 1987. Truth: The Correspondence Theory of Judgment. African
Philosophical Inquiry 1 (1): 19–30.
196 J. Wyatt
———. 2004. Truth and an African Language. In African Philosophy: New and
Traditional Perspectives, ed. Brown, 35–50. Oxford: Oxford University Press.
Wright, C.D. 2005. On the Functionalization of Pluralist Approaches to Truth.
Synthese 145 (1): 1–28.
Wright, C. 2010. Truth, Ramsification, and the Pluralist’s Revenge. Australasian
Journal of Philosophy 88 (2): 265–283.
Wyatt, J. 2016. The Many (Yet Few) Faces of Deflationism. The Philosophical
Quarterly 66 (263): 362–382.
Part II
Logic
Core Logic: A Conspectus
Neil Tennant
1 Introduction
As the title of this study implies, there is a much longer and more detailed
treatment of Core Logic, to be found in Tennant (2017). Here we simply
try to explain and summarize the main points and their relation to debates
about monism and pluralism.
The title of the Storrs conference in 2015 of which this volume is the
proceedings was ‘Pluralism in Truth and Pluralism in Logic’. But the
prevailing intellectual climate is so pluralist that the singulars in the con-
ference title have metastasized so as to occasion, in the title of this vol-
ume, the plural noun ‘Pluralisms’. In this heady atmosphere, the present
author presents a more conservative, but still accommodating, ‘absolutist
pluralism’ about logic—albeit a reformist one. He believes there is such a
thing as ‘core logical’ deductive reasoning, and that it is relevant, in a
sense heretofore not satisfactorily explicated. It comes, however, in two
N. Tennant (*)
Department of Philosophy, Ohio State University, Columbus, OH, USA
e-mail: tennant.9@osu.edu
In the present author’s view, the decision tree for choice of logic, for an
unmitigated absolutist, looks like this1:
Yes No
Is it constructive?
Yes No Yes No
So you’re a constructive mathematician? Then the right logic for you to use
is Core Logic ℂ; or …
So you’re a non-constructive, i.e. classical mathematician? Then the right
logic for you to use is Classical Core Logic ℂ+.
Wir wollen einen Formalismus aufstellen, der möglichst genau das wirk
liche logische Schliessen bei mathematischen Beweisen wiedergibt.
(Gentzen 1934, p. 183)
The difference between the Core logician’s approach and that of Gentzen
is that the Core logician takes seriously the fact that such reasoning
(whether or not it is constructive) is always fully relevant, in the way it
proceeds from premises (say, mathematical axioms) to conclusions (say,
mathematical theorems).2 One’s logical system should furnish formal
proofs that fill in all the gaps that occur in the typically informal proofs
that mathematicians provide, even when they are being (by their own
disciplinary standards) very rigorous. Formal logicians have more
demanding disciplinary standards. Yet their formal proofs should just be
appropriately more detailed homologues of mathematicians’ informal
proofs. Note that this observation applies both to the constructive math-
ematical reasoner and to the non-constructive, or classical one.
So much for the conservativeness of the Core Logician’s formalizing
project.
202 N. Tennant
Øj y
j ®y j ®y
j Ú y Øj
y
Core Logic: A Conspectus 203
The right logic for constructivist reasoning must bear to the right logic
for non-constructivist reasoning the same sort of relation (represented by
the solid horizontal arrows below) that Intuitionistic Logic I bears to
Classical Logic C.
Likewise, the right logic for constructivist reasoning should bear to
Intuitionistic Logic I the same sort of relation (represented by the dashed
vertical arrows below) that the right logic for non-constructivist reason-
ing bears to Classical Logic C.
Our claim is that ℂ is the right logic for constructivist reasoning, and
ℂ+ is the right logic for non-constructivist reasoning.
Note that Core Logic contains the aforementioned Disjunctive
Syllogism (φ ∨ ψ, ¬φ : ψ) and Fregean truth-table respecting infer-
ences (¬φ : φ → ψ and ψ : φ → ψ); but it does not contain either one
of the two closely related Lewis paradoxes φ, ¬φ : ψ and φ, ¬φ : ¬ψ (see
Figs. 1 and 2).
I C
Fig. 1 From classical logic to core logic
204 N. Tennant
In addition to the four systems already mentioned (C, I, ℂ, and ℂ+), the
reader might wonder how Johansson’s system M of Minimal Logic fits
into the picture. Our contention is that M was intended by Johansson to
be a relevantized system of constructive reasoning. And in this project, it
failed, because, although it avoids the positive form of the first Lewis
paradox (φ, ¬φ : ψ), it nevertheless contains the negative form (φ, ¬φ : ¬ψ).
Moreover, M also fails to validate Disjunctive Syllogism. But the rel-
evant validity of Disjunctive Syllogism is unimpeachable; and it is
indispensable to mathematical reasoning (consider: b<a ∨ a≤b, b ≮ a:
a ≤ b).
Figure 2 summarizes these and other well-known differentiae among
the five systems under discussion.
In formulating the core systems ℂ and ℂ+, we take as our point of
departure, but significantly modify and thereby improve upon, Gentzen’s
systems of natural deduction (and the corresponding sequent calculi) for
I and C. The system M, as we have just stressed, is an aberration, consist-
ing as it does of sub-optimally formulated introduction and elimination
rules for the logical operators. Those rules have to be fundamentally re-
thought, and particular ones importantly modified, in order to ensure
relevance even after eschewing the infamous Absurdity Rule
Core Logic: A Conspectus 205
^
j
(i )
Øj
^
(i )
j
j Ú Øj
Dilemma (Dil):
(i ) (i )
j Øj
y y
(i )
y
ØØj
j
(i )
Øj
( CR )
^
(i )
j
(i ) (i )
j Øj
y y
( Dil ) (i )
y
(i ) (i )
j Øj
y ^
(i )
y
Here are some important points to note about the core systems:
1. Major premises for eliminations (MPEs) stand proud. That is, they do
not stand as conclusions of non-trivial proof-work above them.
2. So, all core proofs are in normal form. That is, no core proof contains
any sentence-occurrence standing both as an MPE and as the conclu-
sion of an application of any other rule, be it an elimination, or an
introduction, or a classical negation rule.3
3. Vacuous discharge of assumptions is prohibited where need be.
4. The rules of →-I and ∨-E are liberalized. That is, these rules are formu-
lated in such a way as to obviate the need (in the Gentzen-Prawitz
formulations) to resort to strategic applications of EFQ before they
can be applied.
5. The rule Ex Falso Quodlibet is banned.
6. So, all proofs are relevant—in a very exigent sense about to be
explained.
210 N. Tennant
Δ φ, Γ
Π Σ (where φ ∉ Γ and Γ may be empty)
φ θ
Proof See Tennant (2012b) for Core Logic, and see Tennant (2015a) for
the generalization to Classical Core Logic. The results can also be found
in Tennant (2017).
The main aim of any formal system of deductive logic is to be able to regi-
ment in ultimately fine-grained detail, and in a structure-respecting and
214 N. Tennant
The Core logician (of the disjunctive kind explained above) offers definite
answers to these questions. The challenge, to both absolutists and plural-
ists advancing different choices of system(s), is to show that their rival
choices come anywhere near matching the core systems on the collective
criteria implicit in the notable features just listed above.5
Notes
1. This diagram should be interpreted as involving successive choice points
in response to the questions posed in vertically descending order. Each
question is posed within the overall context created by the successively
narrowing columns containing it.
2. In general, of course, what mathematicians call lemmas and corollaries
can feature both as premises of proofs and as conclusions of proofs.
3. Only in non-core systems like I and C is there the possibility that abnor-
mality could arise by having an MPE stand also as the conclusion of an
application of EFQ.
4. Even a committed anti-realist can recognize the legitimacy of the demand
that one should be able to formalize classical deductive reasoning (on the
part of the realist) in the way described here. The philosophical disagree-
ment between the realist and the anti-realist can be set aside while such a
normative-cum-descriptive project is undertaken. Once the project is
completed, however, one can embark on that deferred debate. The follow-
ing question will then loom large: what is the underlying methodological
or metaphysical or meaning-theoretic commitment involved in acquiesc-
ing in applications of any of the strictly classical negation rules? Our own
answer is given in Tennant (1996): applications of such rules manifest
one’s realist metaphysical outlook, to the effect that the ‘litmus sentences’
Core Logic: A Conspectus 215
References
Anderson, A. R., and N. D. Belnap, Jr. 1975. Entailment: The Logic of Relevance
and Necessity I. Princeton: Princeton University Press.
Friedman, H., and R.K. Meyer. 1992. Whither Relevant Arithmetic? Journal of
Symbolic Logic 7: 824–831.
Gentzen, G. 1934. Untersuchungen über das logische schliessen. Mathematische
Zeitschrift I:176–210. In The American Philosophical Quarterly 1, Translated
by M. E. Szabo as ‘Investigations into Logical Deduction’, 228–306, 1964.
Tennant, N. 1996. The Law of Excluded Middle Is Synthetic a Priori, If Valid.
Philosophical Topics 24: 205–229.
———. 1997. The Taming of the True. Oxford: Oxford University Press.
———. 2010. Inferential Semantics. In The Force of Argument: Essays in Honor
of Timothy Smiley, ed. J. Lear and A. Oliver, 223–257. London: Routledge.
———. 2012a. Changes in Mind: An Essay on Rational Belief Revision. Oxford:
Oxford University Press.
———. 2012b. Cut for Core Logic. Review of Symbolic Logic 5 (3): 450–479.
———. 2015a. Cut for Classical Core Logic. Review of Symbolic Logic 8 (4):
236–256.
———. 2015b. The Relevance of Premises to Conclusions of Core Proofs.
Review of Symbolic Logic 8 (4): 743–784.
———. 2017. Core Logic. Oxford: Oxford University Press.
———. 2018. A Logical Theory of Truthmakers and Falsitymakers. In The
Oxford Handbook of Truth, ed. M. Glanzberg, 355–393. Oxford: Oxford
University Press.
Urquhart, A. 1984. The Undecidability of Entailment and Relevant Implication.
Journal of Symbolic Logic 49: 1059–1073.
Connective Meaning in Beall
and Restall’s Logical Pluralism
Teresa Kouri Kissel
1 Introduction
Jc Beall and Greg Restall (2006) spell out and defend a version of logical
pluralism. A feature of their particular pluralism is that corresponding
connectives in admissible logics mean the same thing. This means, for
example, that negations in classical, intuitionistic and relevant logic are
the same. This is in contrast to a “Carnapian” view, on which change in
logic is a change in the meanings of the logical terms.1
Unfortunately, Beall and Restall have not provided a conception of the
meaning of logical connectives on which this is true. I will show that, as
they describe it, negation in intuitionistic logic and negation in relevant
logic cannot mean the same thing on any reasonable interpretation of
their account of meaning. The fact of the matter is that they have not
settled just what “same meaning” amounts to, and any reasonable way of
spelling out what they say results in connectives which mean different
things.
T. Kouri Kissel (*)
Department of Philosophy and Religious Studies, Old Dominion University,
Norfolk, VA, USA
I will start by describing their pluralism, and what they take the logical
connectives to mean. I will then show that, as defined, the negation con-
nective in intuitionistic logic does not mean the same thing as the nega-
tion connective in relevant logic. Finally, I suggest that, as described, they
have no way around this problem. I close with some suggestions about
what the truth of the matter might be.
That is, he takes it that Carnap holds that variation in logic is due to
variation in connective meanings, which he represents by subscripting
the connectives. Restall here is attributing to Carnap what Shapiro (2014)
calls the “Dummett-Quine-Carnap” perspective (see, e.g., Shapiro 2014,
p. 108). This perspective is one which assumes that, if the logic has
changed, then the meanings of the connectives have changed as well.
From this perspective, then, no two distinct logics can have connectives
with exactly the same meanings.
Connective Meaning in Beall and Restall’s Logical Pluralism 219
GTT: An argument is validx if and only if in any casex in which the prem-
ises are true, so is the conclusion (Beall and Restall 2006, p. 29)
What changes when the logic changes, for Beall and Restall, is what class
of cases we are considering. Cases are, at a minimum, “‘things’ in which
claims may be true” (Beall and Restall 2006, p. 89). An admissible class
of cases is one which generates a logical consequence relation which is
necessary, normative and formal. There are at least three classes of cases,
on this view, one which generates a classical logical consequence relation,
one intuitionistic relation and one relevant relation.
Cases need to be the type of thing such that changing the cases in ques-
tion does not necessarily change the meanings of the connectives in ques-
tion. For, if changing cases resulted in changing connective meanings, we
could not accommodate the “in one language” thesis. Cases, whatever
they are, must be the things such that changing them does not change
connective meaning.
Classical logic is generated by taking cases to be Tarski models.
Intuitionistic logic is generated by taking cases to be what Beall and
Restall call “stages in constructions” (though the clauses they give for
stages are just the clauses for nodes in Kripke structures, see Beall and
Restall 2006, pp. 62–64). Finally, relevant logic is generated by taking
cases to be situations similar to those of Barwise and Perry (1983), but
allowing for inconsistent situations.
Beall and Restall’s description of Tarski models follows the usual
semantics.3 Some examples of clauses, for some model M and some valu-
ation s, are
220 T. Kouri Kissel
The intuitionistic cases, which they call stages in a construction, are nodes
in Kripke structures. A Kripke structure is a pair < W, ≤ > such that W is
a set of nodes, and ≤ is an ordering on those nodes. ≤ is reflexive and
transitive. They give as clauses (for w, u ∈ W)
Lastly, situations are “simply parts of the world” (Beall and Restall
2006, p. 49), except that situations can be inconsistent, and parts of pos-
sible worlds cannot be. There is an important relationship between situa-
tions, called compatibility. Roughly, two situations are compatible if they
do not “contradict” one another.4 Compatibility, according to Restall
(1999), is probably non-reflexive, symmetric and such that every situa-
tion is compatible with something. Some examples of clauses on this
semantics, for s and t situations and C the compatibility relation, are
The clauses can… [all] be true of one and the same [connective] simply in
virtue of being incomplete claims about… [the connective]. What is
required is that such incomplete claims do not conflict, but the clauses gov-
erning negation do not conflict. The classical clause gives us an account of
when a negation is true in a [model]…, and the constructive clause gives us
an account of when a negation is true in a construction. Each clause picks
out a different feature of negation. (Beall and Restall 2006, p. 98)
Connective Meaning in Beall and Restall’s Logical Pluralism 221
This means that, somehow, there is one thing which is (for example)
negation, and the various negation clauses given above all express differ-
ent aspects of that one thing.
To make sense of this claim, we need to fill in more details. I take it
that Beall and Restall (or, at least Restall) want to take the meanings of
the connectives on this picture to be maximal or minimal truth condi-
tions, as described in Restall (2002). The maximal truth condition for a
connective is something like the disjunction of all the clauses for the con-
nectives in the admissible cases. That is, the maximal truth condition for
negation (assuming the only admissible logics are classical, intuitionistic
and relevant) would be something like
and I will argue that at least one of those features is being able to be made
to fit a certain type of minimal truth condition.
If this is acceptable, then we need to ensure that there is no conflict
among clauses where they overlap. This is because Restall (2002) holds
that “the two accounts [classical and relevant] agree where they overlap”
(p. 438), which seems to accord with the claim above that incomplete
claims do not conflict (from Beall and Restall 2006, p. 98). There are two
questions we need to ask now: what, exactly, does it mean for two clauses
to overlap? And do the intuitionistic and relevant clauses agree where
they overlap? I think the answer to the second question is no, but it
depends in large part on our answer to the first. It is to this which I now
turn.
3 Overlapping Clauses
The method that Beall and Restall use to explain overlap involves describ-
ing all three consequence relations in a situation semantics (Restall (1999)
and Beall and Restall (2001)). They state
We… provide a model of how these things could be… to show how our
three different claims about the behavior of negation can be true together…
A world [classical model] is a complete, consistent situation… [We] take
all constructions to be situations… [that are] not only consistent, but
explicitly so. (Beall and Restall 2001, pp. 10–11)
Though this is ultimately the downfall of the view, we need the details to
explain the problem.
This method of explaining overlap involves restricting the clauses for
the connectives in the situation semantics in certain ways to generate
the clauses for the connectives in the other logics. At the very least, the
road is simple enough for finding a classical consequence relation in
such a semantics. We simply restrict ourselves to considering complete
consistent situations, and restrict compatibility to be the identity
relation.
Connective Meaning in Beall and Restall’s Logical Pluralism 223
This is something like the minimal truth condition alluded to above. Let’s
look at precisely how this will work. First, we take the default to be that
D is the entire situation space, and that R is compatibility (we will be
forced to loosen these requirements later). This means that the default of
(*) generates a relevant negation clause. Classical negation can then be
recovered by letting D range over complete, consistent situations, and
letting R be the identity relation. So far so good.
Recall that Beall and Restall (2001) claim that the intuitionistic clause
can be recovered by restricting our attending to explicitly consistent situ-
ations. In the language of (*), this means letting D range over explicitly
consistent situations, and letting R be such that, if c is a stage, and cRs,
then sRs, so that stages only “see” consistent situations.
We are now in a position to articulate a definition for when two clauses
overlap. Two clauses overlap when the intersection of their domains is
non-empty. I think Beall and Restall would agree up to this point. We
can also see now that all three clauses overlap, since the intersection of
situations with complete consistent situations and/or with explicitly con-
sistent situations will not be empty. This means what is left to determine
whether all three negations have the same meaning is to assess whether or
not they conflict where such overlap occurs.7
224 T. Kouri Kissel
4 The Problem
I will focus on showing that the intuitionistic clause conflicts with the
relevant clause for negation. I will demonstrate this by showing the addi-
tion of the intuitionistic clause to a maximal truth condition which
already has the relevant clause as one of its disjuncts makes negation arbi-
trary. If we add the intuitionistic clause to the maximal truth condition
which already has the relevant clause, then we will be able to add clauses
to the maximal truth condition which we ought not to consider part of
the meaning of negation. I take it, showing that the relevant negation
becomes arbitrary upon the addition of the intuitionistic clause demon-
strates that the clauses conflict. So, both clauses cannot be part of the
meaning of negation.
In order to assess whether the intuitionistic clause for negation con-
flicts with the relevant clause, we need to settle just what stages-as-
situations are. Beall and Restall claim that they are merely explicitly
consistent situations. If stages are merely explicitly consistent situations,
then changes in the relationship R in (*) need to be made. I call this horn
“Option 1”. If they are not explicitly consistent situations, then there will
be no overlap between the intuitionistic negation clause and the relevant
negation clause. I call this horn “Option 2”. I will consider each option
in turn, and show that both options make negation arbitrary, and thus no
matter how Beall and Restall go, they will not be able to succeed in their
connective meaning project.
Suppose first that stages are merely explicitly consistent situations. To
do this, we need to make sure that the intuitionistic clause for negation
can appropriately fit (*) when D ranges over explicitly consistent situa-
tions. This means that, in addition to R being restricted in such a way
that stages are only compatible with consistent situations, we also need to
insist that R not be symmetric. Were the parallel relation to be symmet-
ric, we would lose any notion of constructability. The ≤ relation is sup-
posed to be something like a “constructed-from” or “built from” relation,
so that we have w ≤ v just in case v is “built from” w. Suppose that the
relationship between explicitly consistent situations (the “built-from”
relationship) was indeed symmetric. Then anytime v was “built from” w,
Connective Meaning in Beall and Restall’s Logical Pluralism 225
the previous problem: nodes will only be related to the nodes in the
future of their own Kripke structure, and the relationship will not be
symmetric.
There will be no overlap between the clause for negation in situations
and the clause for negation in nodes of Kripke structures. The intersec-
tion of the domains will be empty, since one contains only situations, and
the other contains no situations.
Again, though, we have a similar issue. If all it takes to be part of the
meaning of negation is to be “moldable” into (*) (with no requirements
on the domain) then there is a problem. Again, there are lots of clauses
which would fit this mold, and it is not the case that we would like to
allow all of them to be part of the meaning of negation.
Consider the following. Let us assume that the situation semantics is
rich enough such that every atomic sentence is not true at some situation.
Then, for any list of negated atomic sentences, we can construct a
situation-in-a-model at which they will all be true.10 We do so simply by
restricting our domain in certain ways. Suppose, for example, that the list
we were given contained ¬A1,…,¬An. Then, we simply restrict our domain
to the set/class of situations which make none of A1,…,An true. This will
be non-empty, since because of our assumption, there is at least one situ-
ation which does not make true the conjunction of A1,…,An.11 Further, at
any situation in this domain, the negations of A1,…,An are true, since the
compatibility relationship will be restricted to our domain, and no situa-
tion in the domain makes A1,…,An true. In this case, negation is arbi-
trary. We can make the negation of any sentences we would like true, just
by restricting the domain in certain ways. Thus, allowing restrictions to
the domain in order to get the intuitionistic clause for negation requires
allowing restrictions that gets arbitrary clauses for negation. I claim that
Beall and Restall cannot allow this and still claim that they have suc-
ceeded in pinning down the meaning of negation.
Summing up, we see that if stages are either explicitly consistent situa-
tions with the accessibility relation being a subrelation of compatibility,
or if they cannot be situations at all, negation as described by Beall and
Restall will be arbitrary if it includes both the intuitionistic and relevant
clause. Given the way Beall and Restall discuss stages, and intuitionistic
cases, these seem to be the only options. So, either negation is arbitrary
Connective Meaning in Beall and Restall’s Logical Pluralism 227
such that Thomas’s N relationship (the equivalent of our “R”, above) con-
tains exactly < ω, α > and < α, α >, and then let the other relationships
in the Thomas system (the range of each quantifier, and the tripartite
relationship which is used to make sense of the arrow) be identity rela-
tions. Then, supposing all objects are P at ω, but that we introduce some
new object, a, at α such that ¬Pa, then we will have both ω ⊨ ∀xPx and
ω ⊨ ¬∀xPx. This in effect means that negation will not be compositional,
since φ ⊨ ¬¬φ holds when φ is quantifier free, but does not hold when it
is not. This seems to be allowing the quantifiers to influence the meaning
of negation. Though this is not an issue for Thomas, it poses a problem for
Beall and Restall. A non-compositional negation would not match natural
language negation, and since Beall and Restall motivate their position in
part by the applicability of their various logical systems, the connectives in
their system must at least partly match the natural language connectives.
Since natural language negation is arguably compositional, a non-compo-
sitional negation will not do the trick, and should not be admitted as part
of the meaning of negation. Moreover, Thomas proves that all logics
obtained by this system are subclassical, so this logic will be necessary, nor-
mative and formal as desired.15 Again, we see that if we try to use (*) to
capture the meaning of negation, even in this new system, we will wind up
letting things in which ought not to count as negation.
s peakers actually use. If they want their theory to be applicable, then the
connectives cannot be merely stipulative. So they cannot just define
negation technically, it has to match up with something in the world.
7 A Possible Solution
There is an interesting moral to draw here. There does seem to be some-
thing fundamentally different about intuitionistic negation and relevant
negation. I strongly suspect this has something to do with the fact that
intuitionistic and relevant consequences are both restrictions of classical
consequence in different ways. Intuitionistic logic allows explosion but
not double negation elimination, while relevant logic allows double nega-
tion elimination but not explosion. However, Beall and Restall have pro-
vided a meaning for the connectives on which classical and relevant
negations mean the same thing. We can also do this for classical and
intuitionistic negations. Here, we would use nodes in Kripke models as
intuitionistic cases, and recover classical consequence, and the classical
connective clauses, by restricting ourselves to one-node Kripke structures
(see Beall and Restall 2006, p. 98). If both of these systems are legitimate,
then we have a pluralism which is pluralistic in both logic and language.
We have two model theories, that of the situations semantics and that of
the Kripke structures, and two logics in each, classical and relevant in the
first, and classical and intuitionistic in the second. On the one hand, we
have what Beall and Restall want: two consequence relations with con-
nectives which share a meaning, if sameness of meaning of the connec-
tives in two logics amounts to being able to use the same model theory
for both logics. On the other hand, we have what our earlier characteriza-
tion of Carnap wants: sometimes, logic change does require connective
meaning change. In a sense, this type of system might be thought of as a
synthesis between Carnap and Beall and Restall.
As a sketch of how it might go, consider the following.16 We might
think that the meanings of the connectives are context sensitive in a cer-
tain way; for our purposes, the meanings are sensitive to the goals of
certain uses of logic. That is, the meanings are sensitive to what the logic
in question is being used to do (e.g., formalize a mathematical system,
230 T. Kouri Kissel
8 Conclusion
Beall and Restall (2006) propose a logical pluralism where the connec-
tives for each logic are pairwise synonymous. Due to the manner in which
connectives are given their meaning, relevant negation and intuitionistic
negation cannot mean the same thing. Thus, their pluralism is a plural-
ism of languages and logics, not just logics as desired.17
Notes
1. I argue elsewhere (see Kouri (2018)) that this is not the best way to
understand the Carnapian picture. For the purposes of this chapter,
though, it suffices.
2. It is not clear that Restall is right in such an assessment of Carnap. At the
very least, Restall ought to grant that Carnap’s pluralism is a pluralism of
Connective Meaning in Beall and Restall’s Logical Pluralism 231
logics and languages. Restall’s position, on the other hand, is just a plu-
ralism of logics. Additionally, it is the case that Carnap could claim “A
and ¬A together, classically entail B, but A and ¬A together do not rele-
vantly entail B”. He would claim this is the case precisely because the
connectives mean something different in each logic. In some sense, what
this quote does for Restall is to show that Restall disagrees with the sec-
ond half (“A together with its classical negation entail B, but A together
with its relevant negation need not entail B”.) and not that Carnap dis-
agrees with the first. Thanks to Hannes Leitgeb for pushing me on this
issue.
3. There is a reason to think that the metalanguage Beall and Restall are
using is also classical. Though they never explicitly make a claim about
which metalanguage they are working with, it seems their description of
stages in constructions and situations as subclassical suggests that classi-
cal logic is the strongest logic admissible on their picture, and their rejec-
tion of contra-classical logics suggests the same. If this is the case, then it
seems that it would make a good choice for a metalanguage. For a criti-
cism of their (probable) choice in metalanguage, see Read (2006). It
should be pointed out here that Beall and Restall never explicitly make
the claim that their metalanguage is classical. A referee has suggested to
me that in personal correspondence, Restall has gone so far as to claim
that there is no metalanguage associated with logical pluralism. I am not
entirely sure what to make of this, as presumably, we need some language
in which to discuss the logics which are admissible. Whether the meta-
language is classical, though, or even if there is no metalanguage, most of
what follows can be re-stated accordingly, and so I will not pursue this
complication further here.
4. Unfortunately, very little about the compatibility relation is committed
to by Restall. This poses some problems for Beall and Restall’s project. In
particular, it will turn out that it is important whether compatibility is
primitive, or somehow determined by negation.
5. There is another way we can view this (if Beall and Restall’s metalan-
guage is classical, these will be equivalent). It might be rather than a
disjunction, we have a conjunction of conditionals. The maximal truth
condition for negation would then be
(if M is a Tarski model then M, s ⊨ ¬A if and only if it is not the case
that M, s ⊨ A) AND (if w is a node in a Kripke structure then w ⊨
¬A if and only if, for all u such that w ≤ u, u ¬ A) AND (if s is a situ-
ation then s ⊨ ¬A if and only if, ∀t such that sCt, t ¬ A)
232 T. Kouri Kissel
15. They will be necessary, normative and formal on the basis of Beall and
Restall’s definitions. Formality here seems relatively simple. Necessity
and normativity, on the other hand, are a bit odd, and this is the result
of Beall and Restall’s definitions of them. For Beall and Restall, any
sublogic of classical logic will be necessary, since they hold that relevant
logic is necessary because worlds are a special type of situation. But here,
since, on Thomas’s system, classical logic can be recovered as a special
type of model, and since no model we are concerned with proves any-
thing contra-classical, all models ought to be truth preserving, and so
necessary on Beall and Restall’s picture. This reasoning mimics their
discussion of relevant logic, where they say “The question is, in other
words, whether truth-preservation over all situations guarantees truth-
preservation over all possible worlds. The answer is ‘yes’ if possible
worlds count as (perhaps special) situations” (Beall and Restall 2006,
p. 54). Here, we have swapped “models” for situations, and “classical
models” for possible worlds, but the reasoning is similar. Normativity,
for Beall and Restall, comes down to there being some sort of mistake
associated with reasoning invalidly on this type of logic. But they are
not terribly specific about what types of things count as mistakes. The
mistake associated with relevant logic is irrelevant reasoning, and the
mistakes associated with constructive logic are “mistakes of constructiv-
ity” (Beall and Restall 2006, 70). In this case, then, I claim that not
reasoning properly according to this new logic is reasoning composi-
tionally when one shouldn’t. As one referee has suggested, there is a
good reason for denying that all subclassical logics are normative and
necessary. Though this seems plausible (and many commentators agree
that Beall and Restall’s definitions of necessary and normative are prob-
lematic, see, e.g., Bueno and Shalkowski (2009) and Keefe (2013)),
strictly speaking here all we only need is the Thomas system to match
Beall and Restall’s definitions, and it seems to. On the other hand, if the
Thomas system does not produce necessary, normative and formal log-
ics, then it is not an option for Beall and Restall to begin with.
16. For more details, see Kouri Kissel (2018) and Shapiro (2014).
17. Acknowledgments: I would like to thank Stewart Shapiro, Shay Logan,
Chris Pincock, Kevin Scharp, Neil Tennant and a referee for this volume
for feedback on earlier drafts. Additionally, audiences at the 2016
Central APA, the UConn Conference on Logical Pluralism and Truth
Pluralism, the Munich Center for Mathematical Philosophy and the
2017 Canadian Philosophical Association meeting provided incredibly
helpful comments.
Connective Meaning in Beall and Restall’s Logical Pluralism 235
References
Barwise, J., and J. Perry. 1983. Situations and Attitudes. Vol. 78. Cambridge,
MA: MIT Press.
Beall, J.C., and Greg Restall. 2001. Defending Logical Pluralism. In Logical
Consequence: Rival Approaches. Proceedings of the 1999 Conference of the Society
of Exact Philosophy, ed. John Woods and Bryson Brown, 1–22. Stanmore:
Hermes.
———. 2006. Logical Pluralism. Oxford: Clarendon Press.
Bueno, O., and S.A. Shalkowski. 2009. Modalism and Logical Pluralism. Mind
118 (470): 295–321.
Dunn, J.M. 1993. Star and perp: Two treatments of negation. Philosophical
Perspectives 7 (May): 331–357.
Hjortland, O.T. 2013. Logical Pluralism, Meaning-Variance and Verbal
Disputes. Australasian Journal of Philosophy 91: 355–373.
Keefe, R. 2013. What Logical Pluralism Cannot Be. Synthese 191 (7):
1375–1390.
Kouri, T. 2018. A New Interpretation of Carnap’s Logical Pluralism. Topoi:
1–10. Online first.
Kouri Kissel, T. 2018. Logical Pluralism from a Pragmatic Perspective.
Australasian Journal of Philosophy: 96 (3): 578–591.
Read, S. 2006. Review of Logical Pluralism Notre Dame Philosophical Reviews.
https://ndpr.nd.edu/news/logical-pluralism/.
Restall, G. 1999. Negation in Relevant Logics (How I Stopped Worrying and
Learned to Love the Routley Star). In What Is Negation?, ed. H. Wansing and
D. Gabbay, Volume 13 of Applied Logic Series, 53–76. Dordrecht: Kluwer
Academic Publishers.
———. 2002. Carnap’s Tolerance, Language Change and Logical Pluralism.
Journal of Philosophy 99: 426–443.
Shapiro, S. 2014. Varieties of Logic. Oxford: OUP.
Thomas, M. 2015. A Generalization of the Routley-Meyer Semantic Framework.
Journal of Philosophical Logic 44 (4): 411–427.
Generalised Tarski’s Thesis Hits
Substructure
Elia Zardini
Earlier versions of the material in this chapter have been presented in 2015 at the Veritas
Pluralism, Language and Logic Workshop (Yonsei University); in 2018, at the LanCog Seminar
(University of Lisbon), at the LOGOS Workshop Pluralism and Substructural Logics (University
of Barcelona), at the fifth SBFA Conference (Federal University of Bahia) and at the Workshop
Disagreement within Philosophy (Rhine Friedrich-Wilhelm University of Bonn). I’d like to thank
all these audiences for the very stimulating comments and discussions. Special thanks go
to Agustín Rayo, Colin Caret, Bogdan Dicher, Catarina Dutilh Novaes, Luís Estevinha, Filippo
Ferrari, Ole Hjortland, Luca Incurvati, José Martínez, Ricardo Miguel, Sergi Oms, Nikolaj J. L.
L. Pedersen, Hili Razinsky, Lucas Rosenblatt, Sven Rosenkranz, Diogo Santos, Ricardo Santos,
Erik Stei, Célia Teixeira, Pilar Terrés, Zach Weber, Jack Woods and Jeremy Wyatt. I’m also
grateful to the editors Nathan Kellen, Nikolaj J. L. L. Pedersen and Jeremy Wyatt for inviting me
to contribute to this volume and for their support and patience throughout the process. The
study has been funded by the FCT Research Fellowship IF/01202/2013 Tolerance and Instability:
The Substructure of Cognitions, Transitions and Collections. Additionally, the study has been funded
by the Russian Academic Excellence Project 5-100. I’ve also benefited from support from the
Project FFI2012-35026 of the Spanish Ministry of Economy and Competition The Makings of
Truth: Nature, Extent, and Applications of Truthmaking, from the Project FFI2015-70707-P of the
Spanish Ministry of Economy, Industry and Competitiveness Localism and Globalism in Logic and
Semantics and from the FCT Project PTDC/FER-FIL/28442/2017 Companion to Analytic
Philosophy 2.
E. Zardini (*)
LanCog Research Group, Philosophy Centre, University of Lisbon,
Lisbon, Portugal
International Laboratory for Logic, Linguistics and Formal Philosophy, School
of Philosophy, National Research University Higher School of Economics,
Moscow, Russian Federation
1 Starting Off
Generalised Tarski’s Thesis
considerations would seem to show not only that one ought to reject that
all one’s scientific beliefs are true, but also that one is under no pressure at
all to believe the extremely strong, unlikely, unreliable, arrogant—and
indeed rather silly—proposition that all one’s scientific beliefs are true—
there is no dilemma about believing that not all one’s scientific beliefs are
true, that is just the unambiguously right thing to believe.6
As for the formality condition, there would seem to be a logic of, say,
metaphysical necessity as a particular necessity distinct from, say, epistemic
necessity and governed by principles different from those governing other
necessities (there is no interesting logic of a thin, formal ‘necessity’; it is only
thick, material necessities that have interesting logics). However, I know of
no reasonable sense in which the notion of metaphysical necessity is for-
mal [see Zardini (2018c) for elaboration and variations]. It’s as simple as
that. Having noted all this about BRT2, in the following (really, from the
last paragraph of this section onwards), I’ll rather focus on BRT1.
I’m very sympathetic to logical pluralism as the general view roughly
stated in the fourth last paragraph. To take a natural example, I think that
classical analysis is legitimate (and true)—and so, a fortiori, that its accom-
panying classical logic is legitimate—and that intuitionist analysis [in the
style of e.g. Brouwer (1927)] is also legitimate (and true), and so, a for-
tiori, that its accompanying intuitionist logic is legitimate. Their apparent
incompatibility can quite satisfactorily be explained away by the very
plausible hypothesis that they describe different mathematical structures.7 It
is true that some members of either party make extremely bold claims to
the effect that, as a matter of principle, the mathematical structures pos-
tulated by the other party cannot exist [e.g. Dummett (2000), 250–269].
But I find such claims very implausible and only backed up by very weak
arguments, so that I think that it can safely be concluded that such mem-
bers are plain wrong (there’s room for everyone in Plato’s heaven!). Here
as elsewhere, first-order tolerance requires a modicum of second-order
intolerance. In effect, I thus accept the ‘regionalist’ thesis [already sug-
gested by Putnam (1968)] that different logics apply to different domains. I
submit that the opposite idea that the same logic applies to every logical
domain gets whatever spurious plausibility it has from comparison with
the much more plausible idea that the same physics applies to every physical
domain. But, while physical domains would seem connected and homoge-
240 E. Zardini
neous enough to make the latter thesis rather likely, the whole range of
logical domains—going from the most exotic physical particles through
baldness and truth to the most esoteric mathematical structures—are so
completely disconnected and highly heterogeneous as to make the former
thesis rather unlikely. Given the bewildering variety of logical domains, it
is just plausible to expect that a single logical operation can obey a certain
principle on a certain domain and not obey it on another domain.
I’m not quite sympathetic to logical pluralism as typically understood in
the contemporary debate—where BR have been a major player—for, on
that understanding, even when taking a particular argument and even
when choosing a particular ‘level of formality’, as when, for example, it is
debated in philosophy of logic [e.g. Beall (2007)] whether the particular
argument from ‘The Liar sentence is true iff it is not’ to ‘The Earth is flat’
is valid at the level of the logic of truth, it is supposed to be legitimate to
hold that the argument is valid and also legitimate to hold that it is not
valid, and so, given that in this case presumably it is legitimate to accept
the premise of the argument if (and only if ) it is legitimate to hold that
the argument is not valid,8 it is in effect supposed to be legit to accept the
problematic instances of the T-schema and also legit to reject them! In
addition to undermining the significance of what is in fact one of the most
significant debates in philosophy of logic, pending first arguments such
specific claim is in itself so helplessly relativist that it would seem it ought
to lead to the rejection of any general view entailing9 it.
I’m even less sympathetic to BRT (in particular, to its ‘only’-component).
For, quite generally, BRT is yet another example of trying to impose a for-
mat to which every instance of a kind has to conform (in this case, every
instance of the kind Legitimate Relation of Logical Consequence). Besides
the telling considerations of Wittgenstein (1953), §66 (which would not
seem to have adequately been heeded in much subsequent analytic philoso-
phy), I think that the track record of such attempts in the history of thought
in general and in philosophy in particular (and even more in particular
about such philosophically central kinds as Truth, Knowledge, Cause, etc.)
is extraordinarily negative. And I don’t see any reason for thinking that the
specific case of the kind Legitimate Relation of Logical Consequence will
deviate from this general tendency. Indeed, in the following I’ll corroborate
this point by discussing a particular area that systematically violates GTT.
The discussion is ultimately meant to achieve something more interesting
Generalised Tarski’s Thesis Hits Substructure 241
Substructure
The area that, in section “Generalised Tarski’s Thesis”, has been referred to
as systematically violating GTT is the one of substructural logics [see Restall
(2000); Paoli (2002) for a couple of general, philosophically sensitive sur-
veys]. To recall, operational properties are those properties of a logic which
concern particular logical operations (for example, reductio ad absurdum is
an operational property in that it concerns negation). By contrast, struc-
tural properties are those properties of a logic which concern general fea-
tures that only depend on considering premises and conclusions as a
manipulable structure of unstructured objects (for example, reflexivity is a
structural property in that it only depends on the premise and conclusion
being the same object). Classical logic and many non-classical logics (for
example, intuitionist logic) have a series of noteworthy structural proper-
ties: reflexivity, monotonicity, transitivity, contraction, commutativity and
others. A logic is substructural iff it lacks one of the properties in that series.
I’ll argue then that GTT systematically rules out all kinds of philosophically
interesting substructural logics as legitimate relations of logical consequence.
Given some of my work, I actually have a particular emotional attach-
ment to this exclusion from the realm of ‘logical consequence’ :-( But, as
in many other cases, it’s really not that interesting to determine exactly to
which logics the label ‘legitimate relation of logical consequence’ applies
(either in virtue of the vagaries of the use of that expression or in virtue of
a sheer stipulation).11 What’s interesting is what kind is singled out by
GTT, and how central it is for philosophical theorising about logic. In that
respect, my cumulative argument will show that the kind singled out by
GTT is not central for philosophical theorising about logic. And, as has
already been foreshadowed, there will also be other consequences for
broadly Tarski-inspired theses.
242 E. Zardini
2 The Crash
Reflexivity
(I) φ ⊢ φ holds.
There are philosophical reasons to doubt (I) [Moruzzi and Zardini (2007),
pp. 180–182].12 Firstly, (I) refers to circular arguments—so circular that
they might be thought to violate epistemic and metaphysical principles
governing logical consequence. Epistemically, one might try to forge some
connection between validity and transmission [Wright (2000)] and think
that a valid argument must be such that one can use it in at least some
context to acquire a new justification for believing its conclusion13 [see
Martin and Meyer (1982) for very broadly similar considerations].14 But
in no context can one acquire a new justification for believing φ by infer-
ring it by (I) from φ.15 Metaphysically, one might try to forge some con-
nection between Ableitbarkeit and Abfolge [Bolzano (1837)] and think
that a valid argument must be such that, if its premises are true in virtue
of non-logical facts, its conclusion must be true (also) simply in virtue of
the truth of the premises.16 But x-is-true-in-virtue-of-the-truth-of-y is non-
reflexive (i.e. such that, for some x, it is not the case that x is true in virtue
of the truth of x), even restricting to sentences that are true in virtue of
non-logical facts.17 Secondly, notice that (I) is an atypical structural prop-
erty, in that it licences the validity of rules (things of the form ‘Γ ⊢ φ
holds’) rather than metarules (things of the form ‘If Γ0 ⊢ φ0, Γ1 ⊢ φ1, Γ2
⊢ φ2 … hold, ∆ ⊢ ψ holds’). That would seem already in itself problem-
atic: if at least almost all allegedly valid structural properties licence meta-
rules rather than rules, that is good evidence that all allegedly valid
structural properties licence metarules rather than rules. And it becomes
even more problematic once it is realised that, for this reason, (I) violates
the plausible principle that an argument is valid in virtue of some of the
semantic properties of some expressions occurring in its premises or conclu-
sion—principle that is entailed by some understandings of the plausible
principle that argument validity is a matter of analyticity, for example by
the understanding that an argument’s validity consists in the fact that its
Generalised Tarski’s Thesis Hits Substructure 243
Monotonicity
the pill by saying that LEM is a necessary truth. But that’s not good
enough. One problem with BR’s position is that many relevant logics do
count LEM as a logical truth: for example, the relevant logic BX is defined
in terms of adding LEM as a logical truth21 to the basic relevant logic B
(analogously, a few other relevant logics …X are defined in terms of add-
ing LEM as a logical truth to the relevant logic …), and the even more
influential relevant logic R can be—and often is—partially characterised
in terms of having LEM as a logical truth. Be that as it may with LEM in
particular, the same issue arises with almost any alleged logical truth (save for
absolutely trivial things like ‘For some P, P’). Since the whole hierarchy
of relevant logics as standardly conceived relies on such alleged logical
truths being logical truths in the pertinent logics, and since anyways I’d
like to think that there are relevant logical truths in addition to absolutely
trivial things like ‘For some P, P ’, I deem BR’s position untenable (even
without going into the fact that it does not even speak to the case where
(K) is applied to an argument with premise and conclusion). They should
either, pace philosophical common sense (and their own choice of wit-
ness), deny that relevant logics are legitimate relations of logical conse-
quence or, pace GTT, deny (K).
Transitivity
That is all very exciting stuff, but enter now GTT. If, in every casex
where every element of Γ is true, so is φ, then, in every casex where every
element of ∆, Γ is true, so is every element of ∆, φ. Therefore, if, in every
casex where every element of ∆, φ is true, so is ψ, then, in every casex
where every element of ∆, Γ is true, so is ψ. That holds no matter what x
is, and so non-transitive logics cannot drop out of GTT, and hence, by
BRT1, non-transitive logics are not legitimate relations of logical conse-
quence. That’s a very unhelpful ruling!
Contraction
Commutativity
There are philosophical reasons to doubt (C). Firstly, there might be con-
text change between an occurrence of a premise and another occurrence of
a premise [e.g. Zardini (2014a)]. For example, if each next occurrence
occurs on the day after the day on which the previous occurrence occurs,
〈‘Tomorrow it’ll rain’, ‘It’s not raining’〉 is naturally regarded as inconsis-
tent (and so, let’s assume, as entailing everything), but 〈‘It’s not raining’,
‘Tomorrow it’ll rain’〉 isn’t.23 Secondly, (C) forecloses an attractive account
of what it is for an object to be changing from one state to its opposite
[Zardini (2018d)].
That is all very exciting stuff, but enter now GTT. In every casex where
every element of Γ, ψ, φ, ∆ is true, so is every element of Γ, φ, ψ, ∆.
Therefore, if, in every casex where every element of Γ, φ, ψ, ∆ is true, so
is χ, then, in every casex where every element of Γ, ψ, φ, ∆ is true, so is χ.
That holds no matter what x is, and so non-commutative logics cannot
drop out of GTT, and hence, by BRT1, non-commutative logics are not
legitimate relations of logical consequence. That’s a very unhelpful
ruling!
3 Aftermath
Unireliabilism
In all the cases we’ve considered in Sect. 2, GTT fails in that its satisfac-
tion by a logic in effect requires that the logic capture the fact that, for
every model in a certain class, it is not the case that the premises are true in
the model and the conclusion is not. Ascending to a more abstract perspec-
tive, GTT fails in those cases arguably at bottom because it is thus a spe-
cies of a more general kind of occurrence-insensitive, single-value,
single-model reliabilism: ‘reliabilism’ because it only considers whether the
248 E. Zardini
conclusion has the relevant value when the series of premises do (rather
than considering stronger connections between the conclusion having the
relevant value and the series of premises doing); ‘occurrence-insensitive’
because it only considers which sentences the premises are (rather than
considering, for example, how many times and in what order they occur);
‘single-value’ because it only considers one value relevant for all the series
of premises and the conclusion (rather than considering different values
relevant for different occurrences of the premises or the conclusion); ‘sin-
gle-model’ because, on each test,24 it only considers one model for all
the series of premises and the conclusion (rather than considering different
models for different occurrences of the premises or the conclusion). Since
such reliabilism is hostile to plurality in occurrences, values and models in test-
ing for logical consequence, it may well be labelled for short ‘unireliabilism’.
The issue surfaces slightly when BR note (p. 91) that, because of its
‘preservationism’, GTT rules out non-reflexive and non-transitive logics.
But that’s just the tip of the substructural iceberg—as we’ve seen in Sect.
2, GTT systematically rules out all kinds of philosophically interesting
substructural logics! Quite generally, GTT clashes with substructurality.
And the culprit is not some vague, benign and acceptable ‘preservation-
ism’—as we’ve seen in the last paragraph, GTT clashes with substructur-
ality because it is a species of the much more precise, malign and
questionable unireliabilism, which, obliterating as it does plurality in
occurrences, values and models, as well as intensional connections between
premises and conclusion, makes any version of ‘pluralism’ based on it barely
worth its name. Indeed, once GTT’s unireliabilism has been identified as
the culprit for its clash with substructurality, it becomes clear that, just as
GTT generalises in a very natural way the classical idea that an argument
is valid iff, for every classical model, it is not the case that the premises are
true in the model and the conclusion is not (by relaxing its ‘classical-
model’-component), one can generalise further in equally natural ways uni-
reliabilism itself (by relaxing some of its components). And at least the
specific substructural logics alluded to in Sect. 2—rather than represent-
ing conceptions of logical consequence totally alien to GTT—can actu-
ally be thought of as instances of such further generalisation. To wit,
non-reflexive logics of the kinds alluded to in Sect. 2 can be thought of as
requiring that the value of the conclusion be better than the values of the
Generalised Tarski’s Thesis Hits Substructure 249
might be governed by too weak principles of the logic. Not every interesting
logic has been developed with the purposes of doing syntax or
semantics.
Moreover, picking up from the second disjunct of the second last sen-
tence of the last paragraph, it’s completely up for grabs which logics would
satisfy the logic-relative version of GTT. Indeed, it is likely that quite a
few interesting logics will not satisfy its logic-relative version, and we in
fact know that even some of those that did not create problems for GTT
will no longer satisfy its logic-relative version. For example, the Bočvar
logic B3 [Bočvar (1938)] and the Kleene logic K3 [Kleene (1938)] will not
satisfy the logic-relative version of GTT, since they lack a suitably weak
implication vindicating claims—so crucial to GTT—along the lines of
‘Every F is G’ (for example, ‘In every casex where φ is true, so is φ’). In
fact, we also know that some interesting logics that did create problems
for GTT will still not satisfy its logic-relative version. For example, non-
contractive logics will not satisfy the logic-relative version of GTT, since,
letting ⊗ be a multiplicative conjunction, in typical non-contractive log-
ics φ ⊢ φ ⊗ φ does not hold, but, in the lattice-theoretic semantics asso-
ciated with typical non-contractive logics [e.g. Ono (2003)], ⊗ is
idempotent on designated value (i.e., if φ gets designated value, φ ⊗ φ gets
the same value), so that ‘If φ gets designated value in a certain casex, so
does φ ⊗ φ’ and hence—since, presumably, φ is true in a case iff φ gets
designated value in that case—‘If φ is true in a certain casex, so is φ ⊗ φ’
hold (and presumably do so even when using a non-contractive logic in
the metatheory). GTT’s rulings were definitely unhelpful, but those of its
logic-relative version are likely to be helpless.
Furthermore, while the logic-relative version of GTT is designed to
preserve its letter, it arguably abandons its spirit. For GTT’s spirit argu-
ably incorporated unireliabilism, but the new substructural interpreta-
tions of its universal quantification and implication deprive GTT of one
component or other of unireliabilism. Indeed, there is barely anything sub-
stantial in common at all, unireliabilist or otherwise, between, say, the sense
in which a non-reflexive logic thinks it satisfies GTT and the sense in
which a non-transitive logic thinks it does. The fact that the logic-relative
version of GTT holds for the logics for which it does would seem simply
to show that such logics are in the relevant respects expressively adequate
Generalised Tarski’s Thesis Hits Substructure 251
rather than show that they all conform to the same format. GTT is paid lip
service, but wildly different doctrines are actually pursued behind the
chants of truth preservation.
Finally, it is hard to see what roles cases are playing anyways in the logic-
relative version of GTT. Taking a conditional of the form ‘If every prem-
ise is true, so is the conclusion’, say that its interpretation in terms of
classical [universal quantification and implication] is an ‘[individual-
truth]-preservation conditional’ (in the sense that the individual truth of
all the premises is preserved by the conclusion), whereas its interpretation
in terms of non-classical [universal quantification and implication] (as
these are available in the logic in question) is a ‘[collective-truth]-genera-
tion conditional’ (in the sense that the collective truth of the series of
premises generates the truth of the conclusion).28 (Make the same dis-
tinction if universal quantification is replaced by conjunction.)29 Then,
for the non-classical logics that satisfy the logic-relative version of GTT,
the non-classicality of plain [collective-truth]-generation conditionals
(which are typically much stronger than the corresponding [individual-
truth]-preservation conditionals) will presumably suffice to rule out the
undesired classical arguments, without quantification over cases and with
absolute truth instead of truth (or designated value) relative to cases.30 True,
that might still rule in a lot of other undesired supra-classical arguments
(i.e. arguments whose form is not valid even in classical logic), as, for
example, the argument from ‘Italy will play against Sweden’ to ‘Italy will
lose’ (since the [collective-truth]-generation conditional ‘If “Italy will
play against Sweden” is true, so is “Italy will lose”’ sadly holds). But, for
better or worse, such arguments will be ruled out by BRT2 (especially, by
the necessity and formality conditions). Cases are supposed to be nothing
less than the cornerstone of GTT (pp. 23–24), but they are nothing more
than a frill of its logic-relative version.
valid arguments with at least two premises supports some such extension,
and, if it does, whether it supports a privileged one that can be relied on
in vindicating the grounding in question.35
Secondly, and more importantly, the grounding in question runs afoul
of logics with non-standard behaviour of conjunction or implication. As for
conjunction, a good example is given by Sb (fn 6), which is a philosophi-
cally interesting and technically well-behaved logic, but in which φ & ψ
→ χ can be a logical truth without φ, ψ ⊢ χ holding (essentially because
φ, ψ ⊢Sb φ & ψ does not hold). As for implication, good examples are
given by B3 and K3, which are philosophically interesting and techni-
cally well-behaved logics, but in which φ ⊢ ψ can hold without φ → ψ
being a logical truth (essentially because φ → ψ is indeterminate as soon
as either (B3) or both (K3) of φ and ψ are). (A dual, equally good exam-
ple is given by the logic of paradox LP [Asenjo (1966)], which is a philo-
sophically interesting and technically well-behaved logic, but in which φ
→ ψ can be a logical truth without φ ⊢ ψ holding (essentially because, in
LP, φ → ψ is true as soon as φ is false, even if it is also true).) Notice that
all these examples depend on taking as operative conjunction and impli-
cation those immediately available in the logics. It is certainly technically
possible to extend the logics with further operations with a more stan-
dard behaviour; the problem is that there would typically seem to be no
privileged way of doing so [witness the plethora of non-material implica-
tions proposed for extending K3 or LP; see e.g. Field (2008); Priest
(2006) respectively, and Zardini (2016) for some critical discussion],
which undermines the grounding of logical consequence in logical truth,
since it leaves one with no determinate grounding basis. To elaborate on a
particularly resilient version of the problem, suppose that, for example, a
particular implication ⇒0 does at least the extensional part of the job for
some such logic L, so that, even without restricting to the ⇒0-free lan-
guage, φ0, φ1, φ2 …, φi ⊢L ψ holds iff φ0 & φ1 & φ2 … & φi ⇒0 ψ is a
logical truth in L. A particularly resilient version of the problem is that
there typically is at least one other, equally natural implication ⇒1 whose
logic and so whose meaning differ, respectively, from the logic and so from
the meaning of ⇒0 but which also does at least the extensional part of the
job, so that, even without restricting to the ⇒1-free (or ⇒0-free) lan-
guage, φ0, φ1, φ2 …, φi ⊢L ψ holds iff φ0 & φ1 & φ2 … & φi ⇒1 ψ is a
254 E. Zardini
logical truth in L. One would then be faced with the riveting question of
whether logical consequence in L is grounded in the ⇒0-implicational
logical truths of L or in the ⇒1-implicational logical truths of L (or in
both, or in either, or partly in one partly in the other, etc.).36
Thirdly, and even more importantly, the grounding in question is shat-
tered by logics with a standard behaviour of the determinacy operator [see
Zardini (2014c) for some contrast with a non-standard behaviour of that
operator]. A good example is given by the supervaluationist logic Sp [Fine
(1975)], which is a philosophically interesting and technically well-
behaved logic, but in which φ ⊢ ψ can hold without φ → ψ being a logi-
cal truth (essentially because φ ⊢Sp 𝒟 φ holds but φ → 𝒟 φ is not a
logical truth in Sp). Points analogous to those made in the last paragraph
hold concerning the possibility of extending the logic with a further
implication satisfying the deduction theorem.
Does the [logical-consequence]-first doctrine fare any better than the
[logical-truth]-first one? Well, the Tarskian shift from logical truth to logical
consequence is certainly beneficial insofar as one should also look at logical
consequence, which, for the reasons mentioned in the last three paragraphs,
in a few logics includes features that genuinely go over and beyond those of logi-
cal truth.37 This observation, however, falls dramatically short of vindicating
the [logical-consequence]-first doctrine in its full strength, since, for example,
the doctrine insists that, even for those logics where there is no such differ-
ence between logical consequence and logical truth, logical truth is grounded
in logical consequence. I’m going to attack the [logical-consequence]-first
doctrine precisely by arguing that, for some of those logics (in particular, for
the specific substructural logics alluded to in Sect. 2), it is logical conse-
quence that is grounded in logical truth rather than vice versa.
My argument starts with the plausible if somewhat controversial
assumption that logical consequence is itself grounded in facts about truth,
rather than being primitive [e.g. Field (2015)] or grounded in other kinds
of facts like, for example, facts concerning proof [e.g. Prawitz (2005)]. In
turn, this standard, general semantic conception of logical consequence is
typically accepted in a specific version, in classical logic, according to
which logical consequence is grounded in guaranteed preservation of the
truth of the premises by the conclusion [following in this Tarski again, espe-
cially Tarski (1936)].
Generalised Tarski’s Thesis Hits Substructure 255
and implication, respectively) and the fact that, in LW, by adjunction,
‘Snow is white’, ‘Grass is green’ entail ‘Snow is white and⊗ grass is green’.
According to my proposal, that is grounded in (the logical necessity of )
the fact that, if→ ‘Snow is white’ is true and⊗ ‘Grass is green’ is true, then→
‘Snow is white and⊗ grass is green’ is true, whereas, according to the alter-
native proposal, that is grounded in (the logical necessity of ) the fact that,
if→ snow is white and⊗ grass is green, then→ snow is white and⊗ grass is
green. The former account of the matter is clearly superior, since it
respects the difference in the argument between the series of premises and the
conclusion, and in particular it respects the point of the argument which is
to the effect that the simple ‘Snow is white’ and ‘Grass is green’ together
suffice for the complex ‘Snow is white and⊗ grass is green’. As in the classi-
cal case, it just is illuminating to use in the metalanguage the notions of
conjunction and truth for combining two premises and then, mentioning
a certain single sentence of the object language which involves an opera-
tor expressing the very same notion of conjunction, explaining why that
conclusion follows from the two premises—it just is illuminating to use
the idea of two sentences being both true to make a point about a single
sentence involving ‘and’.47 All this is completely missed by the latter
account, which flattens such wealth of structure concerning the interaction
between premise combination and conjunction into a tautology that has
nothing to do with either.48 Semantic ascent allows us to discern the fine
structure of logical consequence: its combination of premises and its
entailment from premises to conclusion.49
Having resisted the temptation, it only remains to state with due
solemnity a consequence of the approach we’ve been pursuing which has
been looming large for a while (and which, for that matter, is also a con-
sequence of the alternative proposal discussed in the last paragraph).
Namely, and quite literally, on this revolutionary approach, the [logical-
consequence]-first doctrine is turned upside down: in the specific sub-
structural logics alluded to in Sect. 2, not only logical truth is not
grounded in logical consequence, but, vice versa, it is logical consequence
that is grounded in logical truth.50 In those logics, the central property is
logical truth, and logical consequence is grounded in it as being merely a
degenerate limit case of logical truth: those logical truths that are [collective-
truth]-generation conditionals (where the series of premises and the con-
Generalised Tarski’s Thesis Hits Substructure 259
Notes
1. Throughout, page references are to BR unless otherwise stated.
2. Throughout, I follow BR in assuming, merely for simplicity, a multiple-
premise, single-conclusion framework. The extension of my points to the
(superior but) more complex multiple-premise, multiple-conclusion
framework is straightforward.
3. BR claim that necessity is important for logical consequence at least
partly because it guarantees that logic applies unrestrictedly in hypothetical
reasoning (pp. 15–16). But it arguably doesn’t in the first place, since, say,
the law of excluded middle (henceforth, ‘LEM’) is a logical truth but it is
arguably wrong to reason that, if Brouwer were right, it would be the
case that either Goldbach’s conjecture holds or it doesn’t. Notice that the
hypothesis is perfectly possible (Brouwer might have held correct views),
so it is not even the case that logic applies unrestrictedly to possible hypotheses
(which would anyways raise the issue of why it should be more impor-
tant unrestrictedly to apply to possible hypotheses rather than to, say,
interesting hypotheses).
4. Throughout, I work with a standard framework for context-dependent
languages where the truth of a sentence is relative both to a context of
utterance and to a circumstance of evaluation, and I make such relativities
explicit with the relevant parts of the construction ‘φ is
true-as-uttered-in-C-as-evaluated-at-E’.
5. I’d like to forestall two likely reactions to this point. Don’t say that what
is true is ‘For every context C, if “Snow is white” is true-as-uttered-in-
C-as-evaluated-at-C, “Actually, snow is white” is true-as-uttered-in-C-as-
evaluated-at-C’. The proposed Ersatz has the embarrassing problem
(emerging in the opening ‘what is true [my emphasis, EZ]’) that neither
it nor its embedded conditional is non-trivially necessary (so that the
spirit if not the letter of the necessity condition would seem violated).
What the proposed Ersatz indeed offers is universality over contexts, but
that is a far cry from any interesting necessity: for every context C, ‘I exist’
260 E. Zardini
11. Although, should a substructural logic prove to be part of the best solu-
tion to, for example, the semantic paradoxes, I myself would find it ter-
minologically misguided not to label it ‘legitimate relation of logical
consequence’ (surely, whatever logic governs truth deserves to be labelled
‘legitimate relation of logical consequence’!).
12. In the text, I’ll briefly mention what I regard as the best such reasons not
to commit to denial of (I)—an issue which obviously lies well beyond
the scope of this chapter—but simply to make more vivid how interest-
ing philosophical positions denying (I) would unhelpfully be outlawed
by GTT. Analogous comments apply for the other structural properties
to be considered below in the text.
13. That is, by relying in a non-deviant way on one’s belief in the premises of the
argument and on one’s inference from them to the conclusion: one can per-
haps acquire a new justification for believing, say, ‘There are circular
arguments whose conclusion I’ve inferred from their premises’ by ‘going
through’ the relevant instance of (I), but one would thereby be relying in
a clearly deviant way on one’s belief in the premise of the argument and
on one’s inference from it to the conclusion. As elsewhere [see e.g.
Chisholm (1966), p. 30 for the much discussed case of deviant causa-
tion, in whose debate it would ironically seem presupposed that deviance
only affects causation and not also rational connection], it is a tricky issue,
lying beyond the scope of this chapter, to spell out exactly what such
deviance is.
14. Even the crooked simplification argument, say, φ & ψ ⊢ φ passes muster,
since one can be told that Al met with Bob and Cate and acquire a new
justification for believing that Al met with Bob by inferring it by simpli-
fication from what one has been told.
15. This uncontroversial fact would easily be accounted for if, when trans-
mission occurs, the justification for believing the conclusion were simply
identical with the justification for believing the premises, for then it would
be obvious that the justification for believing the conclusion of (I) can-
not be new with respect to the justification for believing its premise. Pace
Moruzzi and Zardini (2007), p. 181, that would seem however a sim-
plistic view of what happens when transmission occurs (a mistake for
which I assume the sole responsibility!), since, presumably, also the justi-
fication one has for inferring the conclusion from the premises is part of the
justification one acquires for believing the conclusion. While the ‘mereol-
ogy of justification’ is still in its infant days, a natural speculation is that
264 E. Zardini
30. To set aside distracting issues concerning the opacity of absolute truth
[e.g. Zardini (2015b)], in this discussion I assume that it makes sense to
extend the target logics with a quoting singular term ⌜φ⌝ for each sen-
tence φ of their original language and with a truth predicate T such that,
for every sentence φ of their original language, T⌜φ⌝ is intersubstitut-
able with φ. Such extensions are straightforward (indeed, on my view,
just as logical as, say, the extension of the conjunction-free fragment of
intuitionist logic with conjunction), contrary to those that would be
needed to develop a theory of truth-in-a-case (as per the third last para-
graph in the text). (Such extensions are also harmless, since they only
licence the intersubstitutability of T ⌜φ⌝ with φ if φ is T-free.) Moreover,
focusing on languages for which such extensions make sense is justified
since, in this discussion, my aim is to defend anti-universalist claims
rather than universalist ones. Thanks to José Martínez and Ricardo
Santos for feedback on some of these issues.
31. I should really be a bit more precise about what it is for a logic to satisfy
the absolute-truth formulation of the logic-relative version of GTT. For our
purposes, anticipating a bit, a natural way of making that notion precise
is to say that a logic L satisfies the absolute-truth formulation of the
logic-relative version of GTT iff [φ0, φ1, φ2 …, φi ⊢L ψ holds iff T ⌜φ0⌝
& T ⌜φ1⌝ & T ⌜φ2⌝ … & T ⌜φi⌝ → T ⌜ψ⌝, as interpreted by L, is a logi-
cal truth in L] (where, by fn 30, the last claim is equivalent with the
claim that φ0 & φ1 & φ2 … & φi → ψ, as interpreted by L, is a logical
truth in L, which makes satisfaction of the absolute-truth formulation of
the logic-relative version of GTT by L a matter of L’s conjunction and
implication correlating in the familiar ways to premise combination and
entailment in L, respectively). Thanks to Ricardo Santos for urging this
clarification.
32. Yes, including non-contractive logics, even given what I’ve said in sec-
tion “A Logic-Relative Version of GTT; an Absolute-Truth Formulation
of a Logic-Relative Version of GTT”, since we’re considering a formula-
tion of the logic-relative version of GTT in terms of plain [collective-
truth]-generation conditionals, without cases or designated values: while
it is the case that, if φ gets a designated value, so does φ ⊗ φ, we do not
have that, if φ is true, so is φ ⊗ φ [Zardini (2011)].
33. Throughout, I understand ‘ground’ and its relatives in a suitably broad
fashion, so as to include also the case of reduction (as is particularly plau-
sible in the case of the [logical-consequence]-first doctrine).
Generalised Tarski’s Thesis Hits Substructure 269
34. The second main conjunct in its full strength is not guaranteed by the
first one, and so, if at all desired, it must be added separately. Still, it
represents the arguably most plausible grounding of logical truth in logi-
cal consequence, and that’s why, throughout, we’re focussing on it.
Notice that ‘most plausible’ doesn’t imply ‘plausible’: following, but from
no premises, would seem to make just as little sense as arriving, but at no
places! Sometimes, the problem is fudged by invoking the empty set and
saying that logical truths are those sentences that follow from the empty
set; however, while it is useful formally to model logical consequence as
a relation between sets (or, in view of substructurality, series), only an
empty-set mystic would think that logical truths are characterised by a
distinctive logical relation to the empty set (as opposed to many other
more relevant objects). Some other times, the problem is fudged instead
by invoking the 0ary operator t and saying that logical truths are those
sentences that follow from t; however, while it is useful to introduce t in
the formal study of logical consequence (especially in view of substruc-
turality), its informal understanding is as of the ‘conjunction’ of all logical
truths, which arguably prevents grounding a sentence being a logical
truth in its following from t. Notice that an analogous problem affects
the related idea that logical falsity is grounded in logical consequence as
being merely a limit case of logical consequence with one premise and no
conclusions (whereas, FWIW, the more general notion of inconsistency of
a series of premises can still be grounded in logical consequence and logi-
cal falsity in terms of the series entailing the ‘disjunction’ of all logical
falsities). Thanks to Ricardo Miguel for pressing me on the formulation
of the [logical-consequence]-first doctrine.
35. Having noted all this, I’ll (pick up from fn 31 and) henceforth focus on
the finite case myself.
36. Even setting aside the problem raised in the text, the extension of these
logics with the desired operations typically requires complications going so
far beyond the basic, natural framework of the logics as to make the [logical-
truth]-first doctrine hardly credible for them. Moreover, the resulting
operations are typically so tailor-made to fit the target relation of logical
consequence that, given on the one hand the valid arguments of one of
these logics and on the other hand its putatively grounding logical truths,
the by far most plausible account is that the putatively grounding logical
truths are what they are because the valid arguments are what they are
rather than vice versa.
270 E. Zardini
theoretic’ values such as truth, falsity, unfalsity, etc., whereas the distinction
among designated values in the model theory of some non-reflexive and
non-transitive logics does not sustain any such interpretation [for exam-
ple, think of the non-[truth-theoretic] interpretation offered by Zardini
(2008), pp. 347–349 of the distinction between two different kinds of
designated values relevant for a wide family of non-transitive logics].
42. We can further argue not only that substructural logical consequence is
not grounded in truth preservation, and not only that substructural logi-
cal consequence is not co-extensional with truth preservation, but also
that substructural logical consequence does not even require truth pres-
ervation. The best example for this is arguably offered by the non-con-
tractive approach to the semantic paradoxes mentioned in section
“Contraction” [especially as further developed in Zardini (2018f )], on
which one can warrantedly accept, say, that the Liar sentence λ is true
while holding that the argument from λ, λ to ‘Snow is black’ is valid, and
so while holding that the premise (which occurs twice, see fn 22) of the
argument is true and its conclusion is not. Substructural logical conse-
quence does not require truth preservation. Thanks to Sven Rosenkranz
for prompting this expansion of the argument.
43. To be clear, the claim is not that the intended class is not co-extensional
with any class defined in purely non-logical terms—not the least because,
under very plausible mathematical assumptions, such co-extensionality
does obtain in the case of many logics! Especially for such logics, the
claim is not that logical consequence is not correlated with individual-
truth preservation over a class of tests defined in purely non-logical
terms; it is rather that logical consequence is not grounded in such truth
preservation, as it is only correlated with it because the class of tests
defined in purely non-logical terms just so happens to be co-extensional
with the intended class, which is in turn only characterisable as such partly
in logical terms [for example, as the class of all and only those tests involv-
ing logically possible models, cf Etchemendy (1990), pp. 107–124].
44. This is not to deny that something of heuristic value can be gained by
working with the notion of a sentence being true in every logically possible
model rather than with the notion of a sentence being logically necessary,
just like something of heuristic value can be gained by working with the
notion of a sentence being true at every metaphysically possible world
rather than with the notion of a sentence being metaphysically necessary:
in both cases, quantificational reasoning, based on the well-understood
272 E. Zardini
47. While this deft movement from use to mention is illuminating, the light
it sheds is admittedly somewhat faint. But then there is only so much
light to be had at these depths.
48. Don’t say that the fact that φ0, φ1, φ2 …, φi ⊢L ψ holds is grounded
instead in the fact that the object-language, conjunction-free conditional
φ0 → (φ1 → (φ2 … → (φi → ψ))) …) is a logical truth in L (cf fn 46), so
that, in particular, the fact that, in LW, ‘Snow is white’, ‘Grass is green’
entail ‘Snow is white and⊗ grass is green’ is grounded in (the logical
necessity of ) the fact that, if→ snow is white, then→, if→ grass is green,
then→ snow is white and⊗ grass is green. Setting aside the adhocness of
such deviation given the route we’ve followed starting from GTT, the
deviation stumbles on exactly the same problem raised in the text when
grounding the fact that, in LW, by modus ponens, ‘If→ snow is white,
then→ grass is green’, ‘Snow is white’ entail ‘Grass is green’, since it
grounds it in (the logical necessity of ) the tautologous fact that, if→, if→
snow is white, then→ grass is green, then→, if→ snow is white, then→ grass
is green. Thanks to José Martínez for comments on some of these issues.
49. Semantic ascent could probably not fulfil this function if alethic defla-
tionism held. Ergo, by modus tollens …
50. In turn, it is plausible that logical truth, even in those logics, is not primi-
tive, and we may expect that its account will appeal to properties of the
logical operations used or mentioned in a logical truth (the latter disjunct
being particularly relevant when the logical truth involves [collective-
truth]-generation conditionals). If so, whether structural properties are
valid is grounded in whether certain sentences involving [collective-
truth]-generation conditionals are logical truths (see fn 51 for a specific
proposal), which is in turn grounded in the properties of logical operations.
Substructurality is a logically interesting but philosophically shallow
phenomenon caused by logically hidden but philosophically active
underlying logical operations. Structure is grounded in operations.
Thanks to Ricardo Miguel for pushing me on this.
51. Although a full treatment of the status of metarules (and metametarules,
and metametametarules …) lies beyond the scope of this chapter, a natural
way of extending the approach we’ve been pursuing to metarules for the
specific substructural logics alluded to in Sect. 2 is to say that the fact that a
metarule is valid (over and above its being admissible) in L is grounded in
the fact that the conditional having as antecedent the conjunction of the
series of [collective-truth]-generation conditionals corresponding to the
‘premise rules’ of the metarule and as consequent the [collective-truth]-gen-
eration conditional corresponding to the ‘conclusion rule’ of the metarule is
274 E. Zardini
a logical truth in L (and then iterate this strategy for metametarules, metam-
etametarules, metametametametarules …). Thanks to Bogdan Dicher
and Lucas Rosenblatt for their questions about the status of metarules.
References
Anderson, A., and N. Belnap. 1975. Entailment. Vol. I. Princeton: Princeton
University Press.
Asenjo, F. 1966. A Calculus of Antinomies. Notre Dame Journal of Formal Logic
7: 103–105.
Beall, J., ed. 2007. Revenge of the Liar. Oxford: Oxford University Press.
———. 2015. Free of Detachment: Logic, Rationality, and Gluts. Noûs 49:
410–423.
Beall, J., and G. Restall. 2006. Logical Pluralism. Oxford: Oxford University
Press.
Bočvar, D. 1938. Ob odnom trexznačnom isčislenii i ego primenenii k analizu para-
doksov klassičeskogo rasširennogo funkcional’nogo isčislenija. Matematičeskij
sbornik 4: 287–308.
Bolzano, B. 1837. Wissenschaftslehre. Vol. II. Sulzbach: Seidel.
Brouwer, L. 1927. Über Definitionsbereiche von Funktionen. Mathematische
Annalen 97: 60–75.
Chisholm, R. 1966. Freedom and Action. In Freedom and Determinism, ed.
Keith Lehrer, 11–44. New York: Random House.
Cobreros, P., P. Égré, D. Ripley, and R. van Rooij. 2012. Tolerant, Classical,
Strict. Journal of Philosophical Logic 41: 347–385.
Dummett, M. 2000. Elements of Intuitionism. 2nd ed. Oxford: Oxford University
Press.
Etchemendy, J. 1988. Tarski on Truth and Logical Consequence. The Journal of
Symbolic Logic 53: 51–79.
———. 1990. The Concept of Logical Consequence. Cambridge, MA: Harvard
University Press.
Field, H. 2008. Saving Truth from Paradox. Oxford: Oxford University Press.
———. 2015. What Is Logical Validity? In Foundations of Logical Consequence,
ed. Colin Caret and Ole Hjortland, 33–69. Oxford: Oxford University Press.
Fine, K. 1975. Vagueness, Truth and Logic. Synthese 30: 265–300.
Girard, J. 1995. Linear Logic: Its Syntax and Semantics. In Advances in Linear
Logic, ed. Jean-Yves Girard, Yves Lafont, and Laurent Regnier, 1–42.
Cambridge: Cambridge University Press.
Generalised Tarski’s Thesis Hits Substructure 275
1 Introduction
Methodology in logic has, on the whole, conceived of a logic as a system
of rules or principles. A correct logic is one which includes correct prin-
ciples, and logical reasoning is reasoning in accord with such correct prin-
ciples. The contemporary versions of this approach use a mathematically
precise formal language to express these principles, which are understood
similarly to the mathematical expression of laws of nature. We call this
orthodox view ‘logical generalism’.
The assumption of generalism is so broadly made that it is difficult to
make sense of giving it up. Generalism might seem to be constitutive of
G. Payette
Department of Philosophy, University of British Columbia,
Vancouver, BC, Canada
e-mail: gpayette@dal.ca
N. Wyatt (*)
Department of Philosophy, University of Calgary, Calgary, AB, Canada
e-mail: nicole.wyatt@ucalgary.ca
2 Validity
Our concern is with the validity of natural language arguments. This is
not to deny that formal logics have many other important applications,
nor that logical systems which have no relevance to natural language
may be of considerable interest. But it is only when logics are under-
stood as theories of a consequence relation in natural language that the
debates between logical monists and logical pluralists make sense: for
any logics to be correct there has to be something that they are correct
with respect to. Following Priest (2014), we call this the canonical appli-
cation for logic.
Monists and pluralists agree that at least one logic is a correct theory
of natural language validity. In contrast, logical nihilists and logical par-
ticularists share the view that no logic provides such a theory. For nihil-
ists like Curtis Franks, this is a reason to give up the canonical application
entirely:
Logical Particularism 279
fixing our sights on [natural language validity] saddles logic with a burden
that it cannot comfortably bear, and that logic, in the vigor and profundity
that it displays nowadays, does and ought to command our interest pre-
cisely because of its disregard for norms of correctness. (Franks 2015, 148)
3 Logics
Our view is that formal logics are mathematical models of validity; admit-
tedly, that is not an uncommon view. But there are important distinc-
tions between our view and other versions of the logic-as-modelling view.
Consider, as a representative example, the view of Roy Cook (2002). On
his account logic aims to model natural language and in particular argu-
ments constructed in natural language—including technical languages
like that of mathematics.4
According to Cook, logical models—proof theory, model theory, for-
mal language—contain certain mathematical objects that accurately rep-
resent some aspect of language/reality. These accurate representations are
the representors.5 Other components of the models are merely mathe-
matical ‘artifacts’ not intended to represent any part of reality. The stan-
dard metaphor is that of a model ship: the model may have certain
components which do not exist in the real ship, and may lack certain
components of the real ship.
Which elements are the representors and which the artifacts in a model
depends on what we want a model to do. For example, a model ship may
not have engines. If we are looking for a model that represents the exte-
rior of a ship, engines are unnecessary. But another model that is meant
for hydrodynamic ‘tank tests’ of the ship may also lack engines, but
include a lead block where the engines are in the actual ship to represent
the position and mass of the engines (though not the moving parts of the
engines since those don’t matter for this purpose). The real ship does not
have a lead block where its engines are; but the lead block is a representa-
tion of a particular, relevant aspect of the real boat. But now suppose we
want to put a model of the ship on a shelf, and we want to make sure that
it doesn’t fall over, so we put in a lead block which happens to be where
the engines are. The representors in this case are again the details on the
exterior of the model. Before, the block was a representor; now it is an
artifact.
Given this understanding of modelling, in order for a logic to correctly
represent validity in natural language, the elements of the model that
stand in relations of consequence must be representors rather than arti-
facts. This means the formulas of the formal languages must represent the
sentences of natural language. On the standard picture, they do so by
282 G. Payette and N. Wyatt
4 Modelling
We have denied that logics represent features of natural language. Contra
the nihilist, we maintain that logics do shed explanatory light on natural
language consequence. To understand how non-representational models
can explain, we turn to modelling in the sciences. We focus on three key
features of explanatory but inaccurate or non-representative models:
abstraction, imagination, and scaffolding.
As Godfrey-Smith observes (2009), scientists from a variety of disci-
plines spend time considering things that don’t exist: examples include
ideal gases, frictionless planes, infinitely large biological populations,
wholly rational agents, and biologically implausible neural networks.
Theoretical investigation of these apparently increases our understanding
of the behavior of actual gases, the movement of physical objects, evolu-
tion in finite populations, human behavior, and human and animal cog-
nition. All of these are part of model-based science, in which one system is
explored as a basis for understanding another system.
This seems puzzling—how can studying things that do not exist lead
to knowledge of things that do? Unsurprisingly, philosophers of science
disagree on exactly how ‘false’ models—models that rest on known false
assumptions—function in scientific practice. But there are a number of
salient themes in the literature.
Firstly, many false models rest upon abstraction in various ways. When
we posit ideal gases and wholly rational agents, we are abstracting away
from details which matter in the actual world. Abstraction depends on
making false assumptions and ignoring details, but increases the
tractability of the problem. For example, in ecology scientists face two
obstacles for studying real systems: (a) the time scale on which they oper-
ate tends to exceed the time available for study, and (b) their complexity
makes it difficult to manipulate them systematically (Odenbaugh 2005,
233). Abstract mathematical models in ecology allow both problems to
be overcome. Moreover, even when a more detailed and perhaps accurate
model is available, an abstracted model may be more cognitively and
computationally tractable, and thus more useful, as Woody (2013,
1574–5) argues with respect to the ideal gas law.8
284 G. Payette and N. Wyatt
142). And the table can function in this way despite the fact that the so-
called periodic law that underlies it does not seem to reveal any causal
structure or identify any mechanisms, nor is it precise enough to support
scientific predictions (Woody 2014, 133–4, 143).10
Some arguments for logical nihilism are motivated by the failure of
formal logic(s) to offer accurate predictions for natural language conse-
quence. But this assumes that the only roles that can be played by logics
with respect to natural language validity are prediction and description.
Franks’ argument for nihilism is, in part, that no system can capture the
variety of logical properties that we might be interested in. But he still
thinks that were a logic useful for explaining natural language validity, it
would have to be correct, that is, as a theory it would be true or as a
model it would have to be accurate. In contrast, the literature on scien-
tific modelling suggests that false models and theories can play a role in
scientific explanation in a variety of ways.
5 Logicians
One common thread in our rapid survey of false models in science is the
attention paid to the way in which scientists make use of these models,
and the roles that abstraction, imagination, and scaffolding play. This
echoes Ronald Giere’s views about modelling in general, which character-
ize representation as something scientists do with models for specific pur-
poses, rather than as a timeless property of a model (Giere 2004, 743).11
When it comes to logics, the question is not whether the formal sys-
tems correctly represent some feature of natural languages. The right
question is how these models are used to explain. Our claim is that logi-
cians interested in natural language use logics to represent in order to
accomplish their explanatory goals.12 Meeting those goals involves
abstraction, imagination, and scaffolding. It turns out that this perspec-
tive makes sense of three practices of logicians which at first may not
make sense on ‘representationalist’ accounts of modelling and veridical
explanatory theories.
Models in logic can be any or all of formal languages, proof theories,
or model theories. Given the target phenomena of logical study we are
286 G. Payette and N. Wyatt
structures. In social choice theory, one lays out a set of axioms for how,
for example, a choice function behaves, and then shows that those axioms
correspond to a certain kind of preference relation.14 In logic, we see rep-
resentation theorems in a variety of forms.
One form we have already mentioned: modal correspondence theory.15
Some object language formulas, when true at every world in every model
on a frame, mean that the relation in that frame has a certain property:
transitivity, reflexivity, and so on. But the most common representation
theorem one sees in logic is completeness.16 Generally completeness
involves the representation of a certain semantic notion of consequence
in terms of a proof-theoretic form of consequence—or semantic truth in
terms of theoremhood.17 Another form of representation theorem is
when one looks for a set of proof rules, usually in a sequent calculus, that
can be characterized by a certain set of properties. Such characterizations
were what Michael Dummett used to argue for the supremacy of intu-
itionistic logic in (Dummett 1991). How do these connect to
imagination?
Representation theorems show the range of variation for a given use of
a model. A completeness theorem shows which logical relationships
between formulas are in and which are out. If we want to include differ-
ent forms, we know we have to change the model theory in some way.
Similarly for sets of sequent proof rules. If we want the set of proof rules
to have an extra property, we know that the consequence relation will
have to change. Correspondence theory shows a closer connection
between model theory and formal language than completeness theorems.
It shows that sometimes general facts about the model theory can be
brought into the object language: properties about the frame relation can
be expressed by the formulas. If we want to change the model theory,
then, we know the logically true formulas will change.
The final role of models, scaffolding, explains why logicians haven’t
discarded classical propositional logic (CPL) despite its limitations. They
add to it and subtract from it in various ways. Sometimes they add prop-
ositional connectives as in modal logics. Sometimes they model sub-
propositional structure as in first-order logic or branching quantifier
logics. Subclassical logicians may weaken it by rejecting, for example,
excluded middle. Logicians give systems which differ from CPL, and
288 G. Payette and N. Wyatt
philosophers of logic tend to debate about the virtues and vices of rival
logics in their propositional forms. We know the fit between proposi-
tional languages and natural language arguments isn’t perfect, but we
keep them around. More importantly, logicians don’t get rid of CPL: it is
what every logician cuts her teeth on. It also acts as a logica franca for
metalogical investigations into logics.
CPL acts as a conceptual scaffolding for understanding other logics.
CPL is really big: you can prove a lot using it. Most of the propositional
logics that are studied are sub-logics of CPL. That is largely because it is
post-complete18 and it is structurally complete.19 Often, logics are studied
by how their proof theory or semantics must be different from those for
classical logic; it is our logical jumping-off point.
In the end, we see all of those modelling practices from other sciences,
which also use inaccurate models, at work in the practices of logicians.
But all of our talk about applications and modelling may be reminiscent
of another perspective in the philosophy of logic that we want to distin-
guish from our view: pluralism.
6 Pluralism
Let us take stock. Like the nihilist, we are convinced by the existence
of counterexamples to every putative logical law that no logic describes
a consequence relation in natural language. Unlike the nihilist, we do
not conclude that logic contributes nothing to our understanding of
natural language validity. According to the particularist, logics can
and do reveal something about the genuine connection between prem-
ises and conclusions in natural language. Logics are best understood as
explanatory but inaccurate/non-representative/fictional models; they
function like other ‘false’ models in the sciences. Before we turn to
pluralism we have to highlight a couple of important features of
particularism.
The common understanding of logics as models assumes, one, natural
language utterances have logical forms, and two, what matters to conse-
quence is stable across contexts and arguments. We think both assump-
tions are false. The former is argued against elsewhere.20 The latter is a
Logical Particularism 289
common ground for the varied particularist projects. As Dancy puts it,
“[t]he core of particularism is its insistence on variability” (2013).21
Whether moral or logical, particularism carries this commitment to
the variable relevance of features. In the logical context, having the form
of modus ponens, under some system of logical analysis, is a relevant
feature to the assessment of some arguments, but not in the assessment of
others. Faced with counterexamples, the particularist feels no need to
offer increasingly recondite analyses of the arguments in order to escape
the conclusion that modus ponens is invalid.22 As Dancy emphasizes, the
particularist takes the view that moral—or in our case logical—reasons
function in ways “not noticeably different from the way in which other
reasons function—more ordinary reasons for action, say, or reasons for
belief rather than for action” (Dancy 2013).
We now want to say a little more about what separates our view from
that of the pluralist. Following Haack (1978), we distinguish between
global pluralism—the view that there are multiple logics which are correct
come what may—and local pluralism, the view that logics are only correct
relative to a particular domain of inquiry. Much of the more recent work
on pluralism takes the view that global pluralism is the only pluralism
worthy of the name, on the grounds that it is only the global pluralist
who claims that a specific argument could be simultaneously valid and
invalid.23 But Shapiro (2014) offers a robust defense of a particular ver-
sion of local pluralism.
Global pluralists and monists share the common assumption that there
is a general-purpose consequence relation for natural language that
logic(s) aim to describe. They differ on how many such relations there
are, but agree that any such relation applies in any context of reasoning.
In contrast, the local pluralist maintains that there are different
consequence relations that apply in different contexts of reasoning. This
might, on first glance, look very much like the particularist view. After
all, we have said that validity is a matter of preserving some property or
other in the move from premises and conclusion, and that what property
must be preserved changes from case to case.
The point of fundamental disagreement between us and the local plu-
ralist lies with the notion of correctness. While the local pluralist allows
what matters to consequence to vary from context to context, they think
290 G. Payette and N. Wyatt
the aim of logic is accurate prediction and adjudication for natural lan-
guage validity. To avoid the pitfalls of the global pluralist, the local plural-
ist needs some way of demarcating the area where each correct logic is in
force. For example, a local pluralist might maintain that there are differ-
ent logics for different areas of inquiry (which might be epistemologically
or ontologically distinguished); or different logics for different sets of
logical connectives (and perhaps that a single natural language connective
can represent different logical connectives); or different logics for differ-
ent sets of possible truth values (e.g. one might think that claims about
the natural world must either be true or false, but that vagueness in social
facts requires a plurality of truth values).
The need for demarcation makes local pluralist vulnerable to the
arguments of the nihilist. As Russell (2017) argues, we can ‘validate’
any purported logical law by restricting the substitution class into that
law. For example, if we allow only true sentences to be substituted for
φ and ψ in the form φ → ψ, ψ ⊨ φ, then this ‘law’—affirming the con-
sequent—would seem to be correct. That isn’t a convincing justification
of affirming the consequent precisely because the restriction on substi-
tution is ad hoc. But then the question becomes, what restrictions are
legitimate?
The local pluralist needs a method of demarcation independent of any
judgments about validity. For example, the local pluralist cannot restrict
substitution to sentences where the laws of classical logic apply, and use
that as an argument that classical logic is appropriate for that domain.
Such a demarcation may be possible, but we are skeptical because any
such demarcation would depend upon a logic neutral account of logical
form.24
In contrast, the particularist has no need of an independent means of
determining which logic applies to a particular argument; from our
perspective the role of logic is not to adjudicate or predict validity but to
explain it. Features and properties have variable relevance to validity, and
we cannot determine in advance of considering a specific argument which
properties are relevant. But, once we have decided what is relevant, logics
can explain how the relevant validity-making properties could be pre-
served between premises and conclusions.25
Logical Particularism 291
7 Particularist Logic
One natural objection to our view rests on the idea that logics are not just
explanatory but also action guiding. Certainly logicians, qua scientists,
may act as we have described, but non-logicians use logic to give argu-
ments. Moreover, they use formal logic to give and check their arguments.
That means they use a logic’s rules to ensure the validity of their argu-
ments by following the rules of that logic. Now for what might seem to
be a U-turn: this practice can be acceptable.
We take logic to be the study of connections between premises and
conclusions. Formal systems show how connections between premises
and conclusions can be made. In classical logic, for example, we use a
conception of ‘connection’ as truth preservation. Classical logic also pro-
vides a precise account of what ‘truth preservation’ means.26 This model
sets a narrow boundary for what can count when evaluating an argument
for validity. The flaw in the generalist way of thinking is the assumption
that model must be applied everywhere, always.
What about premises that can’t all be true simultaneously? Something
other than truth must matter in that case. Here we see logic not just as a
science, but as engineering. We look for some property that might suit
our purposes, and then we see what we can do to model it as a connec-
tion. A favorite example is the notion of level of consistency or just
‘level’—from Jennings and Schotch (1984). Level is the answer to the
question: what is the size of the smallest partition of the set in which all
the sets in the partition are consistent? It measures inconsistency. The
level of {p, ¬p} is two, because it takes at least two sets in a partition to get
consistent sets. The level of {q ∧ p, p ∧ ¬ q ∧ r, p ⊃ (¬q ∧ ¬ r)} is three.
One relation that preserves level is defined by saying t follows from a set
of premises Γ when t follows classically from some set in every partition
of Γ that is the same size as the one used to determine the level of the set,
and each element of the partition is consistent.27 The level of the set of
consequences on this definition will be the same as the level of the
premises.
This level-preserving approach to inconsistent consequence builds one
model atop another as scientists are wont to do. It gives an altogether
292 G. Payette and N. Wyatt
8 Conclusion
Reasons have variable relevance. The reason that a glass dropped on a
cement floor broke is that it is fragile and the cement is hard. In another
case, I might drop an identical glass and it might not break. In this sec-
ond case, the fact that it is fragile and I dropped it on hard cement is not
a reason that the glass broke, for the simple reason that the glass is intact.
Dancy (2013) has pointed out that reasons can have this variability in the
moral context, and we are pointing out that they have that variability in
the logical context.
The logical particularist maintains that an argument analyzed as hav-
ing the form of modus ponens when using classical logic might well be
the reason it is valid in one case. In another case, an argument with that
form in classical logic is not valid, and the classical analysis provides no
reason for taking it to be.
Nonetheless, we think that the standard methodology of logic is pro-
ductive and contributes to the historical goal of explaining validity in
natural language reasoning. Logical particularists use logics as explanatory
294 G. Payette and N. Wyatt
tools and take the explanations to be on par with those offered by other
sciences. Logics do not explain validity on their own. In this chapter,
we’ve sketched an account of the functioning of logics as models which is
continuous with the use of such models in other scientific endeavors.
Specifically, we have argued that models in logic serve our explanatory
goals without aiming at, or succeeding at, representation.
Notes
1. Although we have argued elsewhere that giving the same rules to connec-
tives does not mean that the connectives have the same meaning qua
natural language connectives (Wyatt and Payette 2018), we do not doubt
the mathematical significance of the construction and the purely math-
ematical conception of proof-theoretic meaning used in the construc-
tion. It is the mathematical significance of the constructions that interests
Franks, if we have understood him correctly. We are not at odds on this
point.
2. See Payette and Schotch (2007), Jennings et al. (2009).
3. As will become clear from the subsequent discussion, we do reject the
view on which formal logic sheds light on the validity of arguments by
directly representing the semantics of natural language sentences. See
Stokhof (2007) for a discussion of the relationship between formal lan-
guages and natural language semantics that is somewhat parallel to the
view of the relationship between formal logics and natural language
argument advocated here.
4. Cook’s account of how logics model validity is very similar to that of
Shapiro (2014): the primary difference is that Shapiro takes logics to be
only in the business of modelling mathematical argument, while for
Cook mathematical argument is a special case of a more general goal.
5. Cook refers to models as ‘mock-ups’ (Cook 2002).
6. Notice that this picture is absolutely neutral on the nature of the rela-
tionship between the phonological realization or the syntax of natural
language sentences and their logical forms.
7. See Payette and Wyatt (2018) for more discussion of this point.
8. Abstraction here covers both idealization and fictionalization in science,
where fictional models are those in which no amount of de-idealization—
Logical Particularism 295
that is, no amount of adding in details left out or correcting false assump-
tions—would recover a correct description of the actual world (Bokulich
2011).
9. Hindriks (2008) also argues that models in the economics context
can be better when containing a larger number of unrealistic
assumptions.
10. The periodic table does not, contra the usual historical story, indicate
directly the presence of missing elements, nor does it generate any pre-
dictions of specific quantitative properties. See Woody (2014) for fur-
ther discussion.
11. Giere proposes that the representative features of models fall under the
general schema: S uses X to represent W for purposes P. Some of this
purpose relativity is already present in Cook’s view since which compo-
nents are representors and which are artifacts may depend on the goal of
the model. That goal is, presumably, imposed on the model. It is possible
that Cook would be in closer agreement than we have suggested here,
but note that critics, for example, Smith (2011), describe Cook’s view as
one where representation is a relation between (parts of ) models and
reality, rather than the 4-ary relation above. We won’t enter any further
into debates about how to interpret Cook here since our goal is to
advance our own view.
12. This is not a descriptive claim; that is, we are not claiming that this is
how logicians understand what they are doing, whether collectively or
individually. Rather this is a claim about how we should understand
logical practice. In line with our anti-exceptionalism, we are approaching
the epistemological and metaphysical questions in logic in the same way
in which the philosopher of biology approaches those questions in biol-
ogy. In both cases one must attend to the messy business of how science
actually gets done. For further discussion of our methodology, see Payette
and Wyatt (2018).
13. For example, □p ⊃ p is true at every world in every model on a Kripke
frame (W, R) iff ∀w, wRw.
14. See Sen (1970). One can also consider Cozic and Hill (2015) to under-
stand the larger role of representation theorems in science.
15. See van Benthem (1997) for an overview.
16. Here we are using ‘completeness’ in the (common) sense which includes
both completeness and soundness theorems.
17. See Franks (2010) for more on interpretations of completeness.
296 G. Payette and N. Wyatt
18. You cannot add a formula that isn’t a theorem to CPL as a new theorem
without being able to then prove everything.
19. All of the admissible rules are derivable rules. For more on that distinc-
tion, see Iemhoff and Metcalfe (2009).
20. See Wyatt and Payette (2018).
21. See, for example, the discussions in McKeever and Ridge (2005), Cullity
and Holton (2002) for moral particularisms, and Lakatos (1976) and
Larvor (2001, 2008) for particularism in mathematics. John Norton’s
material theory of induction is, we think, a particularist approach to
inductive argument (2003, 2010).
22. See Kolodny and MacFarlane (2010), Dreier (2009), Cantwell (2008),
or McGee (1985) for some possible counterexamples to modus ponens.
23. See, for example, Cook (2010), Beall and Restall (2006).
24. We argue in Wyatt and Payette (2018) that attributions of logical form
are only possible from within a logic.
25. Logical particularism can be seen as a form of instrumentalism, but not
at the level of theory justification. That is, we don’t take it that a ‘correct’
logic is the most useful one. Rather we take it that the practices and
methods of formal logic are instrumentally justified because of their
effectiveness in explanation.
26. Of course one could give other interpretations of the connection mod-
eled by classical logic, and some people disagree with the classical inter-
pretation of truth preservation, see, for example, Read (1994).
27. The examples above are not very interesting sets. Each has only its parti-
tion into unit sets, and so something follows when it follows classically
from at least one of the formulas.
28. This is at least true for certain parts of the models in Castañeda (1981).
29. This is done in Searle and Vanderveken (1985).
30. a.k.a. mathematically defined logical connectives.
31. Note that we did not say that ⊢ explicates the arguments from L′. We
agree with Woody, if we have interpreted her correctly, that what is
often called ‘explication’ is simply a stage in explanation. Models, par-
ticularly mathematical ones, become part of the understanding of
phenomena.
32. This is similar, we think, to the project in Field (2015) where commit-
ments to relationships between degrees of belief are translated into logi-
cal rules. Particularism assumes that logical connections can be about
things other than degrees of belief.
Logical Particularism 297
References
Beall, J., and G. Restall. 2006. Logical Pluralism. Oxford: Oxford University
Press.
Bokulich, A. 2011. How Scientific Models Can Explain. Synthese 180 (1):
33–45.
Cantwell, J. 2008. Changing the Modal Context. Theoria 74 (4): 331–351.
Castañeda, H.N. 1981. The Paradoxes of Deontic Logic: The Simplest Solution
to All of Them in One Fell Swoop. In New Studies in Deontic Logic, ed.
R. Hilpinen, 37–85. Chichester: Wiley-Blackwell.
Cook, R.T. 2002. Vagueness and Mathematical Precision. Mind 111 (442):
225–247.
———. 2010. Let a Thousand Flowers Bloom: A Tour of Logical Pluralism.
Philosophy Compass 5 (6): 492–504.
Cozic, M., and B. Hill. 2015. Representation Theorems and the Semantics of
Decision-Theoretic Concepts. Journal of Economic Methodology 22 (3):
292–311.
Cullity, G., and R. Holton. 2002. Particularism and Moral Theory. Aristotelian
Society Supplementary Volume 76: 169–209.
Dancy, Jonathan. 2013. Moral Particularism. In The Stanford Encyclopedia of
Philosophy (Fall 2013 Edn.), ed. E.N. Zalta. Metaphysics Research Lab,
Stanford University.
Dreier, J. 2009. Practical Conditionals. In Reasons for Action, ed. D. Sobel and
S. Wall, 116–133. Cambridge: Cambridge University Press.
Dummett, M. 1991. The Logical Basis of Metaphysics. Harvard: Harvard
University Press.
Field, H. 2015. What Is Logical Validity? In Foundations of Logical Consequence,
ed. C.R. Caret and O.T. Hjortland, 33–70. Oxford: Oxford University Press.
Fine, K. 2012. The Pure Logic of Ground. Review of Symbolic Logic 5 (1): 1–25.
Franks, C. 2010. Cut as Consequence. History and Philosophy of Logic 31 (4):
349–379.
———. 2015. Logical Nihilism. In Logic Without Borders: Essays on Set Theory,
Model Theory, Philosophical Logic and Philosophy of Mathematics, ed.
A. Villaveces, R. Kossak, J. Kontinen, and A. Hirvonen, 147–166. Berlin: De
Gruyter.
Giere, R.N. 2004. How Models Are Used to Represent Reality. Philosophy of
Science 71 (5): 742–752.
298 G. Payette and N. Wyatt
1 Outlining the View
The philosophy of logic has been dominated by the view that there is One
True Logic. What is meant by ‘One True Logic’ is sometimes not made
entirely clear—what is a logic and what is it for one of them to be true?
Since the study of logic involves giving a theory of logical consequence
for formal languages, the view must be that there is one true theory of
logical consequence. What it means for a logic to be true is, roughly, for
it to correctly represent something or other.1 What do logics represent? It
is clear from the various uses of applied logic, they can represent many
different sorts of phenomena. But for the purposes of traditional logic,
A. J. Cotnoir (*)
Department of Philosophy, University of St Andrews, St Andrews, UK
e-mail: ac117@st-andrews.ac.uk
Much of the discussion in the philosophy of logic over the last decade
has been devoted to the debate between logical monism and logical plu-
ralism. But logical nihilism hasn’t been given nearly as much attention,
even though the view has historical roots and is philosophically
defensible.
Logical Nihilism 303
that logical nihilism presents us with a unified view across a broad range
of issues in philosophy of logic. I conclude (§4) by considering related
philosophical issues and sketching a general outlook on logic and formal
methods that is nihilist-friendly.
Before presenting these arguments, however, let me respond to an
immediate worry. In presenting and endorsing arguments for a view, one
ordinarily takes them as good arguments, where good arguments are (at
the very least) valid arguments. But if there is no correct theory of logical
consequence, then presumably there are no valid arguments either. Thus,
one might suspect that any attempt to argue for logical nihilism under-
mines itself.
This worry could be serious if logical nihilism entailed that there were
no valid arguments (as in Logical Nihilism 2). But notice that logical
nihilism (of both sorts) is consistent with their being standards governing
good inference in natural language. And Logical Nihilism 1 merely claims
that there is no formal theory that perfectly captures these standards. And
that’s perfectly compatible with these arguments being formally valid in
some regimented language that is adequate for more restricted purposes.
It’s also compatible with arguments conforming to inherently informal
standards on good inferential practice, or an abductive methodology in
the philosophy of logic (e.g. Williamson 2017). So the worry is
ill-founded.
2 Arguments from Diversity
The first batch of arguments for logical nihilism follows a simple recipe.
The first ingredient is an argument for logical pluralism; this establishes
that no single logic can be an adequate theory of inference. The second
ingredient is a constraint on logical consequence that rails against logical
pluralism—that is, a constraint that requires there be at most one single
adequate theory of inference. Combine these two ingredients and stir;
the result is logical nihilism. For if there cannot be one single correct
theory, and any correct theory must not be plural, then there cannot be
any correct theory at all. Fortunately for us, such ingredients are not hard
to come by.
Logical Nihilism 305
Beall and Restall (2006) suggest that the notion of logical consequence
may be analyzed by the Generalized Tarski Thesis (GTT).
Of course, we have yet to specify what sorts of cases are under consid-
eration here. Logicians are frequently concerned with models, but the
notion of consequence itself doesn’t determine that models are the only
possible option. Once this is granted, there is a straightforward argument
to pluralism.
The idea is that settled core of logical consequence (‘the intuitive or
pre-theoretic notion’ (Beall and Restall 2000)) is given by GTT. An
instance of GTT is obtained by a specification of casesx in GTT, and a
specification of the relation of being true in a case. An instance of GTT is
admissible if it satisfies the settled role of consequence, and if its judg-
ments about consequence are necessary, normative, and formal (in some
sense or other). A logic, then, is an admissible instance of GTT.
Beall and Restall contend that there are at least three admissible
instances of GTT. They defend this claim by appeal to the diverse pur-
poses logic may be put to, each of which corresponds to a different kind
of case. There are the complete and consistent cases of classical logic (i.e.
worlds), the incomplete cases of intuitionistic logic (i.e. constructions), and
the inconsistent (and/or incomplete) cases of relevant logic (i.e. situations).
As a result, there is no single correct consequence relation, but many cor-
rect consequence relations relative to which kind of cases we intend.
306 A. J. Cotnoir
Suppose we agree with Beall and Restall that there are a number of
different ways of specifying the notion of a case, each with its own speci-
fication of truth-in-a-case, and hence a number of instances of GTT. We
might still disagree that any one of these instances is admissible because
the resulting consequence relations fail to be necessary. Clearly, it’s a plau-
sible constraint on logic that it must be necessary; an argument is valid
whenever it is necessarily truth preserving. But on the surface, the neces-
sity constraint appears to require us to look at all kinds of cases, if they
really are genuine cases.
This is, in effect, a version of an objection to logical pluralism devel-
oped by Bueno and Shalkowski (2009).
Thus, on Beall and Restall’s account, none of the major families of logics
they consider satisfy the necessity constraint. By their own standard, none
of these are logics at all. Only a very weak consequence relation survives
this scrutiny, according to their accounting of the necessity constraint as
quantification over all cases. Once the partisan spirit of logical monism is
replaced with the open-minded embrace of cases suitable to alternative log-
ics, no commonly promulgated consequence relation seems to satisfy the
necessity constraint. Hence, according to their own accounting of the con-
straints on relations of logical consequence, there are no such relations.
(p. 299f )
It is worth noting that there are actually two sorts of objections here, each
resulting in a different form of logical nihilism. The first objection is that,
on an absolutely general reading of necessity—necessity in the broadest
possible sense—as quantifying over all cases of any type, no instance of
GTT counts as admissible. And hence there are no logics; this is a form
of Logical Nihilism 1.
On the other hand, we might instead specify a general notion of case
as follows: c is a case* iff c is a casex for some x. Similarly, we can give a
general specification of truth-in-a-case as follows: A is true-in-a-case* iff c
is a casex for some x and c represents A to be true-in-a-casex. This seems to
suffice for an instance of GTT. And indeed, this will constitute an admis-
sible instance especially with regard to necessity (read again in the broad-
est sense). But, if we take seriously the vast range of logics which have
Logical Nihilism 307
Another strategy along the same lines applies to Varzi’s (2002) brand of
logical pluralism. According to logical relativism, the correct inferences
depend crucially on what one takes to be logical—as opposed to the non-
logical—vocabulary. Varzi advocates for a kind of skepticism about any
objective criterion on where to draw the logical/non-logical divide.10 Of
course, one might be perfectly happy to accept that there are natural lan-
guage consequence relations for each way of drawing the divide. Indeed,
one might rather appeal to instead to ‘meaning postulates’ or ‘semantic
constraints’ as (cf. Sagi 2014) for all vocabulary to generate semantic
entailment relations. Formal semanticists often deliver such theories.
But there are general reasons why formal semantics does not deliver
the logic of natural language. Glanzberg (2015, §2.2) argues that such
meaning postulates or semantic constraints will, at best, deliver analytic
entailments which should not intuitively count as valid (e.g. ‘John cut the
bread’ lexically entails ‘The bread was cut with an instrument’ because of
constraints on the meaning of ‘cut’). This seems to dictate against count-
ing such semantic ‘entailment’ relations to be logical consequence proper.
Glanzberg (2015, §2.1) also argues that natural language semantics,
whether in its neo-Davidsonian or type-theoretic guises, are primarily
concerned with absolute semantics. That is, they give semantic clauses for
sentences expressing the meanings that speakers actually understand
those sentences to have. They do not give relative semantics, that is,
semantic clauses that are generally specified over a range of models, which
Logical Nihilism 309
The first argument from expressive limitations for logical nihilism appeals
to considerations around semantic closure.
is true with respect to classical formal languages.13 This result has led
many to reject that classical logic is the One True Logic.
But the well-known phenomenon of revenge plagues non-classical for-
mal languages as well. The literature is predictably populated with articles
that find inexpressible concepts for non-classical logics that purport to be
semantically closed. Even absent a general recipe for finding a revenge
paradox for every formal language, the sheer volume of failed attempts at
semantic closure would justify a plausible pessimistic meta-induction14: if
premise 2 has proven true for every formal language so far, then there’s
good reason to think the issue is systemic.
There are somewhat more general recipes for finding revenge problems
even for non-classical approaches to paradox. Here is a basic one: we begin
with three desiderata for any semantically closed formal language 𝕷.
that’s precisely what one disjunct of γ says, and thus it should intuitively
be true. The characterization of paradoxical sentences seems destined to
break down.
There are, of course, a number of options for replying to a revenge
charge like this. The first option is to reject that Characterization is a
desiderata at all. Of course, the characterization constraint has two direc-
tions. One might reject the ‘all’ direction, and attempt to leave some
paradoxical sentences uncharacterized. An immediate reply would be
that one has thereby failed to give a complete account of the semantic
paradoxes. On the other hand, one might reject the ‘only’ direction, and
claim that we must throw out some babies (non-rogues) when throwing
out the bathwater (rogues). Two immediate replies come to mind: first,
this will result in a situation where being a ‘rogue’ sentence is not express-
ible in 𝕷 (even if φ is); after all φ will apply to some non-rogues. Second,
being a ‘rogue’ sentence appears not to be doing any work in the resulting
theory, the characterization of paradoxes happens using φ.
One might opt for second response to revenge: reject Revenge
Immunity—that is, accept that some attributions of φ are themselves
‘rogue’. An immediate reply to this approach, however, is that since the
theory itself includes attributions of φ to the paradoxical sentences, one’s
theory relies on exactly the problematic phenomenon. That is, one is
forced to use rogue sentences in giving one’s theory.
A third response to these revenge charges would be simply to reject
Semantic Closure. This is Tarski’s solution (for formal languages) in that
he rejects any such φ is expressible. But again linguistic appearances sug-
gest the contrary for natural languages, and so this will lead us to a ver-
sion of logical nihilism.
Perhaps one might attempt to agree with Priest (1984) that this merely
shows semantic closure can be had only on pain of inconsistency; an implicit
argument for dialetheism. But Beall (2015) has outlined a general argu-
ment barring semantic closure for formal theories on pain of triviality. The
rough idea is to classify ‘rogue’ sentences of a formal semantic theory
(couched in a formal language together with its consequence relation) as
trivializer-sentences: sentences that, relative to that language’s consequence
relation, yield the theory containing all sentences of the language. Then
there can be no trivializer predicate for that theory characterizing all and
Logical Nihilism 313
Each premise is, of course, controversial so let’s look at the case for
each.17
What evidence is there for premise 1? Certain uses of natural language
quantification appear to require an unrestricted reading. For example,
part of the main function of philosophical uses of quantification is to rule
out the existence of certain objects: ‘The universal set does not exist’
should not be true if the universal set does exist but we just can’t quantify
over it. Even the rejection of unrestricted quantification seems to presup-
pose the availability of unrestricted quantification; saying ‘one cannot
quantify over absolutely everything’ appears to presuppose that there is
something that one cannot quantify over.18 Additionally, McGee (2003)
argues that natural language universal quantification would be unlearn-
able if it weren’t unrestricted.
There are also general reasons for accepting premise 2. Here is an argu-
ment that has been put forward repeatedly in the literature, based on two
key claims.
That is, any theory that rejects the above principles will be inadequate
to natural language inference. But any theory that accepts them reintro-
duces a form of Ex Falso Quodlibet.
B
A→B
¬B ¬B → ¬A
¬A
Here, on the assumption that B and ¬B, we can infer ¬A. But this is
something that simply cannot be valid in a dialetheic setting, since it
would entail the truth of (the negation of ) every sentence
whatsoever.24
There are set theories based on classical logic with a universal set:
Quine’s (1937) New Foundations, and descendants (see Forster 1992).
But Linnebo (2006, p. 156f ) notes, they are technically unappealing
and lack a unified intuitive conception. More importantly, they have a
Logical Nihilism 317
4 Related Issues
An issue I’ve largely set aside is the fact that logic is taken to be normative.
Perhaps this unreasonably privileges psycho-semantic facts about how we
in fact infer, rather than how we ought to infer. On this conception,
appeals to linguistic data and evidence for ‘semantic intuitions’ can appear
irrelevant. The question of logic is: ‘what is best inferential practice?’
But I want to stress that even if logical consequence is conceived as a
fundamentally normative relation, there might well be arguments for
logical nihilism. For example, we might contend that an objective norma-
tive relation would be metaphysically and epistemically strange. Roughly,
we might apply Mackie’s (1977) arguments to logical consequence itself.
This case is made strongly by Field (2009):
Quite independent of logic, I think there are strong reasons for a kind of
an- tirealism about epistemic normativity: basically, the same reasons that
mo- tivate antirealism about moral normativity, or about aesthetic good-
ness, ex- tend to the epistemic case. (For instance, (i) the usual metaphysi-
cal (Humean) worry, that there seems no room for ‘straightforward
normative facts’ on a naturalistic world-view; (ii) the associated epistemo-
logical worry that access to such facts is impossible; (iii) the worry that such
normative facts are not only nonnaturalistic, but ‘queer’ in the sense that
awareness of them is supposed to somehow motivate one to reason in a
certain way all by itself.) (p. 354)
Logical Nihilism 319
This is something the early analytic pioneers of logic, Frege and Tarski,
understood well. Consider Frege:
Notes
1. Of course, there are many ways one might think about truth; one needn’t
think of truth in terms of correct representation. But it is a reasonable
way of thinking about the issue, and I think it is what the intuitive slo-
gan ‘One True Logic’ is getting at.
2. This quote strongly suggests that Beall and Restall would not disagree
with the characterization of logics as being in the business of correct
representation of natural language inferential practice.
3. Compare recent authors like Bueno and Colyvan (2004), “The aim of
logic is taken to be to provide an account of logical consequence that
captures the intuitive notion of consequence found in natural language”
(p. 168). Or Resnik (2004), “As practitioners of inference we make spe-
cific inferences […] As logicians we try to formulate a systematic account
of this practice by producing various rules of inference and laws of logic
by which we presume the practice to proceed. This aspect of our work as
logicians is like the work of grammarians” (p. 179). Or consider Cook
(2010) “[A] logic is ‘correct’, or ‘acceptable’, etc., if and only if it is a cor-
rect (or acceptable, etc.) codification of logical consequence. The idea
that the philosophically primary (but obviously not only) goal of logical
theorizing is to provide a formal codification of logical consequence in
natural language traces back (at least) to the work of Alfred Tarski”
(p. 195).
4. Cook (2010, p. 495f ) gives a detailed account of what it means to say a
logical consequence relation is adequate to natural language inference. I
am assuming something like his definition is suitable for this purpose.
5. Another way of reading Mortensen is as arguing that real broad possibil-
ity outstrips pure logical possibility. In this case, then, there may be logi-
cally impossible scenarios that are not, broadly speaking, impossible.
Mortensen would then not count as a logical nihilist in the sense above.
6. Parallel disputes over the metaphysics of composition. Here universalism
states that composition always occurs, whereas nihilism claims that the
322 A. J. Cotnoir
31. E.g. Scharp (2013) who thinks defective concepts like ‘truth’ need to be
replaced, or Patterson (2009) who argues that we understand natural
language using a false semantic theory, such that strictly speaking natural
language sentences have no meanings.
32. See especially Glanzberg (2015, §IV), but also Cook (2002, 2010),
Shapiro (2006), and Scharp’s (2013) metrological naturalism.
33. This point is made clearly and forcefully in Beall (2007) with respect to
the semantic paradoxes. Cook (2002) argues for a similar perspective
with respect to vagueness.
34. I’d like to thank audiences at the Truth Pluralism and Logical Pluralism
Conference at the University of Connecticut, the Swiss Society for Logic
and Philosophy of Science at the University of Neuchâtel, the Northern
Institute of Philosophy 2011 Reading Party, the University of St Andrews
Philosophy Society, and the students in my 2014 and 2017 Philosophy
of Logic seminars. Their comments and questions led to many improve-
ments and developments in the paper. Special thanks to Colin Caret,
Roy Cook, Matti Eklund, Ole Hjortland, Michael Lynch, Julien Murzi,
Stephen Read, Gillian Russell, Gil Sagi, Kevin Scharp, Stewart Shapiro,
Keith Simmons, Crispin Wright, and Elia Zardini for discussions on
these topics over a number of years. Thanks also to an anonymous referee
for helpful comments on a previous version of the paper. The biggest
debt of gratitude is owed to my PhD supervisor Jc Beall, who disagrees
with many of the ideas in this paper. A reaction against one’s academic
upbringing can be a sign of deep respect; and I hope this paper is taken
in that spirit.
References
Bacon, A. 2015. Can the Classical Logician Avoid the Revenge Paradoxes?
Philosophical Review 124 (3): 299–352.
Beall, J. 2007. Prolegomena to Future Revenge. In Revenge of the Liar, ed.
J. Beall, 1–30. Oxford: Oxford University Press.
———. 2015. Trivializing Sentences and the Promise of Semantic Completeness.
Analysis 75: 573–584.
Beall, J., and J. Murzi. 2013. Two Flavors of Curry Paradox. Journal of Philosophy
110: 143–165.
Beall, J., and G. Restall. 2000. Logical Pluralism. Australasian Journal of
Philosophy 78 (4): 475–493.
Logical Nihilism 325
Forster, T.M. 1992. Set Theory with a Universal Set. Oxford: Oxford University
Press.
Frege, G. 2002. Begriffsschrift, a Formula Language, Modeled upon that of
Arithmetic, for Pure Thought. In From Frege to Gödel: A Source Book in
Mathematical Logic, 1879–1931, ed. J. Van Heijenoort, 1–82. Cambridge,
MA: Harvard University Press.
Glanzberg, M. 2015. Logical Consequence and Natural Language. In
Foundations of Logical Consequence, ed. C.R. Caret and O.T. Hjortland,
71–120. Oxford: Oxford University Press.
Grim, P. 1991. The Incomplete Universe: Totality, Knowledge, and Truth.
Cambridge, MA: MIT Press.
Haack, S. 1978. Philosophy of Logics. Cambridge: Cambridge University Press.
Hofweber, T. 2007. Validity, Paradox, and the Ideal of Deductive Logic. In
Revenge of the Liar, ed. J. Beall, 145–158. Oxford: Oxford University Press.
Jané, I. 1993. A Critical Appraisal of Second-Order Logic. History and Philosophy
of Logic 14: 67–86.
Lavine, S. 2006. Something About Everything: Universal Quantification in the
Universal Sense of Universal Quantification. In Absolute Generality, ed.
A. Rayo and G. Uzquiano, 98–148. Oxford: Oxford University Press.
Lewis, D.K. 1991. Parts of Classes. Oxford: Blackwell.
Linnebo, Ø. 2003. Plural Quantification Exposed. Noûs 37 (1): 71–92.
———. 2006. Sets, Properties, Unrestricted Quantification. In Absolute
Generality, ed. A. Rayo and G. Uzquiano, 149–178. Oxford: Oxford
University Press.
———. 2010. Pluralities and Sets. Journal of Philosophy 107 (3): 144–164.
Linnebo, Ø., and A. Rayo. 2012. Hierarchies Ontological and Ideological. Mind
121 (482): 269–308.
Ludwig, K. 2002. What Is the Role of a Truth Theory in a Meaning Theory. In
Meaning and Truth: Investigations in Philosophical Semantics, ed. O. Campbell,
D. Shier, and J.K. Campbell. New York: Seven Bridges Press.
Ludwig, K., and G. Ray. 2002. Vagueness and the Sorites Paradox. Philosophical
Perspectives 16: 419–461.
Lynch, M.P. 2008. Alethic Pluralism, Logical Consequence, and the Universality
of Reason. Midwest Studies in Philosophy 32 (1): 122–140.
———. 2009. Truth as One and Many. Oxford: Oxford University Press.
MacFarlane, J.G. 2002. What Does It Mean to Say that Logic Is Formal? PhD
thesis, Citeseer.
Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. New York: Pelican Books.
Logical Nihilism 327
G. Russell (*)
Department of Philosophy, University of North Carolina at Chapel Hill,
Chapel Hill, NC, USA
e-mail: gillian_russell@unc.edu
The sentence X follows logically from the sentences of the class K if and
only if every model of the class K is also a model of the sentence X. (Tarski
1936, p. 417)
The relation of implication in one fairly natural sense of the term, viz. logi-
cal implication is readily described with the help of the auxiliary notion of
logical truth. A statement is logically true if it is not only true but remains
true when all but its logical skeleton is varied at will; in other words, if it is
Varieties of Logical Consequence by Their Resistance to Logical… 335
true and contains only logical expressions essentially, any others vacuously.
Now one statement may be said to logically imply another when the truth-
functional conditional which has the one statement as an antecedent the
other as consequent is logically true. (Quine 1981, §5)10
This condition may in fact be satisfied only because the language with
which we are dealing does not possess a sufficient stock of extra-logical
constants. The condition […] could be regarded as sufficient for the sen-
tence X to follow from the class K only if the designations of all possible
objects occurred in the language in question. This assumption, however, is
fictitious and can never be realised. (Tarski 1936, pp. 415–416)
ØXx Ù Yx (1)
Here the non-logical predicates Odd and Prime have been replaced
with the predicate variables X and Y, and the individual constant 2 has
been replaced with the individual variable, x, to form a sentential
function.
336 G. Russell
Sentential functions are then satisfied (or not) by the closest thing in
Tarski’s system to a modern-day model: a sequence of objects. A sequence
of objects which would satisfy (1) above would be one which began (the
set of planets, Ceres, the set of moons, …)—assuming the ‘objects’ in this
sequence are assigned to the non-logical expressions in the order they
appear in the sentence. A sequence of objects which would not satisfy (1)
would be (the set of even numbers, 4, the set of prime numbers). Such
sequences can easily be reconceptualised as functions from the non-
logical expressions to suitable meanings, and hence as an analogue to our
modern-day interpretations.
Another difference between Tarski’s definition and contemporary ones
is that Tarski’s models have no counterpart for the contemporary domain
of quantification. As Etchemendy notes, domains are a very natural addi-
tion to a model if you are thinking of consequence representationally—
we can represent worlds with fewer or more objects—but seem
unmotivated on the interpretational account Tarski favoured12; we
wouldn’t expect a reinterpretation of the non-logical expressions to affect
the number of objects there are. Still, we normally think that variable
domains are important for achieving the right extension for ‘logical con-
sequence’—we wouldn’t want ∃x ∃ y(x ≠ y) (or any other sentence equiv-
alent to the claim that there are at least n objects for some n ∈ ℕ) to be
true in every model (and thus a logical truth) or false in every model (and
thus a logical falsehood). If we think of models representationally, this
naturally motivates the inclusion of a domain, which then allows us to
have variable domain models, resulting in (what we normally think of as)
a sensible extension for the consequence relation.
On the other hand, the representational conception of a model looks
less natural—and the interpretational conception more natural—when it
comes to thinking about why two names, say a and b, might be assigned
the same referent in one model and a different referent in another. On the
interpretations account, this simply represents an alternative semantic
value for at least one of the names, a reinterpretation like any other. Just
as 2 could be a name for Julius Caesar, so Hesperus could be a name for
Mars. On the representationalist account—where we assume that the
semantic value of a name is an object—this looks much harder to explain.
If a refers to Venus and b refers to Venus, it is hard to see how we can
Varieties of Logical Consequence by Their Resistance to Logical… 337
retain those interpretations (as the spirit of the account instructs us to do)
and somehow—by changing the world—end up with a model in which
I(a) ≠ (b).
Here is where we are: we have three varieties of logical consequence:
interpretational, representational, and substitutional. On the interpreta-
tional account, we vary the interpretations of the non-logical expressions
in an argument, and if every way of doing so makes the conclusion true
if the premises are all true, then the conclusion is a logical consequence
of the premises. This view is the historical ancestor of contemporary
model theory, and it makes some limited sense of that model theory, but
it is hard to motivate contemporary models containing variable domains
from this perspective. On the substitutional account, we keep the logical
expressions fixed, and substitute syntactically appropriate alternative
expressions uniformly throughout the argument. If doing so can result in
true premises but a non-true conclusion, the argument is not valid. This
account was endorsed by Quine, but makes the extension of the relation
of logical consequence hostage to the richness of the language. On the
representational account we retain the interpretation of all expressions in
the argument, but consider various different states the world might be in.
If every state that makes all the premises true makes the conclusion true,
then the conclusion is a logical consequence of the premises. This moti-
vates the variable domains of contemporary models quite naturally, but it
is hard to see why different models should be allowed to assign different
interpretations to expressions which actually share an interpretation
(which they are permitted to do in contemporary model theory). Now it
only remains to introduce the fourth and final conception of logical con-
sequence that I will consider.
a=a "x ( x = x )
Fa Ú ØFa "X"x ( Xx Ú ØXx )
Af ® f "X ( AX ® X )
endorsing particular views on these matters. All we need to say for now is
that propositions are the values of sentential variables, much as objects
are the values of first-order variables, and similarly properties are the val-
ues of predicate variables.
Now we generalise to logical consequence. This must be done with
some care since there is no obvious way to close an argument’s shell using
quantifiers (we normally quantify into sentences, not arguments). So
consider an argument with premises, such as modus tollens: P → Q,
¬Q ⊨ ¬ P. Call the result of uniformly replacing the non-logical vocabu-
lary with variables throughout the argument the argument’s shell. The
shell for modus tollens is X → Y, ¬Y ⊨ ¬ X. Then we say that the conclu-
sion of the argument is a logical consequence of the premises just in case
every assignment of appropriate values (objects to first-order variables,
properties to predicates, etc.) that makes both X → Y and ¬Y true also
makes ¬X true. Here is another illustration, this time with a first-order
case:
a=b
Fa
Fb
x=y
Xx
Xy
A ® B, A B ( MP )
A Ú ØA ( LEM )
AÙB A ( ÙE )
Ø(A Ù ØA ( PNC )
A, ØA B ( Explosion )
A A ( ID )
A ® B, B A ( AC )
Call sentences of this form E-sentences (‘E’ for entailment). Many log-
ics reject one or more of these putative laws. Classical logic rejects (AC),
paraconsistent logics reject (Explosion), Strong Kleene and intuitionistic
logics deny that (LEM) is a law. Logical nihilism rejects them all, includ-
ing relatively uncontroversial principles, such as conjunction elimination
Varieties of Logical Consequence by Their Resistance to Logical… 341
and identity. More comprehensively, for any set of sentences K and any
conclusion X, the nihilist is committed to:
K X
What this amounts to will depend on what logical consequence is, that
is, on what ⊨ means. Different conceptions of consequence correspond
to different versions of logical nihilism. Representational nihilism is the
view that for any argument, there is always a case—roughly, a state of the
world—in which the premises are true and the conclusion is not.
Interpretationalist nihilism is the view that in any argument there is always
an interpretation of the non-logical expressions which makes the prem-
ises true and the conclusion not. Substitutional nihilism is the view that
there is always a way to substitute expressions for the non-logical expres-
sions in an argument which makes the premises true and the conclusion
not true. And universalist nihilism will be the view that there is always an
assignment of values to the non-logical expressions in an argument which
makes the premises of an argument’s shell true without making the con-
clusion true.
Nihilism, then, is relatively simple to state and understand, even if, like
logical consequence, it comes in several flavours. But one might wonder
why anyone would entertain such an apparently wild view. My goal in
this section will not be to convince you of nihilism itself, but rather to
convince you that the view is a live option, so that you might be inter-
ested in safeguards against it.15
One of the reasons it is a live option stems from the extreme generality
of logical laws. Claims that are very general are, in one sense, very strong,
and hence take much to establish and little to refute. For example, it is
not sufficient for the truth of modus ponens that the truth of
snow is white
and
guarantees that of
grass isgreen
cut
Varieties of Logical Consequence by Their Resistance to Logical… 343
c orrect. And if we don’t have proofs of the logical laws, then we don’t have
proofs that logical nihilism is false.
(GTT) An argument is valid x if and only if, in every case x in which
the premises are true, so is the conclusion.
The objection draws on a view of logic that is both intuitive and has his-
torical precedent, namely, that logic is supposed to be completely general;
if an argument K ⊨ X preserves truth in (say) arithmetic but not in geom-
etry, or in classical cases but not in dialethic ones—then it is not really a
law of logic after all. Logic is supposed to work in absolutely all cases.
Beall and Restall respond with a reductio ad absurdum on the assump-
tion that logic is completely general: they say that this assumption will
lead to absurdity—in particular to logical nihilism, or some similar
absurdly weak logic:
Elsewhere I’ve argued that not even identity would remain (Russell
2017), but one might think that this is not of the greatest importance.
Most philosophers and logicians regard both all-out logical nihilism and
a view on which only identity is left (or some other very minimal set of
E-sentences—call such a view logical minimalism) as absurd. They think
that if your philosophy of logic leads you to either logical nihilism or
logical minimalism, it must be wrong.
Though discussions of pluralism have tended to assume a case-based
(and so representationalist) approach to consequence, this is inessential to
the debate. A structurally similar issue—nihilism begot by generality—
346 G. Russell
3 Modes of Resistance
So much for the big picture: a commitment to generality in logic seems
to push us towards nihilism, or at least towards weaker and weaker logics.
Still, when we look at the details, the availability, plausibility, and persua-
siveness of certain kinds of counterexample vary with our conception of
consequence. We are now in a position to explore some of this variation
in the third and final section of this chapter. In the coming subsections, I
will look at the issues associated with empty names, vagueness, and self-
Varieties of Logical Consequence by Their Resistance to Logical… 347
Empty Names
Classical logic does not model empty names. In classical model theory, an
interpretation must assign an extension to every individual constant.
Natural languages have empty names—Pegasus, Santa, and so on—and
we might want to know about entailment relations on sentences which
contain them. In particular, a commitment to generality in logic could
lead us to ask what will happen to the logic if we consider all names,
including the empty ones.
What will happen depends on one’s conception of logical consequence.
For example, on the substitution-conception, we are supposed to be able
to substitute any name uniformly for another name throughout a valid
argument without losing validity. But if we substitute Pegasus for a
throughout this classical validity:
a = a $x ( x = a ) (2)
is a not condition that a case can meet or fail to meet if ‘Pegasus’ has no
meaning, any more than making the sky dfsndfnsf is a condition that
something can meet or fail to meet.
But perhaps you think it is too quick to say that the name Pegasus has no
meaning. It could be associated with some descriptive condition—a
sense—that, as it happens, isn’t satisfied, but could be in some cases. Perhaps
you want to say that Pegasus doesn’t exist, but that there are still models on
which Pegasus flies is true, namely, models in which something satisfies the
condition associated with Pegasus, and that thing flies. Such a conception
of names could still be associated with interpretation functions that assign
objects to names, but we might expect that names would be permitted to
lack a referent, that is, the requirement that the interpretation function be
a complete function over the set of individual constants would be dropped.
This is consistent with the representational approach to models. Just think
of the possible changes to the states of the world as including the deletion
or inclusion of the referents of names, and we keep the language the same
by retaining the name’s descriptive sense (which is ignored in the model).
This isn’t my view of names, for broadly Kripkean reasons. But if you like
this view, then you might think that the representational approach is as
threatened by the existence of empty names as the interpretational one.
To sum up the results of this section, empty names motivate a decisive
weakening of the logic in the case of the substitutional conception of
consequence, but not on the worldly conception. Meanwhile defenders
of the interpretational and representational conceptions can make moves
to avoid admitting incomplete interpretation functions on names. The
defender of the interpretational conception would deny that ‘empty’ is a
legitimate interpretation of a name (because interpretations of names are
referents), and the defender of the representational conception would
deny that empty names can be used to place conditions on cases. It is
perhaps worth noting, first, that both these ways of avoiding weakenings
have a decidedly Millian flavour, and second, that on the conceptions of
logical consequence that permit us to avoid weakening in the face of
empty names, we do this by finding a way of thinking of logic and/or
names that allows us to ignore the phenomenon. This is somewhat unsat-
isfying, and this is a point that I will return to at the end of this chapter.
For now, let’s look at a new threat.
350 G. Russell
But one might think that the extensions of vague predicates in natu-
ral languages are incomplete. On one version of the idea, the conven-
tional meanings associated with our colour predicates determine a range
of shades which are red, and a range of shades which are not red, but do
not legislate on all of the shades which fall in between. We might imag-
ine a series of colour cards progressing gradually from the brightest of
reds at one end to the pinkest of pinks at the other. (Represented in
black and white below.) Anything which is pink is not red, but while
there are shades on one end which fall in the conventional extension of
‘red’, and shades at the other end which fall in the extension of ‘pink’
(and so in the anti-extension of ‘red’) there are shades in the middle
about which our linguistic conventions say neither that the shade is red,
nor that it is not red.
red not-red
|P|− (the set of things that ‘are not P’) though these two sets need not
exhaust the domain of the model.19 Then if |a| ∈ |P|+, Pa is true, if
|a| ∈ |P|−, then ¬Pa is true, but if |a| is in neither the extension nor the
anti-extension of the predicate, then both Pa and ¬Pa are neither true nor
false—a third truth status, which I will abbreviate to neither.
To argue for a third truth-status is not yet to have made a case for a
non-classical logic, because a philosopher who believes in three truth-
statuses could still hold that all and only the classical entailments are
valid. But there is a well-worn path from a third truth-status to a non-
classical logic. For if Pa is neither on some interpretations, then what of
super-sentences containing it, such as ¬Pa be or Pa ∨ ¬ Pa? One approach
adopts the Strong Kleene truth-tables, on which ¬Pa is neither if Pa is and
a disjunction takes neither if both disjuncts do. Assuming that a logical
truth is true on all interpretations we have a counterexample to the clas-
sical law of excluded middle when Pa gets the neither truth status.20
A Ú ØA (3)
Self-Reference and Overdeterminacy
extension and the anti-extension of ‘F’, making both the sentence and its
negation true.
heterological heterological
palindrome meaningful
long English
A,Ø A B (4)
4 Conclusion
Things have become quite complicated, but this is only, I think, because
there are four varieties of logical consequence on the table, and some of
these offer more than one option when confronted with a linguistic
expression which does not meet the standard assumptions of classical
logic. The beginnings of a pattern are emerging from this complexity, so
let me try to draw that out. You can take logical consequence to be a rela-
tion that quantifies over various things: expressions, meanings of various
kinds, or over bits of the world (objects and properties and the like.)23
The substitutional view takes it to quantify over expressions, the univer-
salist view takes it to quantify over bits of the world, and the interpreta-
tionalist view can either be understood as quantifying over meanings
or—if you take the line that a meaning just is a bit of the world (this is
what I’ve been calling the ‘Russellian’ take)—over bits of the world. The
356 G. Russell
Notes
1. For example, Beall and Restall (2006, p. 92) and Russell (2016).
2. A note on terminology: in this chapter I will follow the common practice
of treating the expressions ‘(logical) consequence,’ ‘entails,’ and ‘valid’ as
intertranslatable—an argument is valid just in case its conclusion is a
358 G. Russell
(logical) consequence of its premises, and this is so just in case the premises
together entail the argument’s conclusion. Hence the different views of
logical consequence that I will be looking at are just as much different
views of validity and different views of entailment. There are two other
words that it might be tempting to use here: implication (perhaps being
Quineanishly careful to use this for entailment rather than for a
conditional) and inference. But following Harman (1986) I take inference
to be a different topic altogether, and I’ll be cautious about using ‘impli-
cation’ because I think Quine’s suggested regimentation (i.e. restricting
its use to talk of the entailment relation) is less entrenched than would
be ideal for successful communication (Quine 1966, pp. 165–166;
Quine 1981, §5).
3. For example, in the title of Etchemendy (1999). I recognise that
Etchemendy’s word ‘concept’ is fighting talk in some philosophical circles.
I mean to talk—as Etchemendy did—about different views of what fea-
tures an argument has to have to be valid and have no special commitment
to using the word ‘concept’ or to any particular construal of that phrase.
4. Sometimes people respond to this problem by distinguishing different
kinds of necessity. 2 + 2 = 4 and Hesperus is Phosphorus, they might
claim, are metaphysically necessary, but not logically necessary, and it is
the logical modality in terms of which logical consequence is defined.
There are two problems with this response. The first is that it replies on
a controversial claim about the kind of necessity possessed by the propo-
sitions expressed by these sentences. In, for example, Kripke’s modal
argument of the necessity of Hesperus is Phosphorus, the ‘□’ that that
sentence inherits is the very same one applied to Hesperus is Hesperus
(Kripke 1980). The second problem is that the account threatens to be
circular. What is logically necessary if it is not the things which hold in
virtue of logic alone—the logical truths. But this assumes that we already
have an account of logical consequence.
5. Kaplan (1989). Here A is the actuality operator, N the now operator,
and α a singular term, so that informal instances of these sentences might
be Actually it is snowing if, and only if, it is snowing., if it is snowing now,
then it is snowing, I am here now and dthat[the shortest spy]=the shortest spy.
6. An additional problem for the modal slogan is that: necessity and logical
truth come apart in the model theory for modal logics. □p may be true
at some points in a model and false at others. This would make no sense
if logical truth were necessity—necessity might be world relative but
logical truth is not.
Varieties of Logical Consequence by Their Resistance to Logical… 359
15. I’ve looked in more depth at one argument for logical nihilism here:
(Russell 2017)
16. All of our conceptions of logical consequence agree that this is necessary
for logical consequence, though some may require more than this.
17. Admittedly, that argument was intended to be a reductio on someone
else’s view. But one person’s reductio ad absurdum is another person’s
argumentum ad absurdum.
18. There are similar lines of argument in Read (2006, p. 61) and Bueno and
Shalkowski (2009, p. 11).
19. The problems of higher order vagueness make it clear that this is not suf-
ficient to accommodate and explain vagueness, but there are still many
views on which the move is regarded as necessary.
20. And in fact Strong Kleene logic provides counterexamples to all the clas-
sical logical truths.
21. We leave to the previous section discussion of what happens if a does not
exist.
22. That is, The Liar sentence is true and the Liar sentence is not true or
‘Heterological’ is heterological and ‘heterological’ is not heterological.
23. And one of the reasons things are especially complicated is that the most
familiar views can be understood as quantifying over either meanings or
bits of the world.
24. It would be especially interesting in future work to examine this pattern
against the phenomenon of context-sensitivity. Context-sensitivity is a
paradigmatically linguistic, rather than worldly, issue, but there are still
views (such as MacFarlane’s non-indexical context-sensitivity) which
allow it to go deeper than others. Moreover, context-sensitivity is known
to have interesting non-classical effects on logic.
References
Beall, J., and G. Restall. 2000. Logical Pluralism. Australasian Journal of
Philosophy 78: 475–493.
———. 2006. Logical Pluralism. Oxford: Oxford University Press.
Bueno, O., and S. Shalkowski. 2009. Modalism and Logical Pluralism. Mind
118: 295–321.
Etchemendy, J. 1990. The Concept of Logical Consequence. Cambridge: Harvard
University Press.
Varieties of Logical Consequence by Their Resistance to Logical… 361
1 Introduction
In recent years various sorts of pluralisms have come into vogue in the
philosophy of logic. For present purposes, two in particular stand out:
• Logical Pluralism: The claim that two (or more) distinct logics are cor-
rect (or legitimate, or best).
• Alethic Pluralism: The claim that two (or more) distinct accounts of
truth are correct (or legitimate, or best).
R. T. Cook (*)
Department of Philosophy, University of Minnesota, Minneapolis, MN, USA
e-mail: cookx432@umn.edu
pluralism, not two. If, however, the two sorts of pluralism are indepen-
dent of each other, then we truly have two distinct kinds of pluralism—
that is, we have a plurality of pluralisms. The purpose of this chapter is to
argue that domain-specific alethic pluralism does not entail domain-
specific logical pluralism (contrary to Lynch (2008, 2009) and Pedersen
(2014)), nor does domain-specific logical pluralism entail domain-
specific alethic pluralism, and hence we do have such a plurality of
pluralisms.4
To accomplish this, in Sect. 2 I will present a simple model that shows
both how one might be a domain-specific pluralist about logic while
being a monist about truth and how one might be a domain-specific
pluralist about truth while being a monist about logic. I will then, in
Sect. 3, use this model to identify the fallacy in the arguments found in
Lynch (2008, 2009) and Pedersen (2014).5 Section 4 will then further
flesh out the model, distinguishing between different senses in which a
domain might be epistemically constrained. Finally, in a short conclud-
ing section, I will tie up some loose ends and gesture in the direction of
the aforementioned difficulties in extending the present account to
domain-independent pluralisms (and combinations of domain-specific
and domain-independent pluralisms).
Before moving on, a methodological note is in order: Although this
chapter is a response to arguments found in Lynch (2008, 2009) and
Pedersen (2014), the arguments made below do not only apply to ver-
sions of domain-specific alethic pluralism that take truth to be a func-
tional concept manifested by different lower-level properties in
different domains. The arguments are much more general than this,
applying to any view according to which truth obeys different prin-
ciples in different domains, regardless of whether that is a result of
their being different sui generis truth properties in different domains,
or a single truth property that behaves differently in different domains,
or anything in between (e.g. the various accounts of domain-specific
alethic pluralism inspired by Wright (1992)). In short, the argument
is a logical one, and is independent of much of the metaphysical theo-
rizing that is a requisite of spelling out this-or-that particular brand of
alethic pluralism in detail.
368 R. T. Cook
2 Three Theories
Our argument for the independence of domain-specific alethic pluralism
and domain-specific alethic pluralism (i.e. for a plurality of pluralisms)
will proceed by constructing a philosophical model—one consisting of
three distinct domains with differing logics and differing accounts of
truth—and showing that one substructure of the model satisfies domain-
specific alethic pluralism but not domain-specific logical pluralism, while
another substructure of the model satisfies domain-specific logical plural-
ism but not domain-specific alethic pluralism. In order to construct such
a model, we need to consider three ingredients:
Our model will require three distinct domains, which we shall simply
call D1, D2, and D3. Once we have carried out a bit more of the formal
work, there will be much more to say about what sorts of domain D1, D2,
and D2 are.
Now, for each domain, we need to decide on a logic that governs that
domain, and on a theory of truth that governs that domain. First, for
concreteness, let’s assume that we are working in some standard formal
language LPA+T(x) for arithmetic that contains6:
∀, ∃, ¬, ∧, ∨, →
0, s ( x ) , +,×
Following standard usage, we will call this theory CT↾.9 This is our
first theory of truth. Our second theory of truth, which we shall call
CT↾+, is simply CT↾ supplemented with the Arithmetic Law of Bivalence:
(
Biv PA : ( ∀x ) Sent PA ( x ) → (T ( x ) ∨ T ( ¬
x )) )
We will assume that CT↾ is the theory of truth that applied to D1, and
CT↾+ is the theory of truth that applies to D2 and D3. We can sum all of
this up in the following table:
Logic Truth theory
D1 H CT↾
D2 H CT↾+
D3 C CT↾+
intuitionistic logic H, BivPA entails that excluded middle holds for each
sentence in LPA, but it does not entail that excluded middle holds for all
sentences in LPA+T(x). Another way of putting this is as follows: We do not
have:
For all Φ ∈ LPA+T(x):
CT + H Φ iff CT + C Φ
although we do have:
For all Φ ∈ LPA:
CT + H Φ iff CT + C Φ
We need not delve into these details too deeply, however, since for our
purposes the crucial consequence of truth being superwarrant in a par-
ticular domain is that truth is epistemically constrained in that domain—
that is, the following principle of epistemic constraint for truth holds for all
Φ in the language appropriate for that domain:
ECT: If T(‘Φ’) then it is feasible to have a warrant for Φ. (see Lynch (2008,
134))
If this is right, then truth could be correspondence, but the world might
not be determinate enough (in the relevant sense) to guarantee that, for
each sentence Φ, either Φ corresponds to the way the world is or Φ fails
to correspond to the way the world is. As a result, the inference from
374 R. T. Cook
The key to seeing how these observations are compatible with the
independence of domain-specific alethic pluralism and domain-specific
logical pluralism lies in how, exactly, we understand “hold”.
As the subscripts I use in the above reconstruction of the Lynch/
Pedersen argument indicate, the epistemic constraint argument presup-
poses that, with respect to correspondence and superwarrant, there are
only two types of domain possible:
• Those like D1, where the logic is H and (at least some) instances of
excluded middle fail to hold.
• Those like D3, where the logic is C and (all) instances of excluded
middle hold.
… why not say that [intuitionistic logic] is not only the default logic, it is
the only logic, and that domains whose logic appears classical only do so
because we are employing additional principles such as Bivalence which
aren’t part of the one true logic.
Alethic pluralism … is certainly consistent with this suggestion. But the
suggestion comes at a price. According to the suggestion, [intuitionistic
logic] holds in all domains. Bivalence is not recognized as a logical princi-
ple by [intuitionistic logic]. Therefore, in every domain, Bivalence is not
recognized as a logical principle. Therefore in domains which, according to
this suggestion, nonetheless appear classical – and therefore abide by biva-
lence – Bivalence must be true for some non-logical reason. And one might
wonder what that reason might be.
A full assessment of this suggestion, therefore, requires drawing the
boundaries of logic, an issue well beyond the scope of the current essay.
But one small point is worth making: it seems natural that if one domain
allows some inferences as valid and another does not, they have different
logics. And domains where Bivalence holds will allow some inferences as
valid that other domains (which don’t sanction Bivalence) will not. So
the natural thought is that they have different logics. (Lynch 2009,
103–104)
There is a lot that needs unpacking and clarifying here, and in particular,
we are going to do some of the boundary-drawing with regard to logic
that Lynch places outside the scope of his own project. As we shall see,
this will allow us to easily identify the reason that Bivalence (or some-
thing like it) might be true of a domain without being a logical truth
(with respect to that domain).
Lynch’s argument against the truth-but-not-logical-truth of Bivalence
can be reconstructed as follows13:
First, Bivalence is not a logical principle in the relevant sense. The dif-
ference between classical logic and intuitionistic logic is not the validity
of Bivalence, it is the validity of (all instances of ) excluded middle.
Bivalence is a metalogical, semantic principle—one that utilizes the truth
predicate—and hence Bivalence is not a logical truth (not even for the
classical logician). I have argued at length for a similar claim about the
Tarskian T-schema in Cook (2012)—the same argument applies here. To
put this point bluntly, a logical principle is either (i) a sentence that con-
tains only logical vocabulary, or (ii) a schema that contains only logical
vocabulary and metavariables.14 The truth predicate T(x) is not a piece of
logical vocabulary (nor, obviously, a metavariable)—hence Bivalence is
not a logical principle (again, see Cook (2012) for more discussion).
But perhaps the first premise in our reconstruction of Lynch’s argu-
ment, more carefully formulated, is that Bivalence entails a logical prin-
ciple—namely, the logical schema known as excluded middle—and that
if Bivalence holds of a domain D, then excluded middle is true in D.15
Hence, we could reformulate the argument as:
This still won’t do, however. The reason is simple: Even if we reformu-
late (1) in this way, premise (2) is still false.
We should remember that what is at issue here is logical validity—not
merely truth (or even necessary truth). Tarski provided us with a useful
first approximation of what we mean by logical consequence (and hence
of logical validity as the special empty-premise case of logical
consequence):
Consider any class ∆ of sentences and a sentence Φ which follows from the
sentences of this class. From an intuitive standpoint it can never happen
that both the class ∆ consists only of true sentences and the sentence Φ is
false. Moreover, since we are concerned here with the concept of logical,
i.e. formal, consequence, and thus with a relation which is to be uniquely
determined by the form of the sentences between which it holds … the
Pluralism About Pluralisms 377
( ∃f ) ( ∀x ) ( ∃y ) ( f ( x ) = f ( y ) → x = y ) ∧ ( ∃x ) ( ∀y ) ( ¬f ( y ) = x )
expressing the fact that the mathematical universe is Dedekind infinite.
But the vast majority of people holding such views will reject that this
claim is logically true.16
378 R. T. Cook
Now, why does the intuitionist deny the logical truth of excluded mid-
dle even in cases when the domain in question is decidable? Why does the
(second-order) platonist deny the logical truth of the claim that the uni-
verse is Dedekind infinite? Tarski has already given us the answer: In each
case, the principle in question is true, and even necessarily true, but the
principle is not guaranteed to be true merely in virtue of its form. On the
contrary, in the first case we need to carry out a proof of the decidability
of the theory of the domain in question in order to see that all instances
of excluded middle are true, and in the latter case we need to argue that
platonism is, in fact, true (and, in addition, that at least one true mathe-
matical theory has an infinite domain).
Hence, anyone who wants to accept something along the lines of
domain D2 needs to only argue that Bivalence is true of the domain in
question, but is not true (on that domain) in virtue of its logical form.
Perhaps it is true because we have a (non-logical) warrant for the domain
being decidable, or perhaps it is true because we have some sort of meta-
physical guarantee that the domain in question is determinate in certain
respects (along the lines of (Tennant 1996)). Or perhaps we have some
other sort of non-logical guarantee that Bivalence, and hence all instances
of excluded middle, is true of the domain. None of these would in any
way imply that excluded middle is a logical truth, and none of these are
particularly implausible. Hence, the epistemic constraint argument fails,
and we do indeed have a plurality of pluralisms.
The first thing to notice is that ECT entails ECLT: If Φ is a logical truth,
then it is a truth, and so by ECT it is feasible to have a warrant for Φ. But
there seems to be no reason to think that ECLT entails ECT.
Let’s assume that, if ECT holds, then the right logic for the domain in
question is H. The reason is simple and follows the (correct core of ) rea-
soning found in Lynch (2008, 2009) and Pedersen (2014): If logical
truth is epistemically constrained in a domain D, then it would seem that
there is no guarantee that we can know, of every sentence Φ about D, on
purely logical grounds, that either Φ is true or ¬Φ is true. But, of course,
as discussed in the previous section, there might be other non-logical
grounds on the basis of which we can conclude that Bivalence holds, and
hence (if our theory of truth is powerful enough otherwise, as CT↾+ is)
380 R. T. Cook
that all instances of excluded middle (for sentences about the domain, at
least—although not necessarily for sentences involving the truth predi-
cate T(x)) are true.
This leaves three remaining combinations, corresponding to each of
the domains constructed in Sect. 218:
1. Both ECLT and ECT hold of D. Then the logic of D is H, and the
theory of truth should not include BivLD. Hence, the domain is (in
relevant respects) like D1.
2. ECLT holds but ECT does not hold of D. Then the logic of D is H, and
the theory of truth should include BivLD. Hence, the domain is (in
relevant respects) like D2.
3. Neither ECLT nor ECT holds of D. Then the logic of D is C, and the
theory of truth should include BivLD. Hence, the domain is (in rele-
vant respects) like D3.
5 Concluding Remarks
In the previous section, we have seen that, in principle, at least, domain-
specific alethic pluralism is independent of domain-specific logical plu-
ralism, and we also noted that the one simple way to understand this
phenomenon is by noting that logical truth being epistemically con-
strained does not entail that truth is likewise epistemically constrained.
In short, with respect to the domain-specific versions of the two types of
pluralism at issue, we do have a pluralism about pluralisms. We’ll con-
clude the chapter by examining whether similar stories might be told
Pluralism About Pluralisms 381
Notes
1. Further, other sorts of pluralisms—such as metaphysical pluralisms—
also fall on one or the other side of the domain-specific/domain-inde-
pendent divide.
2. See Edwards (2012) and Pedersen and Wright (2013a, b) for good sur-
veys of alethic pluralism and Cook (2010) and Russell (2016) for good
surveys of logical pluralism.
3. On the contrary, I am rather unsympathetic to domain-specific logical
pluralism, and rather sympathetic to domain-independent logical plu-
ralism, despite the good efforts of (Shapiro 2015). And I am wholly
unsympathetic to either sort of alethic pluralism. For some of that story,
see (Cook 2014).
4. It is important to note that I am not arguing that we have a plurality of
true pluralisms (see the previous footnote). The point is that the accounts
in question—domain-specific alethic pluralism and domain-specific
logical pluralism—are distinct positions, regardless of whether they are
correct or not.
5. I am, in what follows, going to concentrate on the argument as given in
Lynch (2008, 2009). Pedersen (2014) characterizes itself as presenting a
slightly more detailed and cleaned up version of Lynch’s argument (and
extending the conclusions to metaphysical pluralism, which is interest-
ing but orthogonal to our concerns here). I would like to note that
Pedersen’s essay was extremely helpful in sorting out how, exactly, the
argument in question is supposed to work.
6. We need both quantifiers and all four propositional connectives in
LPA+T(x) if we are to allow the logic in question to be full intuitionistic
logic, since the familiar equivalences that allow for various (classical)
definitions of one operator in terms of another fail intuitionistically.
7. For an argument that the truth predicate is non-logical, see (Cook 2012).
We will return to discuss the relevance of this fact in detail below.
8. A brief explanation of the notation in these axioms:
16. A slightly snarkier way of making the point: If the truth, or even the
necessary truth, of a logical principle entailed its logical truth, and if
second-order logic is logic, then any platonist would presumably also be
some sort of logicist.
17. I certainly find them interesting, even if, as I noted in endnote 3, I ulti-
mately think neither of them is correct.
18. Here LD is the language about domain D—that is, all sentences about D
not including the truth predicate T(x), hence BivLD is Bivalence restricted
to the sentences in LD.
19. This, despite the fact that I think that the combination of domain-
independent logical pluralism and alethic monism (i.e. option 6) is the
right option. The point is that I don’t think the other options are incoher-
ent—I just think that they are wrong.
20. Thanks are owed to Nikolaj J. L. L. Pedersen, Nathan Kellen, and an
anonymous referee for helpful comments on early versions of this
material.
References
Beall, J., and G. Restall. 2000. Logical Pluralism. Australasian Journal of
Philosophy 78: 475–493.
———. 2001. Defending Logical Pluralism. In Logical Consequence: Rival
Approaches. Proceedings of the 1999 Conference of the Society of Exact Philosophy,
ed. J. Woods and B. Brown, 1–22. Stanmore: Hermes.
———. 2006. Logical Pluralism. Oxford: Oxford University Press.
Cook, R. 2010. Let a Thousand Flowers Bloom: A Tour of Logical Pluralism.
Philosophy Compass 5 (6): 492–504.
———. 2012. The T-schema Is Not a Logical Truth. Analysis 72 (2): 231–239.
———. 2014. Should Anti-realists Be Anti-Realists About Anti-realism?
Erkenntnis 79: 233–258.
Edwards, D. 2012. Pluralist Theories of Truth. In Internet Encyclopedia of
Philosophy, ed. J. Fieser and B. Dowden. http://www.iep.utm.edu/plur-tru/
Lynch, M. 2008. Alethic Pluralism, Logical Consequence, and the Universality
of Reason. Midwest Studies in Philosophy 32: 122–140.
———. 2009. Truth as One and Many. Oxford: Oxford University Press.
Pedersen, Nikolaj J.L.L. 2014. Pluralism × 3: Truth, Logic, Metaphysics.
Erkenntnis 79: 259–277.
386 R. T. Cook
1 Introduction
There is a natural, if inchoate, intuition that truth might have a variable
nature, depending on what the subject matter is. Given the great differ-
ences between natural science, mathematics, ethics, and aesthetics, does
it not stand to reason that ‘Electrons have negative charge,’ ‘1 + 1 = 2,’
‘Torture is wrong,’ and ‘Duck Soup is funny’ are all true, but the truth of
each is a different property from the truth of the others?
The arch-deflationist W. V. Quine had the intuition,1 but it was Crispin
Wright (1992) who brought the prospects of pluralism about truth into
contemporary discussion. These days, a variety of different ways of work-
ing out the details of a pluralist conception of truth are on offer. By
alethic pluralism (AP), I mean the view that there are different, but sub-
stantial, truth-properties in different discourses.
C. B. Wrenn (*)
Department of Philosophy, The University of Alabama, Tuscaloosa, AL, USA
e-mail: cwrenn@ua.edu
2 Alethic Pluralism
The distinctive claim of alethic pluralism is that truth is (in some sense of
‘is’) a different property for claims in different discourses or domains of
inquiry—but truth is also more substantial than deflationists claim it to
be. Pluralists differ in their understanding of the ‘is’ in their distinctive
claim. Some take it to be a matter of identity: the truth predicate desig-
nates different properties when applied to claims from different discourses
(Wright 1992, 2013). Some take it to be a matter of realization or mani-
festation: the property playing the role of truth in one discourse need not
be the same property as what plays the truth-role in another (Lynch 2001,
2009).
Not only can alethic pluralists differ on what it means to be truth for
a discourse, but they can differ on which properties are truth in which
discourses. One pluralist might think correspondence is truth in mathe-
matics, while another thinks mathematical truth is something epistemic,
for example. A main motivation for pluralism is the idea that realism
seems appropriate for some discourses (such as physical science) but not
others (such as aesthetics). Let us call a property a truth-candidate if it is
the sort of property pluralists think might be truth for some discourse or
other. Two prominent truth-candidates are robust correspondence and
superwarrant. A claim robustly corresponds to the world when it repre-
sents the world as being a certain way, the world is that way, and the
world’s being that way is independent of our actual or possible possession
of evidence that that is how it is. (Henceforth, I will omit ‘robust’ and
refer to this as ‘correspondence.’) The notion of superwarrant comes from
Wright’s notion of superassertability (1992). A claim is superwarranted
when there is a stage of inquiry in which it is warranted without defeat
and it would remain that way in all successive stages of inquiry (Lynch
2009, p. 38; see also Pedersen 2014).
Plausibly, something like correspondence is truth in an area such as
physics, chemistry, or everyday discourse about middle-sized dry goods
and their determinate properties. Superwarrant is a plausible contender
in areas where truth seems evidentially constrained. For example, no joke
390 C. B. Wrenn
could be secretly funny—funny despite the fact that no one could know.
Truth and knowability walk together in discourse about what is funny,
and one might suppose the truth about humor must coincide with what-
ever our most considered, unimprovable judgments would say about it.
Mix
Mix’s conclusion and its first premise come from difference discourses.
Suppose truth is something epistemic, such as superwarrant, in discourse
about what is funny, and it is something realist, such as correspondence,
in discourse about the color of snow. Then the argument appears valid,
but it also appears not to transmit truth from its premises to its conclu-
sion, given pluralism. Truth for (1) is superwarrant, but truth for (3) is
not. So, alethic pluralists have to find a way to explain the validity of such
inferences.
An especially promising approach is functionalism about truth (Lynch
2001, 2008, 2009). In Lynch’s version, a number of ‘truisms’ serve to
define the role or job description of truth. They include, among others:
Objectivity: The belief that p is true if, and only if, with respect to
the belief that p, things are as they are believed to be.
and
End of Inquiry: Other things being equal, true beliefs are a worthy goal
of inquiry.
A Plea for Immodesty: Alethic Pluralism, Logical Pluralism… 391
GTT: An argument is valid iff every casex in which its premises are true
is a casex in which its conclusion is true.4
Logical pluralism is just the view that there is more than one genuine
relation of logical consequence. Domain-specific logical pluralism (DLP)
holds that different domains of inquiry or discourses have different log-
ics. The logical consequence relation in ethics, for example, might not be
the same as the logical consequence relation in physics. It might be intu-
itionistic entailment in the former case, classical entailment in the latter.
Such a view could be attractive to alethic pluralists on purely intuitive
grounds: if logic tells us about the preservation of truth in inference, and
different discourses have different truth-properties, why should we expect
different discourses to have the same logic?
Lynch describes a more direct path from AP to DLP, provided we
assume that truth is sometimes non-epistemic and sometimes superwar-
rant (2008, pp. 134–5; 2009, pp. 94–6; see also Pedersen 2014). In
domains where truth is non-epistemic, we might expect the classical logi-
cal laws to hold, and so logical consequence in those domains would be
classical. For example, the classical Law of Excluded Middle (LEM) holds
for those discourses. But in domains where superwarrant is truth, logical
consequence amounts to the following (given that ‘cases’ amount to pos-
sible worlds):
4 Modesty
One might be tempted to try for DLP the same functionalist move Lynch
uses to resolve AP’s problem of mixed inferences. In that case, one would
consider ‘logical consequence’ to be a higher-order relation. One proposi-
tion is a logical consequence of another just in case they bear a relation to
one another that plays the role of (or manifests) logical consequence for
them. Such a move is unhelpful in this case. We need to know what rela-
tion plays the role of logical consequence for Mix. It is hard to see how an
answer to that question could be given while respecting the formality
requirement.
There are two, related worries here. First, the functionalizing move
does not shed much light on the question of whether (3) is, in fact, a logi-
cal consequence of (1) and (2). On the functionalizing move, it is a con-
sequence if it bears the relation to them that plays the role of logical
consequence appropriate for the content of the argument. But that is
merely to avoid the real question: What logical consequence relation is
appropriate for the content of that inference?
The second worry concerns the formality of the higher-order relation
of consequence. Classical and intuitionistic consequence are formal; what
P and Q are about is irrelevant to whether P is a classical consequence of
Q and to whether it is an intuitionistic consequence of Q. However,
whether P and Q are related by a relation that plays the role of logical
consequence for them does depend on their content. If P has a form that
would make it a classical consequence of Q, but not an intuitionistic
consequence of Q, then the question whether P bears a relation to Q that
plays the role of logical consequence for them is one whose answer
depends on what P and Q are about. Two inferences could be identical in
terms of their logical forms, while one of them has a conclusion that bears
the functionalized consequence relation to its premises and the other
A Plea for Immodesty: Alethic Pluralism, Logical Pluralism… 395
does not. So, the functionalized relation fails the formality condition on
being a relation of logical consequence.
Rather than functionalizing logical consequence, Lynch’s own pro-
posal is to be as modest as possible: “[W]hen reasoning across domains,
logical caution is in order: We want to limit the number of logical truths
that we endorse, so as to respect those domains which, by virtue of the
property that plays the truth-role within them, enforce less [sic] logical
laws than others” (2008, p. 138; 2009, p. 101). He codifies that idea in
the following way:
Inessential Mix
( 4) 1 + 1 = 2 and it’s not the case that Jack did not go up the hill.
(5) Therefore, Jack went up the hill.
Essential Mix
( 6) Either it’s not the case that Jack did not go up the hill or 1 + 1 ≠ 2.
(7) 1 + 1 = 2.
(8) Therefore, Jack went up the hill.
A Plea for Immodesty: Alethic Pluralism, Logical Pluralism… 397
7 Chains of Inference
A striking feature of Inessential Mix and Essential Mix is that there are
chains of inferences from the premises to the conclusion that are valid,
even by the lights of MODEST*. The premises of both arguments intu-
itionistically entail:
(9) It’s not the case that Jack did not go up the hill.
400 C. B. Wrenn
And that claim, from a domain whose governing logic is classical, clas-
sically entails the conclusion of both arguments. An advocate of AP +
DLP might hope to find a solution to the problem of mixed inferences in
that fact. The result would be a modification of MODEST* in the fol-
lowing way:
Pure Bald
( 13) If a person with 0 hairs is bald, then a person with 1 hair is bald.
(14) A person with 1 hair is bald.
(15) If a person with 1 hair is bald, then a person with 2 hairs is bald.
(16) A person with 2 hairs is bald.
and so on to:
Mixed Bald
In the non-transitive default governing logic, (18) and (19) can entail10:
(21) (P(1) & M) & (P(1) & M→P(2) & M) & … & (P(99,999)& M →
P(100,000) & M)
The inference from (21) to (22) is valid in its default governing logic:
(22) (P(2) & M) & (P(2) & M → P(3) & M) & … & (P(99,999)& M
→ P(100,000) & M)
Disjunctive Mix
( 23) Either it’s not the case that Jack didn’t go up the hill or 1 + 1 = 2.
(24) Therefore, either Jack went up the hill or 1 + 1 = 2.
The inference from (23) to (24) is not intuitionistically valid, and (23)
doesn’t have any intuitionistic consequences from which (24) can be val-
idly inferred. Both MODEST* and MODEST**, then, count Disjunctive
Mix as invalid. But that seems wrong; the argument is valid.
A Plea for Immodesty: Alethic Pluralism, Logical Pluralism… 403
8 Conclusion
The problems discussed here arise for DLP if we try to solve the problem
of mixed inferences by way of MODEST* or a modification of it. The
advocate of AP + DLP might then try to find a different solution to the
problem. Unfortunately, such a solution looks hard to come by.
The trouble is that mixed inferences can have unmixed conclusions. So,
we want to guarantee that no sound mixed inference has a conclusion
that is untrue in its home discourse. Short of MODEST* and its relatives,
there may be no way to guarantee that. The problem can be seen by con-
sidering an argument such as this (Lynch 2008, pp. 136–7):
NIX: If it is not the case that Sophie’s choice is morally right, then grass is
not green. But grass is green; so Sophie’s choice is morally right.
only invites, the retreat from DLP to VLP, and maybe that is an invitation
easily declined. Nevertheless, it does appear to be a serious problem for
AP + DLP, and there is one obvious way to avoid it: just don’t combine
AP and DLP in the first place.
Notes
1. “Science, thanks to its links with observation, retains some title to a cor-
respondence theory of truth; but a coherence theory is evidently the lot
of ethics” (Quine 1981, p. 63).
2. Lynch (2013) downplays the role of discourses in pluralist theories, but
most pluralists (including earlier time-slices of Lynch) suppose that, at
least for atomic propositions, what truth-property is relevant to a propo-
sition is a function of what discourse the proposition belongs to.
3. Different versions of functionalism characterize the relationship between
truth itself and the various other properties that “realize,” “manifest,” or
“play the role of ” truth differently. For discussion see Wright (2013).
4. GTT gives a generic characterization of logical consequence because an
argument is valid if and only if its conclusion is a logical consequence of
its premises.
5. Lynch also considers a solution on which the governing logic is the
weakest logic among those governing the various domains involved in
the compound proposition or inference. That solution, however, is just a
special case of MODEST*, where the logics in question are ordered in
such a way that one of them is weakest.
6. Some paraconsistent logics reject Disjunctive Syllogism. Maybe some
jokes both are and aren’t funny. Then such a logic might well be the right
one for discourse about comedy.
7. There is room for some debate as to what does or does not constitute a
“logical reason” for a logical law to hold. Here I assume that law holds
for a logical reason when its instances are logical truths. On Beall’s
approach, the instances of ~(p& ~p) are logical truths, so long as they
don’t involve the transparent truth device.
8. I owe thanks to Jack Lyons, Stewart Shapiro, and Sarah Wright for point-
ing out the need to consider versions of this move on behalf of AP +
DLP.
A Plea for Immodesty: Alethic Pluralism, Logical Pluralism… 405
References
Beall, J.C. 2009. Spandrels of Truth. New York: Oxford University Press.
Beall, J.C., and G. Restall. 2006. Logical Pluralism. Oxford: Clarendon Press.
Cobreros, P., P. Egré, D. Ripley, and R. van Rooij. 2012. Tolerant, Classical,
Strict. Journal of Philosophical Logic 41 (2): 347–385.
Lynch, M.P. 2001. A Functionalist Theory of Truth. In The Nature of Truth, ed.
M.P. Lynch, 723–749. Cambridge, MA: MIT Press.
———. 2008. Alethic Pluralism, Logical Consequence and the Universality of
Reason. Midwest Studies In Philosophy 32 (1): 122–140.
———. 2009. Truth as One and Many. Oxford/New York: Oxford University
Press/Clarendon Press.
———. 2013. Three Questions for Truth Pluralism. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 21–41.
New York: Oxford University Press.
Pedersen, Nikolaj J.L.L. 2014. Pluralism x 3: Truth, Logic, Metaphysics.
Erkenntnis 79: 259–277.
Quine, W.V. 1981. On the Nature of Moral Values. In Theories and Things.
Cambridge, MA: Harvard University Press.
Tappolet, C. 1997. Mixed Inferences: A Problem for Pluralism About Truth
Predicates. Analysis 57 (3): 209–210.
Tennant, N. 1987. Natural Deduction and Sequent Calculus for Intuitionistic
Relevant Logic. Journal of Symbolic Logic 52: 665–680.
———. 2012a. Changes of Mind: An Essay on Rational Belief Revision. Oxford:
Oxford University Press.
———. 2012b. Cut for Core Logic. The Review of Symbolic Logic 5 (3):
450–479.
Wright, C. 1992. Truth and Objectivity. Cambridge, MA: Harvard University
Press.
406 C. B. Wrenn
1 Introduction
Following Wright (1992), one might maintain that the debate between
realists and antirealists about a given domain of subject matter—whether
the domain is about scientific, mathematical, or ethical subject matter—
is best formulated as a debate about the nature of the truth property
associated with the domain. This motivates the standard understanding,
due to Wright (1992), of alethic pluralism, according to which there are
distinct, domain-specific truth properties, where domains are associated
with subject matter.1 To illustrate, the correspondence truth property is
associated with the scientific domain, the coherence truth property is
associated with the mathematical domain, and the superassertible truth
property (which takes truth to be epistemically constrained) is associated
with the ethical domain.
Before continuing, let me clarify two simplifying assumptions. First,
pluralists reject truth as it is represented in the familiar account of logic
in favor of an alternative representation that captures its distinctness and
A. D. Yu (*)
Faculty of Law, University of Toronto, Toronto, ON, Canada
domain specificity. I set aside the suggestion that pluralists can unquali-
fiedly grant truth as it is represented in the familiar account, as just T
(“truth”). Not least, pluralists themselves concede that they should say
something more. Crucially, the suggestion does not do justice to plural-
ists’ commitment that there are distinct domain-specific truth properties.
For the commitment suggests that something is missing from the familiar
representation of truth, so that the familiar account stands in need of
revision. Given the intertwining of truth with logic, alethic commitments
motivate corresponding logical commitments. Second, for simplicity,
truth bearers are sentences. That truth bearers are sentences is compatible
with propositions and other entities also being truth bearers.
My focus in this chapter is on how pluralists of various stripes might
go about doing logic. As I explain in what follows, alethic pluralists face
challenges to provide accounts of atomic sentences, quantified sentences,
complex sentences, and logical consequence that capture their view of
truth. But the challenges might not be limited to these. For alethic plural-
ists might also be logical pluralists, in taking distinctions between truth
properties to induce corresponding distinctions between logical conse-
quence relations. Similarly, alethic pluralists might also be ontological
pluralists, in taking distinctions between truth properties to induce cor-
responding distinctions between quantifiers. Alethic pluralists who are
also logical pluralists, or ontological pluralists, should provide a unified
account of logical consequence, or of quantified sentences, that captures
their view of truth and logical consequence, or of truth and quantifiers.
My plan for this chapter is as follows. In Sect. 2, I review the account
I propose on behalf of alethic pluralists.2 In Sect. 3, I extend the account
to accommodate logical pluralism as well as ontological pluralism.
2 Alethic Pluralism
In this section, I outline logical challenges for alethic pluralists and review
my answers to them on behalf of alethic pluralists. In the first subsection,
I outline four challenges, involving atomic sentences, quantified sen-
tences, complex sentences, and logical consequence. In the second sub-
section, I review my answers to them.
Logic for Alethic, Logical, and Ontological Pluralists 409
Alethic pluralists face four challenges, which arise from considering mix-
tures of sentences about distinct subject matters.
The first challenge, to maintain a unified account of atomic sentences,
arises from considering atomic sentences, which can be about distinct
subject matters.3
Pluralism is often illustrated using straightforward sentences. Consider
the following sentences:
These sentences are about exactly one subject matter, and so—given that
they are true—are true in exactly one way. Given that (1) is a scientific
sentence, it is correspondence true; given that (2) is a mathematical sen-
tence, it is coherence true; and given that (3) is an ethical sentence, it is
superassertible true. But what about when matters are less straightfor-
ward? Not all sentences are about exactly one subject matter. Atomic
sentences can be mixed, in that the sentences are about distinct subject
matters; these contrast with sentences that are pure, which are not about
distinct subject matters. But if mixed sentences are true, it is unclear in
what way or ways they are true.
Consider the following mixed atomic sentence:
Given that the scientific sentence (1) is correspondence true and the
mathematical sentence (2) is coherence true, their conjunction (6) is also
Logic for Alethic, Logical, and Ontological Pluralists 411
true. But in what way or ways is (6) true? The options available here are
similar to the options available when considering mixed atomic sen-
tences. Perhaps (6) is both correspondence true and coherence true.
Alternatively, perhaps it is just correspondence true. Finally, perhaps it is
true in neither of these ways, but if so, then in what way or ways is it true?
Here, one version of this last option is salient. Given that a conjunction
is true iff both conjuncts are true, perhaps there is a truth property com-
mon to (1), (2), and (6). But such a property seems to be a generic truth
property. Again, it is unclear to what extent pluralists should grant such
a property significance.
The challenge is for pluralists to maintain an account of mixed com-
plex sentences, and more generally, a standard account of complex sen-
tences, whether pure or mixed. We can put the challenge thus:
( 1) Water is H2O.
(2) 1 + 1 = 2.
(6) Water is H2O and 1 + 1 = 2.
Clearly, the argument from (1) and (2) to (6) is valid: (6) is a logical con-
sequence of (1) and (2). On the standard account of logical consequence,
logical consequence necessarily preserves truth. So perhaps some single
truth property is preserved from (1) and (2) to (6). But what property is
that? Perhaps the truth property is common to (1), (2), and (6). If so,
then the property seems to be a generic truth property. Once again, it is
unclear to what extent pluralists should grant a generic truth property
Logic for Alethic, Logical, and Ontological Pluralists 413
(LC) Γ ⊨ ϕ iff for every model M, if |γi|M = T for every γi in Γ, then
|ϕ|M = T.
exactly one subject matter, while impure domains are associated with two
or more subject matters. Where domains are either pure or impure, pure
domains generate all domains. Each atomic sentence is assigned to exactly
one domain. Negations are always assigned to the same domain as the
negand, while conjunctions and disjunctions may or may not be assigned
to the same domain as each operand, depending on whether or not the
operands are assigned to the same domain. Each atomic sentence is then
assigned a domain-specific truth value, where the relevant domain is the
one it is assigned. The domain-specific truth values of negations, con-
junctions, and disjunctions are determined by the domain-specific truth
values of the operands. Logical consequence necessarily preserves domain-
specific truth.
To incorporate talk of domains, we assume a structure 𝕯 = ⟨𝒟, ⊙⟩,
where D is a nonempty set and ⊙ is a binary operation on 𝒟. Intuitively,
𝒟 is the set of domains, and ⊙ is a mixing operation on domains. To
capture the idea that there are distinct domains (and associated domain-
specific truth properties, as well as domain-specific falsity properties), we
require that there are at least two members in 𝒟: |𝒟| ≥ 2. Let Di denote
a member of 𝒟. Intuitively, Di ⊙ Dj is the mixed domain for Di and Dj.
Since we want the mixed domain for identical domains to be the same
domain and do not want the order in which we mix domains to matter,
we require that 𝕯 is a semilattice:
(PO) Di ≤ Dj iff Di ⊙ Dj = Dj.
Note that ≤ in turn induces a strict partial order <, where Di < Dj iff
Di ≤ Dj and Di ≠ Dj. The following desirable feature follows from the fact
that 𝕯 is a semilattice:
This feature captures the idea that distinct domains where neither
domain is induced by the other induce new domains.
To incorporate talk of subject matter, we assume that there is a distin-
guished subset Pure(𝒟) of pure domains from which all domains are
generated. Accordingly, 𝒟 is the smallest superset of Pure(𝒟) closed
under the mixing operation ⊙. Intuitively, pure domains are associated
with exactly one subject matter, while impure domains are associated
with two or more subject matters. Pure domains are incomparable, in
that for distinct pure domains Di and Dj, we have both Di ≮ Dj and Dj ≮
Di. To determine which pure domains, or subject matters, a given domain
is associated with, we define the subject matter extraction function m: 𝒟
→ ℘ (Pure(𝒟)) from domains to sets of pure domains by m(Di) = {Dj ∈
Pure(𝒟 ) | Dj ≤ Di}.
To incorporate talk of domain-specific truth and domain-specific fal-
sity, we associate with 𝕯 two isomorphic structures, 𝕿 = ⟨𝒯, ⊕⟩ and 𝕱
= ⟨ℱ, ⊖⟩. The semantic values, and more specifically domain-specific
truth values, are drawn from 𝒯 ∪ ℱ. Intuitively, 𝒯 is the set of domain-
specific truth properties, and ℱ is the set of domain-specific falsity prop-
erties. Let TDi as well as Ti—both notations are useful—denote the
domain-specific truth value that represents the truth property associated
with domain Di. Similarly, let FDi as well as Fi denote the domain-spe-
cific truth value that represents the falsity property associated with
domain Di. Since 𝕿 and 𝕱 are isomorphic to 𝕯, the partial orders on
domain-specific truth values mirror the partial order on domains:
(OR) Di < Dj iff Ti < Tj iff Fi < Fj.
(iv) d(ϕ ∨ ψ) = d(ϕ) ⊙ d(ψ).
not associated with the mathematical domain). And that the latter sen-
tence is about mathematical subject matter suggests that it is associated
with the mathematical domain (but it is not associated with the scientific
domain). Similarly, the relevant objects are not enough to determine the
domain. While “2 is a prime number” and “The number 2 is beautiful”
are about exactly the same objects (i.e., the number 2), they are associated
with distinct domains. That the former sentence involves a mathematical
predicate suggests that it is associated with the mathematical domain (but
it is not associated with the aesthetic domain). And that the latter sen-
tence involves an aesthetic predicate suggests that it is associated with the
aesthetic domain.
To capture the above, a domain assignment function now needs to
assign formulas together with variable assignment functions to domains,
where a variable assignment function a based on a model M assigns vari-
ables to members of U. Since only the terms’ denotations (rather than the
terms themselves) matter for determining the domain, we go through
denotation functions and auxiliary domain assignment functions. A
denotation function [·]a based on variable assignment function a is
defined by the following:
(LC*′) Γ ⊨ ϕ iff for every model M, if |γi|M, a = Td(γ, a) for every γi in
Γ, then |ϕ|M, a = Td(ϕ, a).
(OR*) Di < Dj iff Ii < Ij.
A model is a pair M = ⟨d, v⟩, much as before, except that now the valua-
tion function v: Atomic(ℒ) → 𝕿 ∪ 𝕴 ∪ 𝕱 is from atomic sentences to
domain-specific truth values in 𝕿 ∪ 𝕴 ∪ 𝕱. We require that v(α) ∈
{Td(α), Id(α), Fd(α)}.18 Given a model M, the semantic value function | · |M:
ℒ → 𝕿 ∪ 𝕴 ∪ 𝕱 is much as before, except that now it is a function from
atomic sentences to domain-specific truth values in 𝕿∪𝕴 ∪𝕱. To cap-
ture the idea that a sentence is either domain-specifically true, domain-
specifically indeterminate, or domain-specifically false, we first require
that |ϕ|M ∈ {Td(α), Id(α), Fd(α)}. Then, we define | · |M by |α|M = v(α) and the
following:
ìTd Øf iff f = Fd f ;
ïï ( ) M ( )
(ii) f Ùy M
= í Fd (f Ùy ) iff f M = Fd (f ) or y M
= Fd (y ) ;
ï
ïî I d (f Ùy ) otherwise.
ìTd f Ú y iff f = Td f or y = Td y ;
ïï ( ) M ( ) M ( )
(iii) f Ú y M = í Fd (f Ú y ) iff f M = Fd (f ) and y M = Fd (y ) ;
ï
ïî I d (f Ú y ) otherwise.
Next, we define domain-specific logical consequence as follows:
(LC**) G Di f iff for all M such that d (g j ) = d (f )
= Di for all g j in G , if g j = TDi
M
(LO***) (i) |∀Di υϕ|M, a = Td(∀υϕ, a) iff |ϕ|M, b = Td(ϕ), b for all b where b ∼υ a
and b(υ) ∈ U Di ;
(ii) |∃Di υϕ|M, a = Td(∃υϕ, a) iff |ϕ|M, b = Td(ϕ), b for some b where b ∼υ a and
b(υ) ∈ U Di .
Notes
1. Sometimes, it is added that a truth property is whatever satisfies certain
truth-characterizing principles. However, this possibly extra feature does
not concern me here, so I omit mention of it for simplicity.
2. See Yu (2017).
3. This challenge is discussed in Lynch (2005b), Sher (2005, 2013), Wyatt
(2013), Pedersen and Wright (2013b).
4. This challenge is discussed in Lynch (2009), Shapiro (2009, 2011),
Cotnoir (2013), Pedersen and Wright (2013b).
5. See Tappolet (2000), Edwards (2008, 2009, 2012), Cotnoir (2009,
2013), Pedersen (2006, 2010, 2012), Pedersen and Wright (2013a),
Wright (2012) on the extent to which pluralists should grant a generic
truth property theoretical significance.
6. This challenge is discussed in Williamson (1994), Tappolet (2000),
Lynch (2001, 2004, 2005b, 2008, 2009, 2013), Edwards (2008, 2009),
Cotnoir (2009), Pedersen and Wright (2013b), Wright (2013).
7. Where there is little risk of confusion, I am somewhat loose with the
distinction between truth properties themselves and truth values that
formally represent them. For example, I may write of T as either being a
Logic for Alethic, Logical, and Ontological Pluralists 425
References
Beall, J. 2000. On Mixed Inferences and Pluralism About Truth Predicates.
Philosophical Quarterly 50 (200): 380–382.
Beall, J., and G. Restall. 2006. Logical Pluralism. Oxford: Oxford University Press.
426 A. D. Yu
Cotnoir, A.J. 2009. Generic Truth and Mixed Conjunctions: Some Alternatives.
Analysis 69 (3): 473–479.
———. 2013. Validity for Strong Pluralists. Philosophy and Phenomenological
Research 86 (3): 563–579.
Cotnoir, A.J., and D. Edwards. 2015. From Truth Pluralism to Ontological
Pluralism and Back. Journal of Philosophy 112 (3): 113–140.
Edwards, D. 2008. How to Solve the Problem of Mixed Conjunctions. Analysis
68 (2): 143–149.
———. 2009. Truth-Conditions and the Nature of Truth: Resolving Mixed
Conjunctions. Analysis 69 (4): 684–688.
———. 2012. On Alethic Disjunctivism. Dialectica 66 (1): 200–214.
Field, H.H. 1980. Science Without Numbers: A Defense of Nominalism. Princeton:
Princeton University Press.
Lynch, M.P. 2000. Alethic Pluralism and the Functionalist Theory of Truth.
Acta Analytica 15: 195–214.
———. 2001. A Functionalist Theory of Truth. In The Nature of Truth: Classical
and Contemporary Perspectives, ed. M.P. Lynch, 723–750. Cambridge, MA:
MIT Press.
———. 2004. Truth and Multiple Realizability. Australasian Journal of
Philosophy 82 (3): 384–408.
———. 2005a. Functionalism and Our Folk Theory of Truth: Reply to Cory
Wright. Synthese 145: 29–43.
———. 2005b. Précis to True to Life, and Response to Commentators.
Philosophical Books 46 (289–291): 331–342.
———. 2008. Alethic Pluralism, Logical Consequence and the Universality of
Reason. Midwest Studies in Philosophy 32: 122–140.
———. 2009. Truth as One and Many. Oxford: Oxford University Press.
———. 2013. Three Questions for Truth Pluralism. In Truth and Pluralism:
Current Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright, 21–41.
Oxford: Oxford University Press.
McDaniel, K. 2009. Ways of Being. In Metametaphysics: New Essays on the
Foundation of Ontology, ed. D. Chalmers, D. Manley, and R. Wasserman,
290–319. Oxford: Oxford University Press.
———. 2010a. Being and Almost Nothingness. Noûs 44 (4): 628–649.
———. 2010b. A Return to the Analogy of Being. Philosophy and
Phenomenological Research 81 (3): 688–717.
———. 2017. The Fragmentation of Being. Oxford: Oxford University Press.
Pedersen, Nikolaj J.L.L. 2006. What Can the Problem of Mixed Inferences
Teach Us About Alethic Pluralism? The Monist 89 (1): 102–117.
Logic for Alethic, Logical, and Ontological Pluralists 427
1 Introduction
Logicians and Philosophers have developed a large and rich array of logi-
cal systems—intuitionistic logic, relevance logics, free logics, a vast assort-
ment of many-valued logics, multiple conclusion logics, extensions of
classical logic of various types including temporal and modal logics and
many others besides. These are naturally referred to as “logics” and can be
considered and studied in the abstract, giving clear-cut notions of what
follows from what according to the various different logics. Talk of “logic”
(in the singular) often signals something more than a system considered
in the abstract. Logic as the study of what follows from what, looks to
concern more than what follows from what relative to some chosen system.
Similarly for talk of, for example, the study of the logic of our language
or of the world. Not all of the abstract systems are on a par and philoso-
phers have often provided detailed arguments for regarding their chosen
logic as of particular or unique significance, where many of the positions
R. Keefe (*)
Department of Philosophy, University of Sheffield, Sheffield, UK
e-mail: r.keefe@sheffield.ac.uk
and, for example, deny that “p or not-p” must always be true, since
(roughly) p and not-p may both be unknowable. Within certain scientific
domains, by contrast, there may be no such constraints and there may
instead be reasons to presuppose bivalence.
Lynch coins the phrase “domain-specific logical pluralism” (DLP) for
the position at issue here. And Pedersen writes, “Logical pluralists main-
tain that the distinction between different kinds of domains is not merely
verbal, but that it bears on inference. What inferences come out valid
may vary from domain to domain”.6 This may be a misleading portrayal
of most logical pluralists, for whom this does not capture the role of
domains (see below); but it is nonetheless worth considering this domain-
specific logical pluralism and some preliminary objections it faces.
One important general question is the following. If different logics
govern different domains, what logic governs arguments that involve
statements from more than one different domain? This question can be
usefully addressed alongside consideration of the “mixed inference chal-
lenge” to pluralism about truth.7 That challenge takes an argument with
premises from different domains, thus subject to different truth predi-
cates, and considers what validity can amount to if it isn’t necessary pres-
ervation of a single property of truth. Tappolet considers, for example,
“Wet cats are funny; this cat is wet; therefore, this cat is funny”, where the
first premise and conclusion are from a different domain with different
standards of truth than the second premise.
One kind of response to the mixed inference challenge maintains that
what matters for validity is necessary preservation of falling under the
concept of truth, even if what it takes to fall under that concept is differ-
ent for different types of sentence.8 Tappolet complains, “why should we
need the many truth predicates instead of the one that does the inferen-
tial job?” (2000, p. 384): that one thing is necessarily preserved in valid
arguments may seem to provide enough unity to undermine truth plural-
ism. But advocates have shown how truth pluralism nonetheless remains
viable on most plausible forms of the view.
The logic governing a mixed inference, on this approach, would be the
intersection of the logics governing each of the domains involved in the
argument, for arguments valid in only some domains will fail to necessar-
ily preserve some kind of truth in some other domain. This will often be
Pluralisms: Logic, Truth and Domain-Specificity 433
the weakest of the logics for each of the two (or more) domains; for
example, for an argument with realist and anti-realist domains, the logic
would be intuitionistic.9 If the picture is to be as general as possible and
permit a wide range of different logics as applicable to different domains,
including paraconsistent logics, it will not always be the case that the
logic of one domain is contained within the other. If we had an argument
spanning a range of domain representing the full diversity of consequence
relations, the intersection may be nearly empty (compare Beall and
Restall 2006, p. 92.)10
It may be natural, with this approach, to assign particular significance
to that logical consequence relation which is the intersection of the logics
across all domains. Perhaps we should regard that as the logical conse-
quence relation. That would be to abandon domain-specific logical plu-
ralism. It would then be inaccurate to present the key consequence
relation as the intersection across the specific consequence relations, since
those latter will not after all be logical consequence relations. The stron-
ger, apparently justified principles of reasoning within particular domains
would then have to count as non-logical. Lynch says, “… in domains
which, according to this suggestion, nonetheless appear classical … biva-
lence must be true for some non-logical reason. And one might wonder
what that reason might be.” Suppose, then, we resist this option and
uphold domain-specific logical pluralism and the different logics for dif-
ferent domains; there will potentially be other different logics for differ-
ent mixed inferences, depending on what overlapping logics are at issue
and mixed inferences in which it is the very weakest (intersection) logic
at play will be rare.
In the next section, keeping domain-specific logical pluralism on the
table, we consider a different approach.
conclusions are from within a single domain rather than being a mixed
inference. Cotnoir appears to deny this, however, and says, “One can, if
one wishes, insist that propositions in some domains are always classically
evaluated while allowing for non-classical domains of discourse. One
merely stipulates that, at some coordinates in the n-tuples, components
are always only selected from 1 and 0” (p. 14). But that doesn’t make for
bivalence within the domain in the desired sense that all sentences from
that domain are true or false; rather all sentences at all are true or false in
that domain. As he earlier asks, “why think that the negation of a true
descriptive proposition must be morally true, mathematically true etc.?”
It should be compatible with taking mathematical truth to be bivalent to
deny that either “Jack is funny” or “Jack is not funny” is mathematically
true. This seems to make a mockery of the idea of mathematical truth—
limited to the mathematical domain—within a pluralism about truth.
Cotnoir also claims to provide a framework that can accommodate
domains in which truth is constructive and intuitionistic logic holds,14
but, I will argue, the same problem threatens. The original Boolean alge-
bra delivers classical logic and so leaves no room for a constructive notion
of truth, and thus fails to capture the notion of truth that Wright—whose
position he seeks to stay close to—defends for a domain such as humor
and treats as key to his central notion of superassertibility. Cotnoir’s
response to this kind of problem is to develop a variant on the framework
described above which validates intuitionistic logic. Whereas the original
classical version was based on a Boolean algebra, an intuitionistic version
is based on a Heyting algebra. The resulting consequence relation, ǀ=I, is
intuitionistic.
The mechanics of the system and the construction of semantic values
is somewhat different in this case from the previous versions. In the clas-
sical version, the n-tuple consists of 1s and 0s, where the third possibility
of ½ is added in the many-valued variant. The talk of the Boolean algebra
is then introduced at the level of the behavior of the resulting many-place
values. This is how Cotnoir gets quickly to the conclusion that classical
logic (captured by that Boolean algebra) holds for mixed inference, that
is, allowing for the full range of n-tuples that could be assigned. In the
intuitionistic alternative, we are told that the ith place of the n-tuple are
members of the ith Heyting algebra, where for some domains, this could
Pluralisms: Logic, Truth and Domain-Specificity 437
logic of the argument. So, we should not assume that the subject-matter
of an argument can always determine the right logic, as phenomena such
as vagueness can also impact on the argument while cutting across sub-
ject-matters. And the logical relevance of a phenomenon such as vague-
ness is even harder to isolate (in the absence of a specific subject-matter),
since the possible vagueness of a nonborderline sentence is still relevant
to the assessment of validity.
Even if we found a way to isolate sentences within particular domains,
we would again face the problem that typical arguments involving those
sentences also involve sentences from outside those domains. Might there
then be another logic governing cross-domain argument? We have seen
above that treatments of “mixed inferences” can offer a logic that differs
from the logics of other domains.19 Should we then call this a new
domain—the domain of mixed inferences—allowing that sentences are
in this domain as well as in their more exclusive domain? After all, the
idea of DLP was that different logics could be upheld because they were
limited to different domains, so a mixed inference logic suggests a mixed
inference domain. If so, that would be an extreme example of a domain
with no unity of subject-matter. If that domain contains all sentences—
since all sentences could be part of some mixed inference in combination
with sentences from any other domain—then it looks to yield a universal
logic. Other logics will then just look like the kind of enhanced conse-
quence relations described above.20 In short, we need a logic of mixed
inferences since arguments needn’t remain within a single pure domain.
But such a logic threatens to undermine the pluralism of the other so-
called consequence relations since it has a universality.
We can frame this as we did earlier (Sect. 3). Take an argument, A, B
so C, whose premises and conclusion fall within domain m. We then
assess it as validm since it is validated by the enhanced rules specific to that
domain. But the argument is also in the broader domain, SP, and that
same argument is not validSP, since not validated by the weaker rules.21 If
we regard both as genuine logical consequence relations, we have a logical
pluralism beyond domain-specific logical pluralism, since the very same
argument is both valid and not valid. The earlier discussion suggests the
natural move of identifying genuine validity with the weaker consequence
relation. Lynch, however, suggests that we should regard A, B so C as
442 R. Keefe
are the premises all true and the conclusion false. Within DLP, there must
be a restriction to interpretations within the domain; but if the domain is
not clearly demarcated, the test cannot be adequately applied. Suppose,
for example, that we hope to put aside vagueness and focus on the “pre-
cise” domain. Any sentence could be vague, at least as far as its structure
is concerned. So, even if the premises are precise, the consideration of all
possible interpretations will include some vague ones. We thus would
need to limit the interpretations quantified over for the test to the precise
ones. But then we can only reach a conclusion about what follows from
what on the assumption of necessary precision of the relevant elements of
the argument. Since this is clearly a false assumption, the suggestion that
this captures the genuine consequence relation for that domain looks
questionable. When the domains are demarcated by subject-matter (e.g.
morality or humor), then the limitation to interpretations within the
domain may be less problematic than the previous case. But this will still
only do if we can isolate the sentences in that domain in such a way as to
encompass sentences involved in typical arguments concerning that sub-
ject-matter; as indicated above, this is often not easy. Topic neutrality
allows for the kind of generality required.
I have argued that the claim that the right logic varies across different
domains is not sustainable, whether in a version stemming from plural-
ism about truth or not. It is still open to the logical pluralist to reject
topic neutrality as a requirement on logic,23 but I put aside that option
and turn, in the next section, to consider a model whereby different rules
of reasoning are permitted in different domains or contexts, where this
can be superimposed over a universal (and topic-neutral) logic.
endorses only the minimal logic common to all contexts (compare the
discussion above).
To settle the correct consequence relation, however, we would need to
determine what rules are common to all contexts. Someone convinced
that classical logic is the true logical consequence relation might argue, for
example, that reductio ad absurdum is always justified, even if it is dis-
puted. Conversational participants may together have some control over
what rules are added because justified in the context (e.g. the timetable
rule), but, it might be thought, not have the same control over excluding
other rules. Arguably, however, there are contexts in which classical logic
is too strong—it is said, for example, that reasoning about an inconsistent
database requires rules of a relevance logic that block explosion, so that it
is not the case that the recording of contradictory facts on the database
warrants inferring any conclusion at all. Or take another application of a
relevant logic: Lewis argues that a paraconsistent logic is a suitable “logic
for equivocators”.25 If we cannot rule out that A is ambiguous across the
premises not-A and A or B, then we should not infer B; rather the rules of
a paraconsistent logic are better placed to guide our reasoning if we cannot
be sure we are avoiding equivocation. The classicist can respond by
explaining how these cases do not show that there is any argument genu-
inely of the classical form in question (e.g. disjunctive syllogism or reduc-
tio) that fails to be valid. For example, the database argument really needs
its premises and conclusion appended with “according to the database”, so
that what look like P and not-P are actually the non-contradictory “accord-
ing to the database P” and “according to the database not-P”. Similarly,
corresponding to the potentially equivocating English argument, there are
several different propositional contents (corresponding to the various dis-
ambiguations) and none of those invalidate disjunctive syllogism.
The alternative position which does identify logical consequence with
the very weakest common core of all candidate logics may seem truer to
the spirit of the approach to relative validity as explained above, for exam-
ple, in the attempt to “take the argument as it comes”. I will not here try
to settle the choice between these options (and others)—it would require
tackling much-disputed questions about the essential nature of logical
consequence. I have instead illustrated how Smiley’s account of relative
Pluralisms: Logic, Truth and Domain-Specificity 447
relevant. The focus on contexts rather than domains is also more suitable
for dealing with phenomena such as vagueness and the thought that in
many contexts it is appropriate to ignore vagueness and reason with rules
that are legitimate in the absence of vagueness.
We can then allow that the same (weak) underlying logic governs all
domains, strictly speaking, while granting that additional rules are war-
ranted within a specific domain. Even if we maintain that these rules do
not count as logical, since not reflective of absolute validity, they share
many features of logical rules and play a comparable role. They appear a
long way down the continuum from the ad hoc timetable rule “it’s
Tuesday so this is Paris”. In reasoning and in evaluating arguments it is
often relative validity that is paramount and that is most appropriate for
formalising arguments.
Regarding pluralism about truth specifically: the role of the minimal
core logic here mirrors the approach to the mixed inference challenge
that declares an argument valid if it preserves falling under the concept of
truth (allowing for different properties of truth for different domains).
The truth pluralist would thus then need a response to the standard
objection that truth should then be identified with that property that is
necessarily preserved in validity, but if such a response can be given the
framework of relative validity could offer an additional way to approach
the variety of arguments involving the variety of types of truth.
The above context-based framework can thus be employed by the
domain-specific logical pluralist, but that framework also allows for logi-
cal pluralism that does not involve several domains. Different contexts
may demand rules of classical logic, where others only justify intuitionis-
tic rules, even when the same domain (e.g. mathematics) is at issue.26
According to Lynch, the claim that “there is more than one logic gov-
erning our reasoning” is the key claim of logical pluralism (2008, p. 132).
The framework of relative validity allows us to accommodate the idea
here without commitment to multiple logical consequence relations. For
principles and rules can govern our reasoning without their reflecting
logical consequence: other informal rules may be crucial to our reasoning
in certain contexts for pragmatic reasons and stories about how to reason
will involve more than just principles of logic.
Pluralisms: Logic, Truth and Domain-Specificity 449
Notes
1. Uses of logical systems can go beyond capturing logical consequence
relations. For example, it might be useful to employ a paraconsistent
logic to work with databases that may contain inconsistent data. But that
is not to say that there are true contradictions, even though it can be on
the record that p and on the record that not-p. See Keefe 2014, footnote
13.
2. See Keefe 2000, chapter 5, and on pluralism about supervaluationist
notions of consequence, see Keefe 2001. Hjortland 2012 offers the help-
ful term “intra-theoretic pluralism” for this kind of pluralism. A different
relatively modest logical pluralism is one that maintains that there is no
uniquely correct choice of logical constants and so we can construct dif-
ferent equally good logics by selecting different sets from the putative
logical constants.
3. See, for example, Lynch 2009, Wright 1992, 2013 and the volume of
papers Pedersen and Wright 2013.
4. For example, Wright maintains that there is one concept but many prop-
erties and Lynch adopts a functionalism about truth whereby different
concepts play the role.
5. Pedersen (2014, p. 262) maintains that “alethic pluralism argues that
any alethic pluralism that accepts both realist and anti-realist concep-
tions of truth “brings on a commitment to logical pluralism” and Lynch
writes: ‘The alethic functionalist … is not required to endorse …logical
pluralism. But it is likely that she will.’ (2009, p. 104).
6. Pedersen 2014, p. 260.
7. Tappolet 1997, Williamson 1994.
8. See, for example, Wright 2013, p. 133. Beall’s many-valued approach
(2000)—allowing different truth-values, all designated values, for differ-
ent predicates, and validity as preservation of designated value—argu-
ably also falls into this category of approach.
9. See Lynch 2008, p. 137.
10. Lynch 2008, p. 139 responds to the objection that the first approach he
considers—the logic of a mixed inference is the logic governing the
weakest element of it—requires “the assumption that the logics in ques-
tion can be ordered, in the sense that the stronger logics are extensions of
the weaker logics” by maintaining, without further explanation, that this
is not “an unreasonable constraint on those logics that apply only to
specific domains of inquiry”. But he also considers the above objection
450 R. Keefe
of taking the logic to be the intersection of the logics governing the ele-
ments of the inference.
11. Note that this denial of a unique property is compatible with accepting
a unified concept of truth. On some conceptions of properties (e.g. con-
cept nominalism), that would surely be enough for commitment to a
single general property of truth, in which case the distinction between
weak and strong pluralism would collapse.
12. Along the way to showing this, he shows that his set of values with con-
junction and disjunction forms a lattice with the top value taking 1 in
every place and the bottom value taking 0 in every place.
13. This illustration is an odd case of a sentence that doesn’t fall into any
domain. The Law of Excluded Middle also fails for more typical sen-
tences within some domain, as the first disjunct, and thus the second
disjunct too, will typically take value ½ in some of the places and thus
the disjunction will not take the top value (1 in all places). The only cases
where it will take the top value is when the disjuncts have 1 or 0 in each
place, which will be the rare cases of sentences simultaneously in all
domains, for example, perhaps itself a disjunction with a disjunct from
each domain.
14. “In this section, I extend the algebraic account of validity to nonclassical
domains, showing how the account can handle domains for which para-
complete, paraconsistent and intuitionistic logic seem most appropriate”
(p. 13).
15. Might it be represented instead by alternative Boolean algebra, so that
the logic is still classical, even though bivalence does not hold? If biva-
lence is a logical feature of a domain (as Lynch argues, 2008), then surely
only the two-valued Boolean algebra will be able to capture the domain.
16. Domain-specific logical pluralism is not the kind of position Beall and
Restall are interested in, for example. They would, I think, regard it as a
form of relativism, since what is valid is relative to the domain/context.
Beall and Restall’s logical pluralism turns on the different interpretations
of case in the defining principle of logical consequence (GTT): “An
argument is validx if and only if, in every casex in which the premises are
true, so is the conclusion.” Pedersen (2014) seeks to adopt their frame-
work by explaining how different notions of case could be appropriate to
different domains. But (GTT) commands a universality in its bicondi-
tional, and limiting the notion of case to a single domain does not pro-
vide a “precisification” of “case” in the manner required for Beall and
Restall’s framework.
17. Field 2009, p. 344.
Pluralisms: Logic, Truth and Domain-Specificity 451
18. Issues about domains are addressed in Lynch 2009 and Wyatt 2013,
among many other places.
19. The arguments below do not require that the logic of mixed inferences is
distinct from all of the logics of the pure domains—it could, rather, be
the weakest, for example, intuitionistic logic if the domain-specific
options are just classical logic and intuitionistic logic. The same ques-
tions arise as to the status of the stronger logics.
20. Perhaps, then, we should deny that there is any mixed domain—the
domain-specific logical pluralist could consider the logic governing
mixed inferences (the intersection of the logics of the pure domains)
without requiring a corresponding domain. Different logics may then
result from different domains either by being logics of those different
domains or resulting from the interaction of different domains in a more
complex way. Whether there is a mixed domain or not, the argument
regarding the universality of the logic of mixed domains still holds.
21. Or, if SP is not strictly a domain, then the assessment via the most gen-
eral logic still applies.
22. Beall and Restall (2006), for example, identify the settled core features of
logical consequence relations as necessity, normativity and formality: to
qualify as one of the genuine consequence relations of their logical plu-
ralism, a relation must have these features. Although details of Beall and
Restall’s “settled core” are controversial, a requirement of formality is
widely accepted and the domain-specific relations in question meet none
of the candidate more detailed specifications of the formality criterion.
23. See, for example, Shapiro 2014 for a view taking this line, focusing on a
range of fruitful mathematical theories that use different logics (e.g.
intuitionistic analysis).
24. Smiley 1995. See also Keefe 2010 for further discussion of Smiley’s
account of relative validity.
25. Lewis 1982.
26. For other recent papers connecting logical pluralism with ideas about
context, see Caret 2017 and Simard Smith 2018.
References
Beall, J. 2000. On Mixed Inferences and Pluralism About Truth Predicates. The
Philosophical Quarterly 50: 380–382.
Beall, J., and G. Restall. 2006. Logical Pluralism. Oxford: Clarendon Press.
452 R. Keefe
Caret, C. 2017. The Collapse of Logical Pluralism Has Been Greatly Exaggerated.
Erkenntnis 82: 739–760.
Cotnoir, A.J. 2013. Validity for Strong Pluralists. Philosophy and Phenomenological
Research 86: 563–579.
Field, H. 2009. Pluralism in Logic. Review of Symbolic Logic 2: 342–359.
Hjortland, O.T. 2012. Logical Pluralism, Meaning-Variance, and Verbal
Disputes. Australasian Journal of Philosophy 91: 355–373.
Keefe, R. 2000. Theories of Vagueness. Cambridge: Cambridge University Press.
———. 2001. Supervaluationism and Validity. Philosophical Topics 28: 93–105.
———. 2010. Relative Validity and Vagueness. In The Force of Argument: Essays
in Honor of Timothy Smiley, ed. J. Lear and A. Oliver, 127–143. London:
Routledge.
———. 2014. What Logical Pluralism Cannot Be. Synthese 191: 1375–1390.
Lewis, D.K. 1982. Logic for Equivocators. Noûs 16: 431–441.
Lynch, M.P. 2008. Alethic Pluralism, Logical Consequence and the Universality
of Reason. Midwest Studies in Philosophy 32: 122–140.
———. 2009. Truth as One and Many. Oxford: Oxford University Press.
Pedersen, Nikolaj J.L.L. 2014. Pluralism × 3: Truth, Logic, Metaphysics.
Erkenntnis 79: 259–277.
Pedersen, Nikolaj J.L.L., and C.D. Wright, eds. 2013. Truth and Pluralism.
Oxford: Oxford University Press.
Shapiro, S. 2014. Varieties of Logic. Oxford: Oxford University Press.
Smiley, T.J. 1995. A Tale of Two Tortoises. Mind 104: 725–736.
Smith, P. Simard. 2018. Assessment Context-Sensitive Logical Claims. Inquiry
https://doi.org/10.1080/0020174X.2017.1402698.
Tappolet, C. 1997. Mixed Inferences: A Problem for Pluralism About Truth
Predicates. Analysis 57: 209–210.
———. 2000. Truth Pluralism and Many-Valued Logics: A Reply to Beall. The
Philosophical Quarterly 50: 382–385.
Williamson, T. 1994. A Critical Study of Truth and Objectivity. International
Journal of Philosophical Studies 30: 130–144.
Wright, C.J.G. 1992. Truth and Objectivity. Cambridge, MA: Harvard
University Press.
———. 2013. A Plurality of Pluralisms. In Truth and Pluralism, ed. Nikolaj
J.L.L. Pedersen and C.D. Wright, 123–156. Oxford: Oxford University
Press.
Wyatt, J. 2013. Domains, Plural Truth and Mixed Atomic Propositions.
Philosophical Studies 166: 225–236.
Aletheic and Logical Pluralism
Kevin Scharp
There are two main objectives. One is to provide a unified framework for
thinking about pluralisms with respect to truth and pluralisms with
respect to logic—aletheic and logical pluralism. It is tempting to think
that there is a single underlying pluralism that gets applied in these two
different cases, but I think that is wrong. The aletheic pluralists are, for
the most, working with one model, and the logical pluralists are working
with a different model. There are exceptions in each case but the domi-
nant views are distinct. I want to try to bring out this difference and
think about which of these versions is more plausible.
The second objective is to use the version of pluralism that I think is
more plausible to describe a view on the truth paradoxes I have been
defending lately. And then use it again to formulate a position on the
paradoxes that would be a rival to the one that I have advocated. It seems
K. Scharp (*)
Arché Philosophical Research Centre, University of St Andrews,
St Andrews, UK
Department of Philosophy, University of St Andrews, St Andrews, UK
e-mail: ks70@st-andrews.ac.uk
like this view I will present would be a better than my own view. I defend
this rival view a bit and then argue that it is actually not a better approach
to the paradoxes.
1 Pluralisms
Roy Cook writes: “One is a pluralist about a particular phenomenon X iff
one thinks that there is more than one correct account of X” (Cook 2010:
493). Cook then defines logical pluralism as “[T]here is more than one
‘correct’, ‘best’, or ‘legitimate’ logic” (Cook 2010: 492), and Gillian
Russell offers something very similar: “[T]here is more than one correct
logic” (Russell 2016: 1).
On the other hand, aletheic pluralism has been defined by Pedersen
and Wright as “[T]here is more than one way of being true” (Pedersen
and Wright 2013: 1) and in Pedersen “There are several properties … in
virtue of which propositions are true” (Pedersen 2012: 1) and “What
property makes propositions true may vary across domains or from sub-
ject matter to subject matter” (Pedersen and Wright 2013: 2).
Given these kinds of quotations it is tempting to think that there is just
one kind of pluralism that then is applied in various ways. In a moment
it will be clear why that is a mistake.
I am not going to go through all the specific instances of logical plural-
ism or aletheic pluralism. However, under the logical pluralists, I am
thinking of Beall and Restall of course but then also those like Roy Cook
and Stewart Shapiro who think of model theories as scientific models.
And there is Gillian Russell’s version of logical pluralism and Hartry
Field’s version as well.1 Under the aletheic pluralism banner is of course
Crispin Wright’s pioneering work and then different ways of developing
it by Edwards, Lynch, Pedersen, and Ulatowski.2
those used in the conversation, that is, in the speaker’s context of utter-
ance. The domain approach is analogous to subject sensitive invariantism
in the epistemological debate, and the context approach is analogous to
contextualism in the epistemological debate. Keeping these to views sepa-
rate has been thought to be essential to progress in epistemology.
Hopefully those working on pluralism can see the domain vs context
divide as equally important.
It is striking because these two pluralist traditions have developed
somewhat independently of one another. For the most part, the truth
pluralists have been working with one model, the domain approach, and
the logical pluralists are working with a different model, the context
approach.
With that point made, we can return to the contrast between moderate
pluralism and strong pluralism as long as we keep the underlying domain/
context distinction in mind. If we adhere to the domain approach, we
can formulate a moderate logical pluralism in the following way: argu-
ment A is valid-as-such if and only if A belongs to domain 1 and A is
valid1 or A belongs to domain 2 and A is valid2 or … However, if we
adhere to the context approach, it is not obvious how to formulate a
moderate logical pluralism. One problem is that, on a context approach,
arguments are not confined to a unique context. A single argument might
be valid in one context and invalid in another.
For what it is worth, I think the context approach is preferable because
of the implausibility of disjoint domains for truth bearers. As anyone
who is familiar with the aletheic pluralism debate can attest, the literature
is filled with attempts to solve mixing problems, and that is a direct con-
sequence of assuming the domain approach.9 If one does not assume the
domain approach and instead assumes the context approach, then these
sorts of mixing problems never arise.
For another reason to prefer the context approach, consider a common
motivation for logical pluralism: paraconsistent logic is good for data-
bases that might have inconsistent information. This motivation might
go back to one of the important papers in that this tradition is Nuel
Belnap’s ‘How a Computer Should Think’, which advocates first-degree
entailments (FDE) for consequences in a database. However, if we take
the domain approach to logical pluralism, then what becomes of this
458 K. Scharp
In each case, ‘deflationary truth’ refers to a truth property that satisfies the
T-schema, where the biconditional involved in the T-sentences is the
Aletheic and Logical Pluralism 461
obvious one from the logic in question (e.g., in paracomplete logic, the
T-sentences have the paracomplete biconditional; in BX, they have the
BX biconditional; etc.).
This coordinated pluralist theory being presented entails that all ale-
theic and logical expressions of natural language designate properties.
Which property they designate depends on some parameter or variable. I
use ‘context’ as a general term for what determines this parameter. One
might be more specific about this model in various ways. For example,
one might take the context to be a context of utterance and take it to deter-
mine the contents of the word ‘true’ and the logical expressions (e.g., ‘and’,
‘or’, ‘if and only if ’, ‘valid’, ‘entails’). This would be a standard contextual-
ist view. Instead, one might take the context to be a context of assessment
and take it to determine only the extensions of the truth predicate and the
logical expressions. This would be a semantic relativist view.17 Other vari-
ants are possible as well (e.g., non-indexical contextualism, and mixing
options for distinct expressions).
5 Discussion
Objection: Context dependence is incompatible with aletheic or logical
pluralism. These views are competitors that cannot both be true.
My reply: The essence of pluralism is that something’s extension or
reference is sensitive in some way to its domain or its context or whatever.
Contextualism is one way of interpreting any pluralism that utilizes the
context approach. Non-indexical contextualism is another option too, as
is assessment sensitivity. Each of these is a kind of extension-sensitivity.
Hence, contextualists about truth are pluralists about truth (a truth prop-
erty for every content of ‘true’). Pluralists need not be contextualists,
obviously; they could go with some other context-based semantic theory,
the domain approach, or some other way to link the plurality of truth
properties to the single word, ‘true’.
Objection: The coordinated pluralist view is formulated as a kind of
strong pluralism. What about moderate pluralisms? Could they have any
role in a coordinated view?
462 K. Scharp
6 Revenge
I have presented coordinated logical and aletheic pluralism and defended
it from a couple of obvious objections. However, there is a serious prob-
lem with it. I think it ultimately fails, and it fails because it does not do
better on revenge paradox phenomena.
One of the most horrible things about working on the paradoxes
associated with truth is the revenge problem. When a theorist offers a
solution to the paradoxes, it turns out that the proposed solution itself
gives rise to new paradoxes that are structurally identical to the liar para-
dox. And these new paradoxes determine that the proposed solution is
inconsistent. The revenge phenomenon is extremely frustrating and it
affects almost everybody in this literature.21 I say ‘almost’ everybody
because I claim it does not affect my own view. Whether I am right or
not, if an approach to the aletheic paradoxes does not do better on the
revenge paradoxes, then that is a non-starter—one should not even
bother with it.
In the coordinated pluralist view just presented, there are lots of
revenge paradoxes. In particular, sentences like:
In sum, the effects of well-known revenge paradoxes from (1) and (2)
are mitigated by the coordinated pluralist view. That is exactly what it was
designed to do.
However, there are big problems. We have lots of potential contexts of
utterance and in each of them there is some non-trivial combination of
logical expressions and a truth predicate.
Or, more accurately, we have a single language with context-dependent
logical and aletheic vocabulary. In each context of utterance, these expres-
sions have particular contents, and in no context of utterance are the
contents such as to entail triviality. Or so it seems.
So far we have been ignoring the language’s ability to talk about what
goes in different contexts of utterance. Of course, we have to be able to
utilize these sorts of resources to even set up the coordinated pluralist
theory. However, once we focus on expressions like ‘context’ and ‘true in
a context’, we can see why they make trouble. For example, imagine in
Context 1, the truth predicate has the content of a classical gappy truth
predicate (i.e., it obeys T-Out, but not T-In) and the logical expressions
are classical. In Context 2, the truth predicate has the content of a defla-
tionary truth predicate and the logical expressions are paracomplete (e.g.,
Field’s logic). In a conversation with Context 1, someone might utter a
sentence:
contexts that pose the problem. To incorporate each one of the range of
approaches to the paradoxes into a single view, we have appealed to con-
texts. But in so doing, we expose the coordinated view to a new kind of
revenge paradox that is not a problem for any one of the particular
approaches incorporated. It is the incorporation tools (i.e., the frame-
work of contexts) that expose the coordinated view to revenge and, ulti-
mately, triviality.
Given the revenge susceptibility of coordinated logical and aletheic
pluralism, it is better to stick with an approach that does not give rise to
revenge paradoxes at all, even if that view has other costs related to the
truth predicate. The attempt to do better by combining the good aspects
of my view and the good aspects of other views like Beall’s fails.
One might object: If there are revenge paradoxes on the coordinated
view, why doesn’t my own approach fall prey to them? The reason is that
the two truth predicates that I advocate are specially designed to avoid
revenge paradoxes, and they have to be very carefully designed to avoid it.
For example, I have argued that ‘true’ is assessment-sensitive, and that
in situations where truth paradoxes abound, we ought to use new con-
cepts, ascending truth and descending truth, instead of the concept of
truth. Using these resources, we can construct sentences like
Here ‘U’ ranges over contexts of utterance and ‘A’ ranges over contexts of
assessment. It is easy to show that for all U and for all A, (5) and (6) are
ascending true in U from A and not descending true in U from A. There
is nothing paradoxical or contradictory about this result. Indeed, ascend-
ing truth and descending truth have been designed to work together
flawlessly in classical logic.23
When we appeal to contexts to coordinate truth and logic together, that
very careful design is lost, and the result is that we get the revenge paradox.
Appealing to contexts allows us to combine many logic/truth predicate
combinations. But this very mechanism introduces a way of generating
revenge paradoxes that need not be present in any of the specific contexts.
Perhaps there are revenge-immune ways of putting together pairs of logics
and truth predicates, but they don’t seem obvious if they exist.
468 K. Scharp
7 Conclusion
There are two different approaches to pluralism: the domain approach
and the context approach. I have argued against the domain approach
and in favor of the context approach. Moreover, the approach that I
advocate for the paradoxes associated with truth can be understood as a
kind of pluralism. Finally, trying to do better than my approach by merg-
ing many logic/truth combinations into a single pluralist view fails. The
problem is that the pluralist mechanism itself provides the materials for
revenge paradoxes.
Notes
1. See Cook (2002), Beall and Restall (2006), Shapiro (2006, 2014),
Russell (2008), and Field (2009).
2. Cr. Wright (1992, 2003), Lynch (2009), Pedersen (2010), Edwards
(2011), Ulatowski (2017), and the papers in Pedersen and Wright
(2013).
3. See Pedersen (2006), Cotnoir (2013b), Caret (2017), Yu (2017), and
Strollo (2018) for discussion.
4. However, see Berto (2015).
5. See Cr. Wright (1992) and Lynch (2009) for examples of aletheic plural-
ists appealing to domains, and see Beall and Restall (2006) and Shaprio
(2014) for examples of logical pluralists appealing to contexts. Note that
Beall and Restall are inconsistent on this matter: at (2006: 88) they deny
that they appeal to contexts as part of their pluralism, but they explicitly
do so throughout the book (2005, 69, 91, 94, 99, 116, 118). They also
say that there are multiple ‘senses’ of words like ‘valid’ (2005, 29), which
would suggest an ambiguity or context-dependence. See Hjortland
(2015) and Caret (2017) for discussion.
6. See Pedersen (2014) and Kouri and Shapiro (2017) for example.
7. See Fantl and McGrath (2002), Hawthorne (2003) and Stanley (2005).
8. See Cohen (1986), DeRose (1992), and Lewis (1996).
9. See Tappolet (1997).
10. See Scharp (2013).
11. Beall (2009).
Aletheic and Logical Pluralism 469
References
Bacon, A. 2015. Can the Classical Logician Avoid the Revenge Paradoxes?
Philosophical Review 124 (3): 299–352.
Beall, J., ed. 2008. Revenge of the Liar. Oxford: Oxford University Press.
———. 2009. Spandrels of Truth. Oxford: Oxford University Press.
———. 2013a. Shrieking Against Gluts. Analysis 73: 438–445.
———. 2013b. Deflated Truth pluralism. In Truth and Pluralism: Current
Debates, ed. Nikolaj J.L.L. Pedersen and C.D. Wright. Oxford: Oxford
University Press.
Beall, J., and G. Restall. 2006. Logical Pluralism. Oxford: Oxford University
Press.
470 K. Scharp
B C
Beall, Jc, v, 4, 12–17, 20, 26n16, Carnap, Rudolp, 14, 218, 229,
27n22, 27n27, 65, 124n2, 230–231n2, 356
142n3, 190n7, 217–230, 238, Cases
240, 260n6, 302, 305–307, constructions, 13, 220, 305
312, 313, 321n2, 323n24, situations, 13, 219, 221, 225,
324n33, 324n34, 344, 345, 233n11, 305
366, 391, 392, 398, 402, 404n7, worlds, 13, 305, 393
425n8, 425n12, 430, 433, Circularity, epistemic, 185–188
449n8, 450n16, 451n22, 454, Closure, semantic, 310–313,
456, 458, 459, 462, 467, 468n5 318–320
Bivalence, 371, 375, 376, 378, 379, Commutativity, 15, 16, 241, 247,
384n15, 385n18, 397, 398, 262n7, 266n22
432, 433, 436, 442, 450n15 Consequence, logical, 14, 66, 102,
Blackburn, Simon, 9, 26n11, 59n14, 219, 238, 251–259, 301,
64, 66–72, 74, 76–78, 80, 331–357, 376, 388, 408, 430,
81n11–13 455