Chapter 13

NORMS AND MODALITY

Amie L. Thomasson

Claims about necessity and possibility play a central role in metaphysical debates. Consider clas-
sic puzzles about material constitution: a statue may be made of a lump of clay, but it seems that
the lump of clay could survive certain changes in shape that the statue could not survive. But how
could they differ in these ‘modal properties’ when the statue and clay are otherwise identical?
Or consider the ancient Ship of Theseus problem: if the planks of a ship are gradually replaced
with new ones, and then the old planks are reassembled, which is the original ship: the one with
the new planks, or the old? Can a ship survive changes in all of its parts? Can it survive disas-
sembly and reassembly? Other puzzles arise about personal survival and identity, when we ask
whether persons could survive the loss of memories, brain transplants, or teleportation.1
Answering metaphysical questions like these requires us to determine which statements
about what is metaphysically possible or necessary are true. But how can we do that? If we think
of our modal claims as describing the world, it’s natural to think that we have to find out
whether there are the needed truthmakers for our claims of metaphysical necessity or possibility.
However, as I will argue in this chapter, the search for truthmakers for metaphysical modal
claims leads to a morass of ontological and epistemological problems. Here I will argue for a
different approach to understanding metaphysical modal claims: thinking of them not as serving
to describe modal features of the world (nor as other possible worlds), but rather serving a norma-
tive function of conveying semantic rules and their consequences. Understanding metaphysical
modal claims in this way, I will argue, enables us to demystify the ontology and epistemology of
modality, and to clarify the epistemology of metaphysics.

13.1 The search for modal truthmakers

It has become standard to think that, as Tony Roy puts it, “the problem of modality is a problem
about truthmakers for modal propositions” (2000, 56). But what could serve as the truthmakers
for our claims of metaphysical necessity or possibility? Are they modal features, or this world—
modal facts, or modal properties? Call a view a form of ‘heavyweight modal realism’ if it holds
that there are distinctively modal properties or facts that explain what it is that makes some modal
statements true. It is tempting to think that when we say: “A person can survive a transplant of
her brain into another body”, or “A painting cannot survive a process that involves replacing
40% of the original surface paint” these statements simply describe what modal properties people

146

9781138823310_C013.indd 146 20-10-2020 04:21:13

 Norms and modality

or paintings have (the modal property of (not) possibly-surviving-this-change), just as we

describe the person or painting as having other properties (weight, color, etc.). Alternatively, we
might think of modal statements as attempts to describe modal facts in the world, and as holding
true if those modal facts indeed obtain.
But heavyweight modal realism faces what Huw Price (2011) has called the ‘placement prob-
lem’: how are these modal facts or properties supposed to fit into the natural world? They do
not seem to be physical properties or facts like those investigated by the empirical sciences—so
what are these modal facts or properties supposed to be, and why should we think there are such
things? How are they related to the non-modal facts or properties studied by the natural sci-
ences? As the case of the clay and the statue makes clear, this relation is pretty mysterious—
because objects like the statue and clay can have all the same physical properties and yet differ
in their modal properties. As a result, it seems like we can’t think of the modal properties as
‘higher level’ properties that are somehow fixed by an entity’s more basic physical, or other non-
modal properties.
If we think that, to know which metaphysical statements are true, we have to come to know
their truthmakers, we face even more difficult epistemological problems. For how can we hope
to discover what modal properties an object has, or what modal facts obtain? Hume observed
long ago that there seems no prospect of giving an empirical account of our knowledge of
modality—so what sort of account can we give? Even if one thinks, with Barbara Vetter, that one
can acquire knowledge of dispositions or potentialities empirically (having observed many
glasses shattering, we infer that glasses are disposed to break on sharp impact) (Vetter 2015,
11–13), the same methods cannot ground knowledge of distinctively metaphysical modality. For
the same observations of the statue/lump of clay lead to different modal conclusions: that the
statue would not survive a squashing while the lump would. The so-called ‘grounding’ problem
arises precisely because no non-modal/non-sortal properties that can be empirically known
seem capable of grounding the difference in the metaphysical modal properties (identity and
persistence conditions) attributed to the statue and the clay.2
One prominent alternative to heavyweight modal realism is David Lewis’s (1986) possible
worlds realism. On this view, we need not accept that there are any distinctively modal features
of this world. Instead, we accept that there are many other concrete worlds causally and spatio-
temporally disconnected from our own—call these the (merely) possible worlds (our actual
world is also a possible world). Lewisian possible worlds realism still enables us to understand
modal statements as descriptions, but they are seen as attempted descriptions not of modal fea-
tures of our world, but of non-modal facts in one or more of the possible worlds (including our
actual world).3
However, few have been willing to accept that there are such possible worlds, or that they
could adequately serve as non-modal truthmakers for our modal propositions (Jubien 2007;
Divers and Melia 2002). Even if we are content to let the ontological issues slide and allow that
there are possible worlds, a massive problem of relevance remains—a problem that was raised in
reviews of The Plurality of Worlds by Nathan Salmon (1988), Graeme Forbes (1988), William
Lycan (1988), and Allen Stairs (1988). The problem is that our modal statements just don’t seem
to be about what goes on in other worlds, even if there are such worlds (Stairs 1988, 344); these
don’t seem like relevant truthmakers.4 The fact that even if we accept the ontology, its relevance to
the modal question remains in doubt, might already be taken to suggest that the modal claims
aren’t aiming to describe facts about other worlds at all.
Lewis’s possible worlds realism also faces daunting epistemic problems like those that trouble
the heavyweight realist. For given the causal isolation of the other possible worlds from our own,
it remains unclear how we could know anything about them, or about what claims they make

147

9781138823310_C013.indd 147 20-10-2020 04:21:13

 Amie L. Thomasson

true. Lewis (1986, 104–115) does suggest that we commonly come to have the modal beliefs we
do by way of engaging in imaginative experiments guided by a principle of recombination
(113–114). But it is not at all clear, on his view, why this sort of procedure should give us any-
thing that counts as modal knowledge: why should imagination guided by this principle give us
any information about what is going on in causally and spatio-temporally disconnected con-
crete worlds? Lewis himself makes it clear that he does not take himself to be answering that
question.5

13.2 Exposing the descriptivist assumption

As we have seen, a daunting array of puzzles and problems arises from the attempt to find truth-
makers for our metaphysical modal statements—whether we think of the truthmakers as modal
properties, modal facts, or possible worlds. What I want to call attention to in this chapter is that
they all arise from an assumption so common that it has become almost invisible: the assumption
that modal statements are descriptive or representational in character—that is, that they function to
describe or track certain ‘features of reality’, which can then serve to make modal claims true.6
I will argue that this assumption should be brought to light and reexamined. Before we start
asking what the truthmakers are for modal claims, we should step back to ask a more basic ques-
tion: what function does it serve to make metaphysical modal claims? Do they serve a kind of
describing or world-tracking function, or do they have another function entirely—in a way that
might make the search for truthmakers otiose?
In one sense it might seem just obvious that metaphysical modal statements are descriptive.
If I say ‘Mary is necessarily human’, this sounds parallel to saying ‘Mary is unusually tall’; both
are indicative in form and are naturally thought of as ‘describing’ (or ‘representing’) Mary in
certain ways. I certainly do not wish to deny these obvious truths.7 The philosophical assump-
tion of descriptivism, however, goes beyond these truisms: it involves an assumption about func-
tion. Many of the basic terms in our language seem to serve the function of tracking certain
features of our environment, with which they are meant to co-vary, enabling us to get around
better. So, for example, it is plausible that terms such as ‘wolf ’ and ‘river’ serve such a descriptive
tracking function.To assume that a term is descriptive in our sense is to assume that it serves that
function. Huw Price (2011) calls terms that serve this function ‘e-representations’, those whose
job “is to co-vary with something else—typically, some external factor, or environmental condi-
tion” (20). Where terms serve this kind of function, it is natural to think of them as aiming to
correctly represent what there is (and is not) in the world—and answerable to the world in the
sense that we look to the world to determine if what we say using those terms is true or false.
When discourse is descriptive in this way, it seems natural to look to the world to find ‘truth-
makers’8 for our claims about wolves or rivers, features of the world that ‘explain how sentences
about the real world are made true or false’ (Mulligan et al. 1984, 288).
But there are plausible and interesting philosophical accounts of some central, philosophi-
cally interesting terms that treat them as serving very different functions from describing or
tracking elements of our environment. For example, Paul Horwich (1998) treats the function of
the truth predicate as serving as a device of generalization; Stephen Yablo (2005) suggests that
nominative vocabulary for numbers enables us to simplify our expression of scientific laws; and
moral expressivists like Simon Blackburn (1993) argue that moral discourse serves not to
describe moral facts, but rather to enable us to coordinate our attitudes in certain useful ways.
Even where modal discourse is concerned, the descriptivist assumption wasn’t always so
invisibly dominant. In the early days of analytic philosophy, it was common to deny that modal
statements of various kinds serve a descriptive function. The approach was suggested by early

148

9781138823310_C013.indd 148 20-10-2020 04:21:13

 Norms and modality

conventionalists like Schlick (1918),9 who argued that necessary statements of mathematics and
logic are not descriptive statements, but serve as implicit definitions of concepts. It was devel-
oped in a new way by Wittgenstein in the Tractatus (1922/1933), who treated the propositions
of logic as tautologies which say nothing about the world.10 The approach reappeared in a more
sophisticated vein in the work of the later Wittgenstein, and in Ryle’s work on statements of
scientific laws, which he took to serve not to describe the world but rather to serve as ‘inference
tickets’ (1950, 121). Sellars (1958) develops a similar treatment of statements of scientific laws,
which he treats as having the function of justifying or endorsing inferences from something’s being
an A to its being a B. More recently, approaches to modality along these lines have been devel-
oped by Simon Blackburn (1987/1993) and Robert Brandom (2008).
In the remainder of this chapter, I aim to lay out a view in that tradition, on which metaphysi-
cal modal statements fundamentally serve not to describe features of this or other worlds, but
rather serve a basically normative function. On reflection, it shouldn’t be surprising that modal
terms serve a normative function. As a grammatical group, modal terms include not only the
metaphysician’s ‘possible’ and ‘necessary’, but also such terms as ‘can’, ‘may’, ‘must’, and ‘shall’,
which are characteristically used in issuing requirements and permissions, and in stating com-
mands and rules in an impersonal indicative form. These different sorts of modal terms (for
alethic, deontic, and epistemic modalities) tend to come together across a wide range of lan-
guages (Papafragou 1998, 371), and children tend to learn to use modal terms for obligation,
necessity, and possibility at around the same time (about age three) (Wells 1985, 159–160, 253).
So it would make sense to think that they have something in common—perhaps that that they
all enable us to convey norms in useful and perspicuous ways.

13.3 What function does modal vocabulary serve?

But why might it be useful to have modal terms convey rules and norms? There are surely other
ways of communicating rules and norms, such as in non-modal imperatives, or with a stick. Let
me begin by discussing why modal terms might be useful in conveying rules in general, and then
turn to the particular role of metaphysical modal terms in conveying semantic rules.
To see what functions modal terms might serve in conveying rules, think about the way rules
of games are expressed in language. Some rules may be expressed in imperatives, for example, for
checkers:

• “Move only on the dark squares”

But rules can also be expressed in declarative sentences (in the indicative mood), as:

• “Players move their pieces only on the dark squares”, or

• “Black always moves first”

Putting rules in the indicative mood has certain advantages over the imperative. For starters, one
can’t easily state “Black moves first” in an imperative form without knowing who the black
player is, and addressing him or her directly. Secondly, expressing rules in the indicative mood
enables us to make explicit our ways of reasoning with rules, so that we can, for example, embed
them in conditionals and say “If black always moves first, then red never moves first”, whereas
we cannot put an imperative in the antecedent of a conditional.
But there are also dangers of expressing rules in the indicative form, since they might be
mistaken for descriptions of what does happen or has happened—making it hard to distinguish the

149

9781138823310_C013.indd 149 20-10-2020 04:21:13

 Amie L. Thomasson

expression of the rule that black always moves first from the red player’s misguided complaint
that black always moves first.
We can, however, add a modal verb, and say instead:

• “Players must move their pieces only on the dark squares” or

• “The black player must move first”

Expressing rules in this modal form preserves our ability to make explicit our ways of reasoning
with rules—since it is still in the indicative. But it also brings other advantages. First, it clearly
distinguishes these statements of rules from mere descriptions of what does happen. Second, it
enables us to express permissions as well as requirements. Neither the imperative form nor the
simple indicative enabled us to do that: if we ask, “Does each player take a turn every round?”
the only way to give a negative answer to this question, as Ryle (1950/1971, 244) pointed out, is
to add a modal verb, and say “No, a player may choose to skip a turn”. These observations lead
to the hypothesis that at least one function it serves to have modal terminology in our language
is to give us a way of expressing rules or norms in the indicative mood, in a way that makes the
regulative status more explicit, enables us to make explicit our ways of reasoning with rules, and
enables us to express permissions as well as requirements.

13.4 The function of metaphysical modal claims

But if modal terms have as their function conveying rules or norms, what rules or norms might
be at stake in metaphysical modal claims? The heart of the normativist view of modality is to see
metaphysical modal claims as functioning to convey semantic rules. The interesting and tricky
feature, however, is that they do not do so by describing what the semantic rules are. If we said
that they describe what the rules are—that the adoption of semantic rules is a truthmaker for our
metaphysical modal claims—then we would still be stuck with the old descriptivist assumption
that we need truthmakers for our modal claims. Moreover, we would fall into the problems of
classical conventionalism—our necessary claims would only hold contingently, as it is a contin-
gent matter that we adopt some rather than other semantic rules.
But nor do metaphysical claims of necessity convey semantic rules by stating these rules in a
metalanguage: metaphysical modal claims are object-language claims. Instead, they convey these rules
by using those very terms—remaining in the object language. Consider the following dialogue:

• Child: “Mom, is Aunt Sophie always going to be a bachelor?”

• Mom: “Bachelors must be men, dear”.
• (Pretentious philosophical mom: “It is necessary that bachelors are men, dear”.)

What is going on in this dialogue? In the response, Mom just uses the term ‘bachelor’ and
states a necessary truth: It is necessary that bachelors are men. But what she is doing thereby is
communicating a rule that could be stated in the metalanguage, as “The term ‘bachelor’ may only
be applied to men”.
Now consider a more philosophical dialogue.

• Question: “Can a ship survive having all of its parts gradually replaced?”
• Response: “Yes, as long as the replacement process is gradual. For all that is essential to artifact
identity over time is a continuous history of maintenance, not the retention of any particular
material part”.

150

9781138823310_C013.indd 150 20-10-2020 04:21:13

 Norms and modality

Again, here we have a dialogue conducted in the object language, about ships and other
artifacts. But what is being done through this dialogue, on the normativist view, is not describing
some modal properties of ships that the philosopher pretends to have discovered, but rather a
way of communicating some rules of use for ship names (and names for other artifacts): that we
are permitted to say this is ‘the same ship’ as before, or to re-apply the name ‘The Queen Mary’,
as long as there has been a continuous history of maintenance.
Although the primary function of metaphysical modal claims, on this view, is to convey
semantic rules rather than to report metaphysical discoveries, it is still useful to do so by just using
the terms, in the object language. We often use terms as a way of demonstrating or implicitly
commenting on how the term is (to be) used, or whether it should be used at all. Chris Barker
(2002) calls these ‘metalinguistic’ uses as contrasted with ‘descriptive’ uses, where metalinguistic
uses are those in which a term is used to “communicate something about how to use a certain
word appropriately”, rather than to communicate (other) information about the world.11 For
example, we engage in what has become known as ‘metalinguistic negation’, when one speaker
says, “The performance was good tonight”, and another replies, “It wasn’t good, it was spectacu-
lar!” In a case like this, the second speaker apparently uses language in order to show what choice
of words she thinks was appropriate. In other cases, we may sometimes demonstrate how vague
terms are appropriately used in a context by using them in certain ways. In some contexts, for
example, one might communicate what the standards for tallness are around here by pointing to
a man (whose height is not in doubt) and saying “Jones is tall” (see Barker 2002, 1–2)—in which
we are not adding information about Jones’s height, but rather using the term ‘tall’ in a way that
communicates information about how it is appropriately used in this context.
In short, it is not unusual or idiosyncratic for us to communicate standards for language use
by using it in certain ways. That is exactly what the normativist thinks is going on with claims
about what is metaphysically possible and necessary: they are claims in the object language, and
so in that sense are world-oriented, ‘about the world’, not about language (just as the aforemen-
tioned claims are about the performance or about Jones).Yet their function is to convey how the
terms ought to be used—to convey norms. In the case of metaphysical modal claims, these are
semantic norms, typically concerning actual and hypothetical cases in which the term should be
applied and refused, or applied again ‘to the same thing’. What the addition of the modal verb
does, over the simple indicative, is to ‘flag’ this regulative function, making it more explicit, and
enabling us to convey permissions as well as requirements.
So, in sum, on the analysis given here, there are two odd features of metaphysical modal
claims. First, like other modal statements, though they fulfill a normative function, rather than
being expressed in imperatives they are expressed in indicative form—for good, functional rea-
sons. Second, though simple utterances of claims of metaphysical necessity are in the object lan-
guage, they involve implicitly metalinguistic uses of the terms—as ways of conveying something
about how the relevant terms are to be used. Both of these features can lead us astray into think-
ing of modal statements as if they are worldly descriptions in need of truthmakers. But once
we’ve noticed the commonalities with other cases in which normative modal language is
expressed in indicatives, and object-language claims are used to serve a metalinguistic purpose,
we can see that they are not so very strange after all, and they can see our way clear to a more
plausible, and less problematic, analysis of modal discourse that does not begin from the drive to
seek modal truthmakers in this or other worlds.
But although the normativist doesn’t think of modal claims as needing truthmakers, a nor-
mativist can nonetheless allow that our modal claims are true or false. Given the rules that (on
this view) govern the use of our modal terms themselves, we are entitled to add ‘necessarily’ onto
any object-language expression of an actual semantic rule. So we can begin from “All bachelors

151

9781138823310_C013.indd 151 20-10-2020 04:21:13

 Amie L. Thomasson

are men” and (since that claim is an object-language expression of a semantic rule) add ‘neces-
sarily’ and assert “Necessarily, all bachelors are men”. We then need only adopt a deflationary
understanding of truth (see, e.g. Horwich 1998), according to which the concept of truth is
simply governed by the equivalence schema: <p> is true iff p, to recognize the equivalence of
this with “<Necessarily, all bachelors are men> is true”.The uncontroversial equivalence schema
applies just as well to modal as to non-modal indicatives, so there is no problem in allowing that
modal claims may be true, stated in propositional form, and used in reasoning.

13.5 Remaining challenges and hopes

The road to developing a full modal normativist approach is long and full of challenges. Some
of the challenges I have discussed elsewhere12—including showing how it avoids the problems
that plagued classical conventionalism, and showing how to avoid the Frege-Geach problem by
giving the meaning of modal discourse, not just its function or use. Another major and familiar
challenge is showing how this view can accommodate not only simple a priori necessities like
“Necessarily, all bachelors are men”, but the sorts of a posteriori and de re necessities Kripke
(1980) famously called attention to: such as ‘Water is necessarily H2O’ and ‘Elizabeth Warren is
necessarily human’. The first key to handling such necessities is to accept that even names and
kind terms have some conceptual content.13 The second key is to note the varied and often
ostensive and world-deferential forms semantic rules can take. There is no space to develop that
solution here, but the original approach was developed in Sidelle (1989) and is applied specifi-
cally to the normativist approach in Thomasson (2020).
At any rate, if we can make a normativist view along these lines work, it will be very attractive.
One advantage of this view is ontological, and comes from not thinking of modal properties,
modal facts, or possible worlds as things our modal claims aim to describe or track, and that are
capable of explaining what makes them true.The normativist may allow that there are modal facts
and properties—and even other possible worlds—but not in the sense of ‘positing’ them to ‘explain’
the truth of our modal claims. Instead, the ontological entitlement to say that there are such things
is explanatorily ‘downstream’ from such truths.That is, we can start by making metaphysical modal
claims (which have another function entirely); for example, “Necessarily, all bachelors are men” or
“Obama is necessarily human”. From these modal truths, we can make trivial inferences to the
existence of modal facts (that it is a fact that it is necessary that all bachelors are men) and proper-
ties (that Obama has the property of being necessarily human). But we arrive at talk of these modal
facts and properties by hypostatizations out of our modal expressions; we do not need to start by
‘discovering’ these worldly modal features to figure out which modal claims to accept.
The most important advantage of a normativist approach is epistemological, that of avoiding
the notorious difficulties heavyweight realist views and possible worlds realist views face in
explaining our knowledge of modal facts. If we think of ourselves as trying to track and describe
the modal properties, modal facts, or possible worlds (the presence of which could make the
claims true), it is very difficult to see how we could have any (quasi-)perceptual or intellectual
access to these modal facts or properties or worlds. The normativist demystifies modal knowl-
edge by considering the move from using language to knowing basic modal facts to be a matter
of moving from mastering the rules for properly applying and refusing expressions (as a compe-
tent speaker), to being able to explicitly convey these constitutive rules and their consequences in
the object language and indicative mood.
A third, related, advantage of the normativist view is its helpfulness in clarifying the method-
ology of metaphysics and justifying the use of intuition in modal debates. If we think of meta-
physicians as trying to ‘detect’ modal properties of objects, the hopes of adjudicating debates

152

9781138823310_C013.indd 152 20-10-2020 04:21:13

 Norms and modality

about the modal features of persons or works of art seem slim: for no one, it seems, has any useful
answer to the question of how they are supposed to be detected. Moreover, although it is com-
mon to rely on intuitions to support metaphysical views, it’s not clear how to justify why intu-
ition should be thought a reliable guide to the modal features of the world (when it certainly fails
to be a reliable guide about most other features) (see Sosa 2008, 233). But on the normativist
view, we have good reason for thinking that intuitions of competent speakers may play a useful
role in revealing and making explicit the actual semantic rules, and thereby in coming to express
modal truths in the object language—and signaling this with the addition of modal verbs.
Nonetheless, as I have argued elsewhere, some uses of metaphysical modal claims may be fruit-
fully seen as engaged in what David Plunkett and Tim Sundell (2013) have called ‘metalinguistic
negotiation’: as ways of advocating for changes in the rules—whether to precisify them or alter them
in other ways, in order to serve various purposes, rather than simply as ways of conveying the rules
there are.14 The function of the metaphysical modal claims may still be normative, but it may have
to do more with pressing for changes in the rules than with communicating or enforcing the
extant rules.The fact that, in the object language, we may often engage in this kind of metalinguis-
tic negotiation of what how our terms should be used enables us to account for the fact that debates
about metaphysical modality are often enduring and hard to resolve—even among competent
speakers. For what is at issue is not just what rules do govern the terms (though these, too, may be
imprecise, open-ended, contextually variable), but what rules should govern our terms—where this
is sensitive to a range of other issues about what we should value and how we should live.
But whether they are used with the force of communicating those semantic rules there are,
or of pressing for the rules the speaker thinks there ought to be, seeing metaphysical modal
claims as having a normative function, to do with conveying and enforcing semantic rules, prom-
ises to do a great deal to demystify the epistemology of modality—and with it, the methodology
of metaphysics.

Notes
1 While there are many other sorts of modal claim (e.g. asking what is physically possible or logically
possible), I will focus here on claims like these—about what is metaphysically possible or necessary—
since those play a central role in many metaphysical debates.
2 For discussion of the grounding problem, see Burke (1992), Zimmerman (1995), Bennett (2004), and
Thomasson (2007, ch. 4).
3 I do not mean to suggest that Lewis himself thought of his possible worlds as truthmakers for modal
propositions, only that possible worlds realism along Lewis’s lines is capable of supplying the truthmak-
ers that the truthmaker theorist needs.
4 In the latter form this is the famous ‘Humphrey’ objection Kripke (1980, 45 n. 3) raises against Lewis.
5 For further critical discussion of Lewis’s reply to the knowledge problem, see Bueno and Shalkowski (2004).
6 Huw Price uses the terminology ‘representational’ and puts the point in terms of denying that all dis-
course is ‘e-representational’. See Price (2011) for the original formulation and criticisms of the repre-
sentationalist/descriptivist assumption.
7 Price similarly notes that those he calls ‘non-facutalists’ (our non-descriptivists) may accept that moral
claims, for example, are ‘statements in some minimal sense’ (‘Semantic Deflationism and the Frege
Point’, p. 3 in pdf).
8 Though I do not mean to endorse truthmaker theory here—only to point out that, however plausible
it is in descriptive cases, it leads us astray in others.
9 Schlick, in turn, was developing ideas originating in Hilbert’s Foundations of Geometry and attempting
to generalize them to the cases of logic and mathematics. See Baker (1988, 187ff).
10 A neo-conventionalist view has also been developed by Alan Sidelle (1989).
11 Barker distinguishes when adjectives have a ‘descriptive’ use from when they have a ‘metalinguistic’ use
(2002, 1).The thought here is that modal terms signal that the assertion in question has a metalinguistic
rather than a descriptive use.

153

9781138823310_C013.indd 153 20-10-2020 04:21:13

 Amie L. Thomasson

12 See my (2020).
13 I (Thomasson 2007) and others (Devitt and Sterelny 1987) have argued for this position elsewhere.
14 See Plunkett (2015) and Thomasson (2020) for developments of the idea that certain metaphysical
debates may be seen as engaged in metalinguistic negotiation.

References
Baker, Gordon. 1988. Wittgenstein, Frege and the Vienna Circle. Oxford: Blackwell.
Barker, Chris (2002). “The Dynamics of Vagueness”. Linguistics and Philosophy 25: 1–36.
Bennett, Karen (2004). “Spatio-Temporal Coincidence and the Grounding Problem”. Philosophical Studies
118: 339–371.
Blackburn, Simon (1993). Essays in Quasi-Realism. New York: Oxford University Press.
Brandom, Robert (2008). Between Saying and Doing: Towards an Analytic Pragmatism. Oxford: Oxford
University Press.
Burke, Michael. (1992). “Copper Statues and Pieces of Copper: A Challenge to the Standard Account”.
Analysis 52(1): 12–17.
Devitt, Michael and Kim Sterelny (1987). Language and Reality. Cambridge, MA: MIT Press.
Divers, John and Joseph Melia (2002). “The Analytic Limit of Genuine Modal Realism”. Mind 111(441):
15–36.
Forbes, Graeme (1988). “Critical Study: The Plurality of Worlds”. Philosophical Quarterly 38(151): 222–240.
Horwich, Paul (1998). Truth. 2nd edn. Oxford: Oxford University Press.
Jubien, Michael (2007). “Analyzing Modality”. Oxford Studies in Metaphysics 3: 99–139.
Kripke, Saul (1980). Naming and Necessity. Oxford: Blackwell.
Lewis, David K. (1986). On the Plurality of Worlds. Oxford: Blackwell.
Lycan, William (1988). “Review of ‘On the Plurality of Worlds’”. Journal of Philosophy 85(1): 42–47.
Mulligan, Kevin, Peter Simons, and Barry Smith (1984). “Truth-Makers”. Philosophy and Phenomenological
Research 44(3): 287–321.
Papafragou, Anna (1998). “The Acquisition of Modality: Implications for Theories of Semantic
Representation”. Mind and Language 13(3): 370–399.
Plunkett, David and Tim Sundell (2013). “Disagreement and the Semantics of Normative and Evaluative
Terms”. Philosopher’s Imprint 13(23) (December): 1–37.
Price, Huw (2011). Naturalism without Mirrors. Oxford: Oxford University Press.
Roy, Tony (2000). “Things and De Re Modality”. Nous 34: 56–84.
Ryle, Gilbert (1950/1971). “‘If ’, ‘So’, and ‘Because’”. In Collected Papers,Vol. 2. London: Hutchison.
Salmon, Nathan (1988). “Review of ‘On the Plurality of Worlds’”. Philosophical Review 97(2): 237–244.
Schlick, Moritz (1918). Allgemeine Erkenntnislehre. Berlin: Springer.
Sellars,Wilfrid (1958).“Counterfactuals, Dispositions and the Causal Modalities”. In Herbert Feigl, Michael
Scriven, and Grover Maxwell, eds., Minnesota Studies in Philosophy of Science,Vol. 2: Concepts,Theories and
the Mind-Body Problem. Minneapolis: University of Minnesota Press, 225–308.
Sidelle, Alan (1989). Necessity, Essence and Individuation: A Defense of Conventionalism. Ithaca: Cornell
University Press.
Sosa, Ernest. 2008. “Experimental Philosophy and Philosophical Intuition”. In Joshua Knobe and Shaun
Nichols, eds., Experimental Philosophy. Oxford: Oxford University Press, 231–240.
Stairs, Allen (1988). “Review Essay: ‘On the Plurality of Worlds’”. Philosophy and Phenomenological Research
49(2): 333–352.
Thomasson, Amie L. (2007). Ordinary Objects. New York: Oxford University Press.
——— (2020). Norms and Necessity. New York: Oxford University Press.
Vetter, Barbara (2015). Potentiality. Oxford: Oxford University Press.
Wells, Gordon. 1985. Language Development in the Preschool Years. Cambridge: Cambridge University Press.
Wittgenstein, Ludwig (1922/1933). Tractatus Logico-Philosophicus. Translated by C. K. Ogden. London:
Routledge.
Yablo, Stephen (2005). “The Myth of the Seven”. In Mark Eli Kalderon, ed., Fictionalism in Metaphysics.
Oxford: Oxford University Press.
Zimmerman, Dean (1995). “Theories of Masses and Problems of Constitution”. Philosophical Review 104:
53–110.

154

9781138823310_C013.indd 154 20-10-2020 04:21:13