1
Friederike Moltmann fmoltmann@univ-paris1.fr
January 2013
In every day speech, talk about ‘cases’ is abundant. The noun case in English in fact is an extremely useful linguistic device, and not only in the more particular contexts of medicine and law. Surprisingly, though, the noun case has hardly received any attention in semantic theory.
1
The noun case plays a role in a variety of constructions that are of significant interest not only to linguistic semantics, but also to philosophy, more specifically ontology and the debate concerning the notion of truthmaking.
The following main constructions with case can be distinguished:
Nominal case -NPs:
(1) a. the case of the stolen statue
b. a case of flu
c. the case of a defeat
Clausal case -NPs:
(2) a. the case in which it might rain
b. the case in which a student fails the exam
Conditionals and modals
(3) a. In case it rains, we won’t go. In that case we will stay home.
b. John might go to the party. In that case, I will go too.
The noun case has counterparts in a number of other European languages, such as German
( Fall ), French ( case ), Italian ( caso ), and Spanish ( caso ), which exhibit more or less the same constructions.
1 Sometimes, though, philosophers have made use of the noun case in a somewhat technical philosophical sense, as when Lewis (1975) talks about adverbs of quantification like something ranging over ‘cases’ and Woltersdorff
(1970) calls tropes ‘cases’.

2
The noun case is challenging for a semanticist since the different constructions in which case can figure call for a unified semantic analysis. The noun case also bears on the semantics of complex noun phrases in general as well as that of conditionals and modals.
The semantics of constructions with the noun case is also of interest to ontology. Case -NPs are referential NPs evidently referring to ‘cases’. But cases are not ordinary objects. A medical case is distinct from a patient and a legal case from the act of the crime. We will see that cases are not identical to any sort of familiar object. In particular, they are not identical to material objects, events, tropes, facts, states of affairs, or ‘possibilities’. Rather they form an ontological level of their own. In fact, they generally come with their own existence predicate, shared with no other type of entity. Thus, to clarify the nature of cases is a challenge to ontological theory.
Cases often seem to play the role of situations, in particular in conditionals and in regard to modals. The notion of a case thus could clarify or even perhaps replace the notion of a situation with respect to some of the semantic roles it is taken to play.
Cases finally may be of considerable interest philosophically given the role they appear to play in sentences with the predicate is the case :
(4) a. It is not the case that S.
b. It is sometimes the case that S.
Is the case , at first sight, seems to have just the same semantic function as is true . However, that there are significant semantic differences. In fact, is the case does not act as a truth predicate, but rather, as I will argue, expresses the truth-making relation, relating situationlike cases to the that -clause. Thus, (4b) involves existential quantification over situation-like cases as truth makers for the sentence S. The occurrence of the noun case in that construction indicates that cases play the role of truth makers, and it provides a linguistic ‘inspiration’ to consider cases in general to play that role, as entities distinct from any of the more familiar types of entities. We will see that there are philosophical considerations that provide motivations of their own in favor of cases playing that role.
I will propose that cases are ‘filtered objects’ or ‘filtered pluralities of objects’. They are obtained, so to speak, by reducing an ordinary object or plurality of objects to just some of its properties (including relational properties) relative to a general condition.

3
I will start with a number of observations about clausal case -NPs. I will then look more closely at nominal case -NPs and propose a particular ontological account of the sorts of entities they describe. This account will then be extended to the entities described by clausal case -NPs. Finally, I will argue that is the case expresses truth-making with cases being suitable candidates for the truth-making role in general, and I will sketch semantic analyses of the various case -constructions on the basis of the notion of truth-making.
Sometimes, I will make use of examples from German and French, since constructions with German Fall and French cas sometimes are more explicit and systematic. In general, all that is said about English case -constructions holds for the corresponding constructions in the other European languages as well that have a noun for ‘case’.
1. Clausal case-NPs
1.1. Situation-like cases
There are two kinds of clausal case -NPs, that is, NPs with case as head nouns and a clausal modifier specifying the ‘content’ of a case. The first type, illustrated in (5), appears to describe a particular possible situation; the second type, illustrated in (6), a kind of situation: 2
Particular clausal case -NPs
(5) a. the case in which John will return
b. The case in which it might rain
Generic clausal case -NPs
(6) a. the case in which it rains on a Sunday
b. the case in which a student fails the exam
2 In German clausal case -NPs are of a somewhat different syntactic construction:
(i) der Fall, dass Hans zurueckommt
the case that John returns
This construction is of the very same type as complex NPs in English of the sort the fact that S or the belief that
S. Characteristic of that construction is that the that -clause does not have the status of a complement or an adjunct of the noun and the determiner is obligatorily the simple definite determiner. Semantically, there does not seem to be a difference between the English and the German construction.

4
Particular clausal case -NPs require a modal (or future tense) in the which -clause. Without a modal, they are unacceptable, as seen below:
(7) a. We discussed the case in which John might get elected.
b. We discussed the case in which John won’t return.
c. ?? We discussed the case in which John returned yesterday.
Obviously, the modal indicates the status of the situation referred to as a merely possible one.
Clausal case -NPs are somehow not allowed to refer a factual situation, which would be indicated by the absence of a modal.
The noun case itself is not subject, though, to a restriction to be applicable only to merely possible situations. First, the present case in in the present case can anaphorically refer to a factual situation-like case. Moreover, generic clausal case -NPs are not subject to a modal requirement, and they can include an actual situation-like case in their denotation:
(8) What happened to John is terrible. The case in which a student failed the exam has never
occurred before.
Generic clausal case -NP are kind-referring NPs in the sense of Carlson (1977) (even though they are not of the form of bare plurals or mass nouns). They refer to a kind whose instances are specific situation-like cases. Evidence for kind reference comes first of all from the applicability of typical kind predicates, as in (8) and below:
(9) a. The case in which someone passes the exam is rare / unusual.
b. The case in which someone passes the exam does not occur often.
c. The case in which someone passes the exam has never occurred.
Furthermore, generic clausal caseNPs exhibit the existential reading characteristic of kind terms with stage-level predicates (Carlson 1977):
(10) a. John cannot remember the case in which anyone failed the exam.
b. I have never encountered the case in which a candidate was unable to speak during
the oral exam.

5
Clausal case -NPs behave quite differently from fact descriptions, descriptions of the sort that fact that S. The latter are not subject to a modal requirement and cannot be used to refer to ‘kinds of facts’.
The fact that someone failed the exam, with an unspecific reading of someone , for example, refers to a quantificational fact. Clausal case -NPs cannot refer to
‘quantificational’ cases, but only to cases involving particular individuals. If many people failed the exam, there are many cases in which someone failed the exam, not a single case in which someone failed the exam. However, there is only â single fact described by the fact that someone failed the exam . Facts described by explicit fact descriptions are non-worldly facts in the sense of Strawson (1949) (entities that correspond to true propositions), not worldly facts in the sense of Austin (1961b) (entities that are part of world).
1.2. Characteristics of situation-related cases
The question now is, what are situation-like cases? For example, are they worldly facts (or possible facts), possibilities, events, or states of affairs?
3
A remarkable feature of situation-like cases is that they come with their own existence predicate. This requires a brief remark concerning existence predicates in natural language in general. Existence predicates have a particular semantic characteristic in that they may yield true sentences with empty subjects and negation, as we see with the verb exist below:
(11) Vulcan does not exist.
Other predicates under the same circumstances generally yield sentences that intuitively are neither true nor false.
Existence predicates in natural language are to an extent restricted to particular types of objects. Thus, exist applies to material and abstract objects (or empty terms describing them) as in (11) and (12a), but not to events, as seen in (12b):
(12) a. The number four exists.
b. ??? The accident existed yesterday.
3 To say that situation-like cases are situations does not help much since the notion of a situation in semantics is a rather technical one and allows for various sorts of construals.

6
The existence predicates that select events are instead happen and take place . They in turn resist material and abstract objects:
(13) a. The accident never happened / took place.
b. ??? The planet / The number four happened / took place.
Obtain is yet another existence predicate, reserved for entities like facts, laws, and conditions:
(14) The law / condition no longer obtains.
None of those existence predicates naturally apply to cases:
(15) a. ??? The case in which John will not return might exist / might take place / might
happen.
b. ??? The case in which it rains on a Sunday has never existed / happened / taken place /
obtained.
Instead, there are special existence predicates for cases, namely present itself and occur (the latter can apply also to certain types of events):
(16) a. The case in which John will not return could occur / present itself.
b. The case in which it rains on a Sunday has never presented itself / has never occurred.
In some other European languages, the choice of a ‘case’-specific existence predicate is even more remarkable. Thus, German choses eintreten ‘enter’ is the existence predicate reserved for cases:
(17) a. Der Fall, daβ Hans nicht zurückommt, ist nicht eintreten.
‘The case that John might not return could enter’.

7
Eintreten as an existence predicate applies to no other sort of entity (except to certain sorts of events). In French, the existence predicate for cases is se produire
‘produce itself’ (which also applies to certain types of events):
(17) b. Le cas ou Jean retourne ne s’est pas produit.
‘The case that John returns did not produce itself’
If the choice of existence predicates is taken seriously as an indication of ontological categories, then several conclusions can be drawn. First of all, cases are not (worldly or nonworldly) facts. Cases can be merely possible, whereas facts cannot. Moreover, cases do not go along with the existence predicate obtain , the existence predicate specific to facts and similar entities. Second, cases are not ‘possibilities’. Possibilities as ‘mere’ possibilities
‘exist’ ( the possibility that John may never return exists ). By contrast, merely possible cases do not ‘exist’, that is, occur or present themselves. If the existence predicate occur or the existence predicate present itself is true of a case, it means that the case is a factual one, not a merely possible one. Finally, cases are not states of affairs. States of affairs exist whether or not they obtain. By contrast, merely possible cases do not ‘exist’ and cases can never be said to obtain.
1.3. Conditional case-constructions
Situation-like cases are also involved in case -constructions of a conditional or quasiconditional sort:
(18) a. In case it rains, we won’t go.
b. We will take an umbrella in case it rains.
In English, the construction is grammaticalized, containing no determiner before case and no complementizer that after it.
4
4 In this construction, case seems to act itself act as the complementizer.

8
(19) a. * in the case it rains
b. * in the case that it rains
Note, though, that the more explicit construction below is marginally acceptable – and, it appears to reflect the semantics that is at stake explicitly:
(21) ? In a case in which it rains, we won’t go.
Moreover, explicit versions of conditional case -constructions are found in German and
French:
(20) a. Im Fall, dass es regnet, werden wir nicht gehen.
in the case that it rains we won’t go
b. Au cas ou il pleut, nous n’allons pas.
in the case where it rains we won’t go
Whereas the explicit version in English involves a singular indefinite case -NP, the German and French versions involve a definite case -NP. The English version contains a particular case -NP, whereas the French and German versions a generic caseNP.
Given the more explicit versions, what appears as the antecedent of the conditional serves to introduce situation-like cases. This is supported by the observation that in a subsequent sentence, a case -NP can be used anaphorically so as to refer back to the case introduced by the case -phrase serving as the antecedent of the conditional:
(21) a. In case it rains, we won’t go.
b. In that case / In such a case, we will stay home.
c. Let’s better not think about that case.
Conditional case -constructions display two readings: a proper conditional reading and a quasi-conditional reading. The conditional reading is exemplified by (18a), the quasiconditional one by (18b). On the quasi-conditional reading, the case -NP refers to a possible future case whose actualization will not be a condition for the main clause to obtain, unlike on the ordinary conditional reading. Rather, the main clause is said to be true in view of one

9 possible future course of events, namely in which a situation-like case as described by the case -NP occurs. Thus, the two readings of conditional case -constructions involve different relations regarding the situation-like case described by the clausal case -NP: a conditional relation and a relation that one may call the ‘in view of’-relation.
1.4. Cases and modals
Modal sentences do not involve an explicit caseconstruction. However, modals of possibility may introduce a situation-like case to serve as the referent of a subsequent anaphoric case -
NP:
(22) John might go to the party. In that case, I will go too.
Note, though, that modals of necessity cannot support case -anaphora, even though, as universal quantifier ranging over circumstances, they are expected to:
(23) a. ??? John must be at home. In those cases, Mary is at home too.
b. ??? John has to submit a proposal. In those cases, Mary will help him.
I will later suggest an explanation of that difference.
2. Nominal case-constructions
How should situation-like cases be conceived of ontologically given that they cannot be identified with facts, possibilities, events, or states of affairs? I will pursue the view that situation-like cases belong a special and more general ontological category, that of ‘cases’.
That is, I will propose that situation-like cases be considered as being of the very same sort as the kinds of cases that are described by nominal case -NPs. In this section, I will take a closer look at nominal case -NPs and develop an account that can then be carried over to situationlike cases as well.
Three types of nominal caseNPs need to be distinguished: those describing cases as instances of universals, those describing cases as correlated with objects, and those describing cases as correlated with events.

10
2.1. Cases as instances of universals
Examples of cases described as instances of universals may be medical or legal cases or cases of a particular art movement:
(24) a. This is a case of insanity.
b. What John has is a case of schizophrenia.
d. The incident is a case of fraud.
e. John’s behavior toward Mary is a case of harassment.
f. This building is an unusual case of art deco.
Not all instances of universals are cases, tough. For an instance of a universal to be a case, it needs to have a particular complexity that does not make it too obviously an instance of the universal. Thus, whiteness and darkness do not have instances that are cases, but wisdom and fraud do:
(25) a. an instance of whiteness / darkness
b. ?? a case of whiteness / darkness
(26) a. an instance of wisdom
b. a case of wisdom
(27) a. an instance of fraud
b. a case of fraud
The cases described above as instances of universals are trope-like or event-like, which is of course because the universals in question have as their instances tropes or events. But also individuals and quantities may be described as cases of universals:
5
5 This is noted for the original French version of (28a) by van de Velde (ms):
(i) J’ai connu des cas de journalistes honnêtes.
Van de Velde also notes that cases correlated with universals may be universals themselves, species of a genus:
(ii) Ménon propose une multiplicité de cas de vertu.
‘Menon offered a variety of cases of virtue.’

11
(28) a. I knew cases of honest journalists.
b. I am not aware of a case of outstanding Indian wine.
What kind of entity may be described as an instance of a universal depends on the universal itself, whether it is a quality, a kind of event, a kind of individual, or a kind of quantity.
2.2 Object-related cases
Case -constructions that describe cases as related to objects are those below:
(29) a. the case of that incident
b. the case of the man that has suffered from this illness for more than 20 years
c. the case of the stolen statue
Here the complement describes what I call the correlated object of the case.
There are NPs of the same construction type that seem to express an identity relation between the referent of the complement and the referent of the entire NP, for example the city of Munich . But this is not so with the case -construction in (29). Generally, a case has different properties than its correlated object and should be considered an entity distinct from it.
First, a case and its correlated object behave differently with respect to predicates expressing object-related attitudes. This is so whether the correlated object is considered a material object or a complex feature or trope, as illustrated by the semantic differences among the following sentences:
(30) a. We studied the case of the disabled student. (as a medical / legal case , ..)
b. We studied the disabled student.
c. We studied the disability of the student.
Understanding (30a) requires understanding what kind of case the case is supposed to be, a legal or medical case, for example. What the case is, in turn, depends on which features of the student or his disability are relevant in the context. It depends on, let’s say, whether the features are features relevant from a medical or legal point of view and thus constitutive of a

12 medical or a legal case. Importantly, the features may include not only intrinsic features of the object in question, but also relations it enters to other things, relational features.
No filtering condition is not required for (30b) and (30c). Here the object of study may simply be the student himself or his disability.
Cases and their correlated objects behave quite differently also as objects of discussion and evaluation:
(31) a. We discussed the case of the book.
b. We discussed the book.
(32) a. The case of the stolen statue is interesting.
b. The stolen statue is interesting.
c. The theft of the statue is interesting.
(33) a. John compared the case of the first student to the case of the second students.
b. John compared the first student to the second student.
Again case -NPs require a contextual filter reducing the correlated object to just relevant properties.
Cases and their correlated objects also display different part structures. In general, a case does not inherit its part-structure from its correlated object. Rather, its part structure is ‘reset’, that is, entirely driven by the relevant features (intrinsic or relational) obtained by filtering the correlated object through the relevant general condition:
(34) a. Part of the case of the stolen statue is familiar.
b. Part of the stolen statue is familiar.
c. Part of the theft of the stolen statue is familiar.
Cases differ from their correlated objects in other respects. Generally, it is difficult for a case to have properties of concreteness. Thus, cases generally do not have a spatial location, even if their underlying object has:
(35) a. ??? The case of the stolen statue is on the table.
b. The statue is on the table.

13
Moreover, cases generally do not act as objects of perception;
(36) a. ??? I saw / noticed the case of the broken vase.
b. I noticed the broken vase.
The condition is not strict, though. Under special circumstances, cases do act as objects of perception:
(37) This case of musical experimentation sounds horrible.
The acceptability of such examples can be traced to the perceivable quality forming part of the filtered features constituting the case.
Finally, cases are generally not causally efficacious (except, of course, as objects of mental attitudes):
(38) a. An overweight baby caused the cradle to break apart.
b. ?? The case of an overweight baby caused the cradle to break apart.
Again, particular circumstances allow for exceptions:
(38) c. This one case of cholera was the cause of a great epidemic.
Thus, we can conclude that what properties cases have depends entirely on the filtering condition. Some properties of the correlated object may pass through the filter. But other properties of cases will be reset on the basis of what features are filtered through. This in particular holds for attitudinal and evaluative properties as well as for properties based on the part structure of cases.
2.3. An account of object-related cases
We have seen that cases can be described by two different constructions: one involving reference to a universal and the other involving reference to the object correlated with the case. The two constructions together in fact reflect the way cases are individuated. We have

14 seen that an implicit general filtering condition plays a role also in the individuation of cases described by reference to the correlated object. Thus, cases in general are driven both by a general condition or universal and a correlated object. The general condition serves to filter the underlying object, reducing it to relevant intrinsic and relational features. Cases thus will be filtered entities .
6
The general condition, as one may say, serves as an ontological case filter .
7
Given the implicit presence of a filter, the lexical meaning of case can be considered a three-place relation taking the general condition and a particular as an explicit or implicit argument. The noun case actually is generally used as a functional noun, expressing a function mapping an object and a general condition onto the case that is the result of filtering the object through the general condition. Thus we have (39), where C is the contextually given filter:
(39) [ The case of the stolen statue ]
C
= [ case ](C, [ the stolen statue ])
How would this account for situation-like cases? Let us consider the clausal construction the case in which John marries Mary . The future case described by this NP will involve John and Mary and their standing in the relation of one marrying the other. This relation may itself be considered the filter. Thus, the proposal is that situation-like cases are pluralities of objects reduced to the properties and relations specified by the that -clause in question.
How would this work for quantificational clausal case -NPs such as the case in which someone discovers a solution ? We will see that the cases described by such case -NPs are not constituted by existential quantifiers, but involve particular individuals, in this case a particular person and a particular solution. This matches the account of the semantics of clausal case -NPs based on truth-making proposed later.
A generic clausal case -NP will stand for a kind whose instances are situation-like cases, which means, given the proposal, pluralities of objects filtered through the complex condition specified by the that -clause.
6 Note that filtered entities are entities reduced to some of their features are not tropes or features themselves, the tropes that have the unfiltered entities as bearers. While cases may share some of their properties with the correlated objects, as we have seen, tropes hardly every share properties with their bearers. Cases or filtered entities are not tropes, but ‘tropers’ , to use Loux’s (ms) term.
7 For the more familiar, but unrelated syntactic notion of a Case Filter see Chomsky (1981).

15
Cases described by nominal case -NPs may go along with the existence predicate exist or the existence predicate happen , unlike cases described by clausal case -NPs:
(40) a. The case of the cancer patient that Mary described exists / ??? occurred / ??? presented
itself.
b. That case of fraud happened yesterday.
The reason is that cases described by nominal case -NPs may inherit some properties from their correlated objects. These properties include the mode of existence expressed by the existence predicate that the correlated object goes along with. By contrast, cases described by clausal case -NPs cannot inherit properties, but rather set up their own mode of existence.
2.4. Event-related cases
There is another, more special kind of object-related case, namely event-related cases. A case -
NP like the case of bad weather or the case of a defeat refers to a case whose correlated object is a merely possible event. This is why such case -NPs (in in -phrases) may lead to sentences having the status of conditionals:
(41) We will cancel the event in the case of bad weather.
Again, a case of an event is not identical to the event itself. The difference is clear with object-related attitude verbs. Compare (42a) and (42b):
(42) a. Napoleon imagined a defeat.
b. Napoleon imagined the case of a defeat.
While (42a) may describe a situation in which Napoleon imagined the details of an event of defeat, the imagination described in (42b) is likely to focus on the consequences of an event of defeat, not the way it happens.
The difference between (42a) and (42b) indicates that a case of an event is an event reduced to its ‘mere occurrence’. Only some object-related attitude verbs fail to distinguish between the case of an event and the event itself, for example prepare :

16
(43) a. He is prepared for a (possible) defeat.
b. He is prepared for the case of a possible defeat.
Predicates of description also show a clear difference between cases of events and events:
(44) a. He described a possible defeat.
b. ?? He described the case of a possible defeat.
Predicates of descriptions hardly apply to cases of events, which are ‘descriptively closed’, lacking internal detail.
Cases of events, moreover, generally do not have typical event properties. They cannot be said to ‘last’, to be ‘sudden’, or to be ‘visible’:
(45) a. A snowfall might last several hours.
b. ??? The case of a snowfall might last several hours.
(46) a. A snowfall might be sudden.
b. ?? The case of a snowfall might be sudden.
(47) a. A snowfall would be visible.
b. ??? The case of a snowfall would be visible.
Cases of events and events finally differ in what prepositions they may go along with:
(48) a. during the snow
b. ?? duringa case of snow
(49) a. In the case of a defeat, we should retreat.
b. ?? In a defeat, we should retreat.
Event-related cases can form the object of object-related attitudes as long as the details of the correlated event do not matter, but they otherwise hardly share any properties with the latter. A case of an event has lost the descriptive properties and the temporal structure of the event and thus comes close to a fact. Unlike a fact, however, an event-related case generally is correlated with a merely possible event, not an actual one.

17
3. The is the case-construction
3.1. Differences between is true and is the case
A particularly interesting construction that case engages in is is the case , a syntactic predicate allowing a thatclause or a sentential pronoun as subject:
(50) a. That it is raining is not the case.
b. John feared that it might rain. That was in fact the case.
A common view among philosophers is that is the case and is true mean the same thing:
(51) That it is raining is not true.
However, there are significant semantic differences between is true and is the case :
8
The most important semantic difference concerns the behavior of is true and is the case with adverbial modifiers. First, is true and is the case differ in their acceptance of location modifiers. Location modifiers are perfectly fine with is the case , but they are often hard to make sense of with is true :
(52) a. In our firm, it is not the case that one gets fired without explanation.
b. ??? In our firm, it is not true that one gets fired without explanation.
(53) a. In John’s family, it is not the case that children respect their parents.
8 One might think that the case unlike true does not have the status of a predicate. However, standard linguistic criteria for predicatehood diagnose is the case as a predicate syntactically.
First, is the case is able to act as the predicate in small-clause constructions, a standard criterion for predicatehood:
(i) a. I consider it true that John is a genius.
b. I consider it clearly the case that John is a genius.
Moreover, like true , the case can go with other copula verbs than be , such as remain and seem :
(ii) a. That John is the best player will always remain the case.
b. The generalization remained true despite the changing circumstances.
(iii) a. That John is happy does not seem the case.
b. That John is happy does not seem true.

18
b. ?? In John’s family, it is not true that children respect their parents.
Whereas (52a) and (53a) are perfectly natural as statements of facts, (52b) and (53b), if not unacceptable, at least convey a somewhat particular metasemantic notion of location-relative truth.
Furthermore, is the case is fine with adverbs of quantification, with which is true is hardly acceptable:
(54) a. Given that she has developed Alzheimers, it will often be the case that Mary forgets
something.
b. ?? It will often be true that Mary forgets something.
(55) a. It was twice the case that someone was absent.
b. ??? It was twice true that someone was absent.
The use of adverbs of quantification with is the case shows that the subject clause may be evaluated with respect to the various situations that the adverb of quantification ranges over.
Propositional anaphora such as that in the context of again show the same thing:
(56) It was once the case that no parent was at the school meeting. Today that is the case
again.
By contrast, the thatclause with is true needs to be propositionally complete. That S in that S is true is understood as complete regarding context-dependent elements, such as quantifier restrictions, tense interpretation, spatial location etc., though of course the proposition expressed may involve ‘unarticulated constituents’.
A further difference between is true and is the case shows up with adverbs that may act as degree quantifiers such as German kaum ‘hardly. With is the case , such adverbs can act only as adverbs of quantification, whereas with is true they most naturally act as degree modifiers:
(57) a. Es ist kaum der Fall, dass Hans Kaffee trinkt.
‘It is hardly the case that John drinks coffee.’
b. ? Es ist kaum wahr, dass Hans Kaffee trinkt.
‘It is hardly true that John drinks coffee.’

19
Whereas (57a) means that there are only rare cases of John drinking coffee, (57b) claims that it can hardly be said that John drinks coffee.
3.2. A semantic account of is the case based on truth-making
The semantic behavior of is the case with respect to adverbial modifiers supports an analysis based on truth-making.
9
Truthmaking is not an uncontroversial philosophical notion, and I will address some of the controversies it has given rise to in the next section. For the time being, it should suffice to clarify that truthmaking is a relation that holds between an entity and a true sentence just in case the sentence is true in virtue of that entity. The entity in virtue of which the sentence is true is the truthmaker of the sentence. Instead of holding between an entity and a sentence, truth-making may also hold between an entity and a proposition.
10
Making use of truth-making, that S is sometimes the case will involve existential quantification over cases (with sometimes ) and claims that some cases make S true. The semantics of is the case with a location modifier and an adverb of quantification will thus be as follows, where s
╞
is the truth-making relation:
(58) a. For a prepositional phrase PP, PP it is the case that S is true iff for the maximal actual
‘case’ s such that PP(s), s
╞
S.
b. It is sometimes the case that S is true iff for some ‘cases’ s, s
╞
S.
I will assume standard conditions on truth making of conjunctions, disjunctions, and existential quantification as follows:
11
9 For the notion of truth-making see, for example, Mulligan/Simons/Smith (1984), Armstrong (2004), and
Rodriguez-Pereyra (2006).
10 A notion related to the notion of truth-making when applied to situation-related ‘cases’ is the relation of exemplification in Kratzer (online), a relation that holds between situations and propositions.
11 The condition for sentences with universal quantification and conditionals are less obvious and in fact controversial and I will not specify truthmaking conditions for them here. See Armstrong (2004) for a discussion and a proposal.

20
(59) a. s
╞
S and S’ iff for cases s’ and s’’, s = sum(s, s’) and s’
╞
S and s’’
╞
S’.
b. s
╞
S v S’ iff s
╞
S or s
╞
S’
c. s
╞  xS iff s
╞
S[x/d] for some individual d.
The relation between a ‘case’ and the that -clause that is involved in the semantics of the is the case -construction needs to be the relation of exact truth-making.
12
That is, it must be the relation that holds between a case and a that -clause only if the case is wholly relevant to the truth of the that -clause. This is clear from the way adverbs of quantification are understood:
(60) a. It was twice the case that John made a mistake.
b. It was only once the case that John lost the game.
Twice in (60) counts situations that are completely relevant for the truth of John made a mistake , that is, situations that include nothing more than John, a single mistake, and the
‘making’-relation holding between the two.
Twice does not count larger situations or sums of situations. Similarly, once in (60) counts situations of a single event of losing only.
If adverbs of quantification with is the case count truthmaking cases, then those should be part of the world and not nonworldly facts (that is, facts that correspond to the truth of a proposition). The latter could be quantificational and disjunctive, the former cannot. The way indefinites and disjunctions contribute to the ‘cases’ that adverbs of quantification count makes this very clear:
(61) a. It was only once the case that someone failed the exam.
b. It was three times the case that John or Mary received a gift.
Once in (61a) counts cases involving exactly one individual failing the exam; it does not count a single fact, the fact that someone failed the exam. Three times in (61b) counts cases involving either John or Mary; it does not count a single disjunctive fact.
The implicit reference to cases on this analysis of is the case is not achieved, as one might
12 This is the truth-making relation that is used in Rodriguez-Pereyra (2005), Moltmann (2007) and Fine (to appear) .

21 have thought, by the expression the case itself. The case in that context does not have the status of a truly referential NP. Several diagnostics show that. First, the case in it is the case requires the simple definite determiner:
(62) a. * It is not that case that S,
b. * It is not a case that S.
Moreover, the case in it is the case does not permit adjectival or relative-clause modifiers:
(63) a. * It is the unfortunate case that S.
b. * That S is not the case that we expected.
The case in is the case, moreover, cannot act as the antecedent of an anaphoric case -NP:
(64) That no one came was recently the case. ?? We did not like that case.
The case in is the case rather appears to be a mere ‘referential residue’.
13
Is the case in fact does not involve reference to a particular case (as one might expect from the presence of the case ), but quantification over cases. This is obvious from adverbs of quantification such as something , but also the interpretation of negated is the case -sentences.
Thus, it is not the case that S states that there is no ‘case’ that supports S. Without an adverb
13 There is a potential alternative analysis of the is the case -construction that one might think of, namely as a specificational sentence (Higgins 1979), as a sentence of the same sort as those below:
(i) a. That John is innocent is the truth.
b. That we would all go is the idea.
c. That one can walk home is the advantage.
d. That John is incompetent is the problem.
But there are major differences. First, is the case does not permit extraposition, unlike specificational sentences:
(ii) a. * It is the truth that John is innocent.
b. * It is the idea that we would all go.
Moreover, is the case does not permit inversion, unlike specificational sentences:
(iii) a. The truth is that S
b. The idea is that S.
c. * The case is that S.
Thus, an alternative analysis as a specificational sentence is not an option.

22 of quantification or negation, is the case is not very good, unless it occurs in the antecedent of a conditional:
(65) a. ??? It is the case that John is married.
b. If it is the case that John is married, then Mary should stop seeing him.
c. If it is the case that someone fails the exam, then he should be given the chance to
repeat it
(65c) makes clear that it is the case that S involves existential quantification over cases that make S true.
The truth-making relation has applications to the semantics of other case -constructions as well. First, it helps to account for anaphoric reference to cases introduced by modals of possibility, as in (22b), repeated below:
(66) John might go to the party. In that case, I will go too.
The truthmaker of the first sentence will be a merely possible case, which will then act as the semantic value of that case .
Sentences with modals of necessity do not support plural anaphora, as we have seen in
(23a, b), repeated below:
(67) a. ??? John must be at home. In those cases, Mary will be at home too.
b. ??? John must write the report. In those cases he will have done his duty.
Given the standard account of modals as universally quantifying over circumstances, this is surprising. In general, universal quantifiers support plural anaphora:
(68) Every woman disliked the cake. They preferred the cookies.
Obviously, anaphoric case -NPs do not refer to the circumstances that the modal quantifies over on the standard account of modals. This suggests an account of modals on which modals

23 do not act as quantifiers over circumstances but have the status of primitive operators.
14
This means that a true sentence with a modal of possibility provides a referent for an anaphoric that case because it has a possible situation-like case as truth maker, not because it expresses existential quantification over circumstances. By contrast, a true sentence with a modal of necessity, whatever its truthmaker may be, will not have a single situation-like case as a truthmaker suited for anaphoric reference with that case .
Truthmaking is also suited for the analysis of conditional and quasi-conditional case constructions, such as (18a, b), as well as event-related case -constructions that are of a conditional sort, such as (41). Ordinary conditionals in fact have been analysed on the basis of truth-making (Fine, to appear), and it is expected that such an analysis is suited for conditional case -constructions as well.
15
For conditional case -constructions, truth-making may come into play in roughly the following way. The truth of a conditional in case S, S’ requires that for any case making S true, there will be a larger case part of which makes S’ true.
16
The truth of a quasi-conditional in case S, S’ roughly requires that there be a case making S’ true and that the case making S’ true be part of an extension of a case making S true.
17
Truth-making should also be involved in the semantics clausal case -NPs. Clausal case -NPs will refer to cases making the which -clause true, which on the present view would amount to the cases being filtered pluralities of objects. Note that clausal case -NPs with a modal will not refer to a unique case, but a plurality of cases, as below:
(69) a. [ the case in which might S] w = the cases ss such for any s, s < ss and s R w, s
╞
S.
Here R is the relevant relation of accessibility and < the relation that holds between an object and a plurality of objects of which the former is part.
14 For such a modalist view of modality see, for example, Forbes (1989).
15 I will not go into the details of Fine’s analysis of conditionals, but adopt a much simplified analysis.
See also Kratzer (online) and references therein for analyses of conditionals on the basis of situations.
16 For a more developed analysis of if -conditionals based on truth-making states see Fine (to appear). Such analysis may apply to in -phrases with clausal case-NPs as well.
17 There are differences between ordinary if -conditionals and case -conditionals. If-conditionals can go along with adverbs of quantification, of which the ifclause appears to act as a restriction, but case -conditionals cannot:
(i) a. If a student fails the exam, he usually tries again.
b. ??? In case a student fails the exam, he usually tries again.

24
Truth-making may even be used to account for the semantics of case -NPs describing event-related cases such as the case of a defeat . The analysis would have to invoke an implicit predicate might occur , to ensure reference to something like the possible occurrence of an event:
(69) b. For an event noun N,
[ the case of an N] = the cases ss such that for any s < ss, s
╞ an N might occur
Truth-making can also be used to specify the semantic values of clausal caseNPs referring to kinds of cases. Kinds of cases may be viewed as pluralities of possible and actual cases so that we have the following very simple analysis:
18
(69) c. [ the case in which S] = the ss such that for any s < ss, s
╞
S.
Truth-making may also be applied to the semantics of nominal case -NPs describing objectrelated cases. This might be done as below, where C is the contextually given filtering condition, and what is made true is now a structured proposition consisting of that condition and the object in question:
19
(69) d. [ the case of X]
C
=
 s[s
╞
<C, [X]>], for a referential NP X.
It is the case that S appears to involve as its interpretation the sort of semantics that Austin
(1950) proposed for independent sentences in general. On Austin’s view, with the utterance of
18 Note that for this interpretation to work, the which -clause should not have a tensed interpretation.
19 A somewhat different treatment is required when case occurs predicate-initially, as below:
(i) John’s illness is a case of cancer.
Does this mean that John’s illness is identical to a case, even though it would not necessarily be a filtered object?
It is plausible that (i) is not an ordinary subject-predicate sentence, but rather is on a par with (ii):
(ii) True is a truth value.
The predicate in (i) arguably does not just attribute a property to the semantic value of the subject, but involves its reification as an object that is a truth value (Moltmann 2013, Chapt. 6). Similarly, the predicate in (i) would involve filtering of the subject referent as a case.

25 a sentence, a speaker refers to an (actual) situation and claims that the situation referred to is of the type specified by the sentence uttered. That is, the situation referred to acts like a truth maker of that sentence. On the present view, this is only part of the constructional meaning of is the case . With is the case
, adverbs of quantification range over ‘cases’ and location adverbials act as predicates of cases. Austin’s motivations for implicit situation reference were in fact quite different from the present ones. The situation referred to, for Austin, is responsible for contextual restrictions on quantification domains, the interpretation of tense etc. The present motivation for invoking truth-making is quite simply the semantics of the is the caseconstruction.
3.3. Cases as truthmakers
The truth-making idea as a general view about truth says that if a sentence is true, it is true in virtue of something in the world that makes it true. The truth-making idea is a highly controversial philosophical view, though.
20
The question thus is, given that case -constructions involve the truth-making relation how does this bear on the philosophical issues surrounding truth-making? In particular, what general view of truth-making is reflected in the semantics of the case -construction given that the construction does involve truth-making?
What is controversial about the truth-making idea is first of all the view that grounding requires an entity to act as a truth maker. Some philosophers such as Lewis (2001) and
Hornby (2005) agree that the truth of sentences should be grounded, but disagree with the view that they need to be grounded in entities acting as truth makers; rather the truth of sentences should be grounded in how thing are. The grounding of truth on that view does not require what is considered reification of entities as truth makers.
If truth-making just plays a role in the semantics of caseconstructions, then the use of truth-making for the semantics of that construction appears compatible with the view that truth is grounded in how things are and does not require entities as truthmakers. Given that view, case could be considered a ‘nominalizing’ or rather reifying expression , mapping the way things are to support the truth of a sentence onto the set of objects that would act as the truthmaker of the sentence. Formally, case can then be considered an expression denoting a
20 For an overview of the truthmaking debate see Rodriguez-Pereyra (2006) and the contributions in
Beebee/Dodd (2005).

26 function mapping a world w and a proposition onto the set of objects that are reifications of whatever it is in w that makes S true.
(70) [ case in which S] w
= {s : [ case ](w, [S]) (s)}
This account, of course, makes sense only if truth-making is not taken to be involved in the semantics of natural language elsewhere, independently of nominalizing expressions.
21
A subject of great controversy regarding truth making is the question whether the truth of all sentences requires a truthmaker. A number of philosophers thus shy away from truthmaker maximalism, as the view is called, but take truth-making to be a requirement for the truth of only certain types of sentences. Negative sentences pose particular difficulties for truthmaking. What kind of entity is there in the world that could make the sentence John failed to show up true or the sentence no one is satisfied ? On some views of truth-making, negative sentences do have truthmakers; on others, they don’t. The notion of truth-making that I argued is involving in the semantics of clausal case -NPs clearly is one that requires truthmakers for negative sentences. Negative sentences pose no difficulty for the referentiality of clausal case -
NPs:
(71) a. the case in which John fails to show up
b. the case in which noone is satisfied
Another kind of sentence considered problematic is sentences expressing the predication of essential properties. However, with an epistemic modal, such sentences are perfectly suited for forming case -NPs:
(72) a. We should not exclude the case in which 388767 might be a prime number.
b. We took into consideration the case in which Sasha might be a cat.
There are some types of sentences that cannot be used to form referential case -NPs, though. In particular, conditionals are excluded, both with a modal and when they are of a generic sort:
21 Fine (seminar at NYU) pursues the view that truth-making plays an important role in semantics in general.

27
(73) a. ??? The case in which we might stay home if it rains.
b. ??? The case in which a stone falls down if it is dropped.
A third controversy surrounding the concept of truth-making concerns the nature of truthmakers. Some philosophers that pursue the truth-making idea make no assumptions what exactly truth makers are, whether or not they are fully individuated entities (see for example
Rodriguez-Pereyra 2005). Fine (to appear) takes truth makers to be ‘states’, but makes no further assumptions about what states are except that they are the entities that play the truthmaking role and stand in certain relations to each other (in particular the relation ‘part of’).
Other philosophers, such as Mulligan / Simons / Smith (1984) and Lowe (2006), take truthmakers to be fully individuated entities that play an independent role in the world.
22
In particular, events and tropes on that view act as truth makers, entities that play independent rolesas objects of perception and relata of causal relations.
23
Different types of entities thus are proposed as truthmakers for different types of sentences:
(74) a. John is happy truthmaker: John’s happiness (a trope)
b. John is walking. truthmaker: John’s walk (an event)
c. John exists. truthmaker: John (an individual)
d. John is a man. truthmaker: John or a trope (‘John’s humanity)
There are difficulties for this view, however. First, fully individuated entities do not seem to fulfill the condition of exact truthmaking. There are always features about John’s particular walk that may not be relevant for the truth of John walked , for example, the location and time of the walk and the way John walked. Moreover, there are many features of John that do not matter for the truth of John is a man or John exists.
Truthmakers suited for exact truthmaking, it appears, need to be thinner than fully individuated objects.
22 Armstrong (2004) takes truth makers to be states of affairs, which for him also act as causal relata.
23 This is also the view adopted in Moltmann (2004), which applies truth-making to the semantics of event- and trope-nominalizations.

28
The view also faces the difficulty of not being compatible with presentism, the view that only objects at the present moment exist (Sider 2001, Merricks 2007). Given presentism, most true sentences will fail to have a truth maker or will at some point lose their truthmaker.
Truthmakers as ‘cases’, it seems, avoid the two difficulties. As filtered objects or pluralities of objects, they are suited for exact truth-making. Moreover, there is reason to consider cases not as entities that exist in time, but time-independently – and thus as existing, derivatively at any time. The use of tense in natural language is indicative of that. Existential quantification over cases whose correlated objects are events or tropes in the past is possible with sentences in the present tense, whereas existential quantification over past events requires past tense. Talking about events in the past, (75a) and (76a) are perfectly fine (that is, possibly true); but (75b) and (75b) are not, as opposed to (76c) and (76c):
(75) a. There are three known cases of this disease.
b. ??? There are only three outbreaks of this disease.
c. There were only three outbreaks of the disease.
(76) a. There are only three known cases in which someone crossed the border.
b. ??? There are only three crossings of the border.
c. There were only three crossings of the border.
This suggests that cases should be considered objects reduced to certain tensed features. As such their existence will not be restricted to a particular time; rather they will be present at any time. We have already seen that object-related cases generally do not have a spatial location and it is not expected that they should have a temporal duration, unless a temporal duration is part of their constitutive features.
There is another type of entity that shows the same time-independence as event-related cases and makes a good candidate for being a filtered object. These are entities constituted by the lasting legacy of an actual person, such as philosophical figures. Below, we see that present tense can be used to quantify over philosophical figures that, as people, no longer exist, which is not possible with people viewed simply as people
(77) a. There are three famous philosophers that had studied in Tuebingen, Hegel, Fichte and
Schelling.
b. ??? There are three people that had studied in Tuebingen and became famous

29
philosophers, Hegel, Fichte, and Schelling.
c. There were three people that studied in Tuebingen and became famous philosophers,
Hegel, Fichte, and Schelling.
It is quite plausible that philosophical figures are filtered objects, persons reduced to their philosophical views and achievements. As such, they share their time-independent existence
(once they have come into existence in the first place) with cases.
4. Conclusion
In this paper, I have argued that c ase -constructions in English (and some other European languages) are revealing as to the notion of truthmaking and the nature of truthmakers. This raises the question how general such constructions are across languages. As a matter of fact, not all languages have case -constructions, not even all European languages. Chinese lacks them, as does Swedish, to mention just two. This does not undermine the view, though, that cases are entities of a particular sort, act as truthmakers, and play particular semantic roles.
Cases may to an extent play the very same roles in the semantics of a language that does not have expressions to describe them explicitly, for example in the semantics of conditionals.
Moreover, cases may play the role of truthmakers regardless of whether or not a language has expressions to refer to them or to quantify over them. And finally, cases may be what was proposed that they are (filtered object) independently of any language.
Acknowledgments
This paper has greatly benefited from several workshops on case- and fact descriptions held in in Paris in the context of the project ANR-DFG Nominalizations: Linguistic and
Philosophical Aspects in 2011. The financial support from that project is gratefully acknowledged. Special thanks go to Danièle van der Velde, Katya Paykin-Arrouès, Faissal
Talayati, and Lucia Tovena. I would also like to thank the audiences of the conferences
Truthmakers and Proof Object s organized by Per Martin-Loef (Paris September 2011) and the audiences at talks in Hong Kong, Goettingen, and Chicago, as well Kit Fine’s seminar on truth makers at NYU in the spring of 2012.
References

30
Armstrong, D. (1997): A World of States of Affairs . Cambridge: Cambridge UP.
----------------- (2004): Truth and Truthmakers . Cambridge UP, Cambride.
Austin, J. L. (1950): ‘Truth’.
Aristotelian Society Supplement Vol.
24, 111-129. Reprinted in
Austin (1961).
------------- (1961a): Philosophical Papers.
Ed. By J. O. Urmson / G. J. Warnock,
Clarendon Press, Oxford.
-------------- (1961b): ‘Unfair to Facts’. In Austin (1961a), 102-222.
Beebee, H. / J. Dodd (eds.) (2005): Truthmakers: The Contemporary Debate . Oxford UP,
Oxford.
Carlson, G. (1977): ‘A Unified Analysis of the English Bare Plural’.
Linguistics and
Philosophy 1, 413-457.
Chomsky, N. (1981): Lectures on Government and Binding . Foris, Dordrecht.
Fine, K. (to appear): ‘Counterfactuals without Possible Worlds’.
Journal of Philosophy .
Forbes, G. (1989): Languages of Possibilities . Blackwell, Oxford.
Higgins, R. (1979): The Pseudo-Cleft Construction in English . Indiana University
Linguistics Club.
Hornsby, J. (2005): ‘Truthmaking without Truthmaking Entities’. In H. Beebee / J. Dodd
(eds.): Truthmakers. The Contemporary Debate . Clarendon Press, Oxford, 34-48.
Kratzer, A. (online): ‘Situations in Natural Language Semantics’. Stanford Encyclopedia of
Philosophy.
Lewis, D. (1975): ‘Adverbs of Quantification’. In E. Keenan (ed.): Formal Semantics of
Natural Language . Cambridge UP, Cambride.
------------ (2001): ‘Truthmaking and Difference-Making’. Nous 35.4., 602-6014.
Loux, M. (ms): ‘Constituent Ontology’. Ms University of Notre Dame.
Lowe, J. (2006): The Four-Category Ontology . Oxford UP, Oxford.
Merricks, T. (2007): Truth and Ontology . Clarendon Press, Oxford.
Moltmann, F. (2007):
‘Events, Tropes and Truthmaking’. Philosophical Studies 134,
363-403.
----------------- (2011): ‘On the Semantics of Existence Predicates’. I. Reich et al. (ed.):
Proceedings of Sinn und Bedeutung 15 (2010), Saarland University Press,
Saarbruecken,
31-54.
---------------- (2013): Abstract Objects and the Semantics of Natural Language . Oxford UP,
Oxford.

31
Mulligan, K. / P. Simons / B. Smith (1984): ‘Truthmakers’.
Philosophy and
Phenomenological Research 44, 287-321.
Rodriguez-Pereyra, G. (2005): ‘Why Truthmakers?’. In H. Beebee / J. Dodd (eds):
Truthmakers: the contemporary debate , Oxford University Press, Oxford, 17-31.
-------------------------- (2006) Truthmakers. Philosophical Compass 1, 186-200.
Sider, T. (2001): Four-Dimensionalism . Oxford UP, Oxford.
Strawson, P. F. (1949) : ‘Truth’. Analysis 9.6.
Van de Velde, D. (ms): ‘
Cas et Fait
’. Ms University of Lille.
Woltersdorff, N. (1970): On Universals . Chicago UP, Chicago.