On Weak Truthmaking Nicola GUARINO a,1 , Daniele PORELLO a , Giancarlo GUIZZARDI b a Laboratory for Applied Ontology, ISTC-CNR b Free University of Bolzano-Bozen, Italy Abstract. Informally speaking, a truthmaker is something in the world in virtue of which the sentences of a language can be made true. This fundamental philo- sophical notion plays a central role in applied ontology. In particular, a recent non- orthodox formulation of this notion proposed by the philosopher Josh Parsons, which we labelled weak truthamking, has been shown to be extremely useful in ad- dressing a number of classical problems in the area of Conceptual Modeling. In this paper, after revisiting the classical notion of truthmaking, we conduct an in depth analysis of Parsons’ account of weak truthmaking. By doing that, we expose some difficulties in his original formulation. As the main contribution of this paper, we propose solutions to address these issues which are then integrated in a new precise interpretation of truthmaking that is harmonizable with commonsense as well as with existing four-category foundational ontologies. Keywords. Strong and Weak Truthmaking, Ontological Analysis 1. Introduction Suppose the sentences ‘John is in love with Mary’, or ‘this Rose is Red’ are true in a given situation. If this is the case, it is because they accurately represent something about the world. In other words, if a sentence is true, it is because there is something real (e.g., the existence of certain things bearing certain properties) that makes them true. The philosophical literature terms a truthmaker of a sentence this entity in virtue of which the sentence is true, i.e., truthmaking is a relation that binds the sentences of a language with pieces of reality. The literature about truthmakers is very rich [1, 2, 3], and this notion plays a fun- damental role in ontological analysis. Indeed, we may regard ontological analysis as the activity that, at its core, consists of searching for the truthmakers of the sentences that concern a certain domain of interest [4]. In the latest years, based on the analysis of truthmakers of relational sentences, some of us have revisited a number of problems in the representation of relations [5], proposed a revised ontology of relation types [6], and isolated a number of modeling patterns for their proper representation [7]. In such work, we relied on a non-orthodox notion of truthmaking proposed by the philosopher Josh Parsons [8, 9], which we termed weak truthmaking. 1 Corresponding Author: Nicola Guarino, ISTC-CNR Via Alla Cascata 56C, 38123 Trento, Italy; E-mail: nicola.guarino@cnr.it. Copyright c 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). In this paper, we revisit the notion of truthmaking and conduct a in depth analysis of weak truthmaking. We expose some difficulties in Parsons’ original formulation of the latter and propose some solutions to address them. In particular, we advance a new precise interpretation of weak truthmaking that is harmonizable with foundational on- tologies that countenance the existence of individual qualities, such as DOLCE [10] and UFO [11]. The remainder of this paper is organised as follows: in Section 2, we elaborate on the standard philosophical notion of truthmaking. In Section 3, we revisit Parsons’ non- standard albeit insightful idea of weak truthmaking. We expose some of its difficulties, and propose our new interpretation of this notion. In section 4, we present some final considerations to conclude the paper. 2. Truthmaking As previously mentioned, the truthmaker of a sentence is an entity in virtue of which the sentence is true. The truthmakers debate is specifically about providing a meaning- ful grasping of the expression in virtue of. According to the mainstream theories, the truthmaker of a sentence is something the existence of which at a world w is necessary, sufficient, or both, for the truth of that sentence at w. For the purpose of this paper, we shall only present the definition of the entailment principle (T) of truthmaking: [12, 2]: (T) x is a truthmaker of φ at w iff the existence of x at w entails the truth of φ at w. Note that we phrased the principle (T) in modal terms and we are defining the rela- tion of truthmaking with respect to a world w. (T) is traditionally assumed in combina- tion with the following principle (E), that is, the relation of truthmaking must be invariant across worlds (truthmaker essentialism). (E) expresses the assumption that it is because of the very existence of the truthmaker, that the sentence holds. (E) if x is a truthmaker of φ at w, then for every w0 , x is a truthmaker of φ at w0 . A number of points are worth noticing. First, this definition of truthmaking has been severely criticised on the ground of its logical consequences, if a few quite natural prin- ciples are assumed [13]. One of such principles is the so-called truthmaker maximalism, according to which every true sentence must have a truthmaker. If this principle is as- sumed, then arbitrarily complex sentences need to have their own truthmakers, and, if not carefully constrained, principle (T) leads to a devastating conclusion from the onto- logical point of view, namely that every true sentence has the same set of truthmakers (truthmaker monism). In this paper, we cannot face this delicate issue and we restrict ourselves to discussing truthmakers of atomic predicative sentences, without claiming for truthmaker maximalism. Perhaps, as Parsons suggests, “the best we can demand is that every contingent truth either has a truthmaker, or is a conjunction, disjunction, nega- tion, or some other more complicated Boolean function of propositions that would have truthmakers were they true.” [9, p. 169]. A second problem is that principle (T) overgenerates, in the sense that it allows for truthmakers that intuitively should not be considered as such [1]. For example, it fol- lows from (T) that if t is a truthmaker of φ and φ is true at w if and only if ψ is true in w, then t is a truthmaker of ψ as well. Thus, two logically equivalent sentences have the same truthmakers. This consequence may be convincing, only if we assume a suffi- ciently large number of possible worlds that is capable of separating the meaning of the required propositions. However, this move only implicitly distinguishes between the rel- evant truthmakers of the sentences at issue. We prefer to explicitly assume a further con- dition on truthmaking, which is traditionally labelled the relevance assumption: “what makes something true must— in some sense— be what it is ‘about’ ”([1]). (R) x is a truthmaker of φ at w only if t is relevant to φ at w As it is well known, it is very difficult to capture relevance formally. In Section 4 we shall discuss a simple strategy that allows us to take relevance somehow into account. Note finally that, since principle (T) is formulated in terms of the existence of a certain entity, the truthmaking relation is intertwined with our ontological assumptions, since only the things endorsed by our ontology can be considered as truthmakers. For instance, if we say that a certain mass quality is the truthmaker of the sentence ‘a has mass’, then of course we commit to the existence of qualities. 3. Weak truthmaking In 1999, Josh Parsons proposed an interesting non-orthodox approach to truthmaking [8, 9] that denies the so-called truthmaker essentialism, i.e., the idea (implicit in principle (T)) that a truthmaker is essentially such that it makes a certain proposition true. We labelled this approach weak truthmaking2 [4], in contrast with strong truthmaking, which relies on principle (T). Parsons describes it as follows, proposing two criteria that for him are equivalent3 : (i) for every true sentence, there is some thing such that the sentence cannot become false without a qualitative change, a non-Cambridge change, in that thing. That thing, whatever it is, is the sentence’s truthmaker. (ii) [Or,] the truthmaker for a sentence is that thing that is intrinsically such that the sentence is true. [8, 328] (sentence numbers added) So, according to Parsons’ intuitions, it is not the mere existence of the truthmaker that entails the truth of the sentence, it is rather the way the truthmaker intrinsically is. In his view, truthmaking depends not so much on whether something exists, but on how something is in itself (either contingently or necessarily), independently on anything else. In this sense, weak truthmaking appears to be more informative than strong truthmaking, as it manifests the intrinsic way an entity grounds the truth of a proposition. Note also that, since a weak truthmaker may be contingently such that the sentence is true, the truthmaker essentialism (E) is rejected. 2 Friedrike Moltmann speaks of strong and weak truthmaking in a different sense, to distinguish a truthmaker theory that admits truthmaker maximalism from one that does not [14, p. 90] 3 Note that the two criteria below, which we have numbered in order to discuss them, are actually presented contiguous to each other in [8], while only the second criterion is mentioned in [9]. 3.1. Being intrinsically such that Let us now try to understand Parsons’s proposal more in detail. Criterion (ii) is for him the central one [9], which may be formulated as follows4 : (WT) t is a weak truthmaker for proposition that φ at w iff t is intrinsically such that φ at w. A crucial point in this definition is the notion of being intrinsically such that φ . Parson adopts Lewis’ definition of intrinsicality in terms of duplicates [9]. That is, for him a property P holds intrinsically for x (at w) if and only if P holds for every duplicate of x (at w). Being a duplicate of x and having the same intrinsic properties of x are therefore inter-definable notions, thus it is a matter of choice whether one prefers to start with a definition of duplicates or to start by defining intrinsic properties. An alternative (and perhaps more informative) way of defining intrinsicality is in terms of loneliness (being the only contingent object in the world), or accompaniment (presupposing the existence of other contingent beings): a property P holds intrinsically for x iff “having or lacking the property is independent of accompaniment or loniless” [16]. In the light of this, the tentative definition that we shall adopt is as follows: (INT) A property holds intrinsically for x at w iff it holds at w independently of the presence or the absence at w of anything5 wholly distinct from x, where x is wholly distinct from y means that x is not identical to y nor to any of its proper parts. Consider now properties of the form being such that φ . These are known as ’vacu- ous’ (a.k.a. ’propositional’) properties since they do not convey much information about their argument: x is such that φ just in case x exists and φ holds. As Lewis noted, such properties may hold intrinsically for some entities and extrinsically for others6 . For in- stance (Parsons’ example), being such that Socrates is snub-nosed holds intrinsically for Socrates and not for Plato, since in the latter case it requires the presence of something else (Socrates) wholly distinct from Plato. Therefore, according to (WT), the (weak) truthmaker of ‘Socrates is snub-nosed’ is just Socrates, and not Plato, because only the former is intrinsically such that the proposition holds. Let us now go back to the (alleged) alternative way of expressing the same intu- itions, conveyed by criterion (i) in the above quotation. What are the things such that the sentence ’Socrates is snub-nosed’ “cannot become false without a qualitative change, a non-Cambridge change”, in that things? For sure, Socrates itself is one of such things. But what about his nose? Of course, a suitable qualitative change in that thing, too, 4 We report here Rami’s re-formulation of Parson’s definition [15, p. 23]. As Rami explains, “Parsons himself gives the following formulation: x makes p true iff x is intrinsically such that p. But this formulation is incorrect, because ’p’ must be regarded as a nominal variable relative to its first occurrence and as a sentential variable relative to its second occurrence”. Here and in the following, we shall use φ as a sentential variable, and p as a nominal variable standing for the proposition that φ . So, ‘intrinsically such that φ ’ is equivalent to ‘intrinsically such that p is true’, and ‘φ holds’ is equivalent to ‘p is true’. 5 The cited definition refers to the presence or the absence of any contingent entity wholly distinct from x, but for our purposes we prefer to adopt a stricter notion of intrinsicality, allowing us to consider John’s property of liking the number 7 as holding extrinsically just because John and the number 7 are wholly distinct from each other. Philosophers are still debating these definitions. 6 Such properties are called therefore conditionally intrinsic. A property is (unconditionally) intrinsic iff it holds intrinsically for all its instances. would falsify the sentence. So, Socrates’ nose seems to be a perfect weak truthmaker of ’Socrates is snub-nosed’ according to (i). Indeed, it would be the minimal weak truth- maker, assuming that a truthmaker is non-minimal iff another truthmaker of the same sentence is somehow internal7 to it. Such minimal truthmaker would definitely carry an important role in explaining why the proposition holds, grounding therefore its meaning: Socrates is snub-nosed because of the way his nose is (and not, say, his legs). Unfor- tunately, however, Socrates’ nose does not satisfy criterion (ii), since ‘Socrates’nose is such that Socrates is snub-nosed’ implies the existence of something (Socrates) wholly distinct from Socrates’ nose8 , so the nose is not a weak truthmaker of our proposition. Similarly, assume that a certain rose, Rosie, has an individual color quality q. As in the DOLCE or UFO ontologies, that means that q is located, in a certain region of the space of colors, for example, in the region of ‘red’. Therefore, q has the property of being a Red Color. According to criterion (ii), such quality would not count as a (minimal) weak truthmaker of ‘Rosie is red’ (contra criterion (i)), because it is not intrinsically such that the proposition is true. The reason is that being such that Rosie is Red would imply the existence of Rosie, which is wholly distinct from q9 . So, there is a problem for Parsons: criterion (ii), which he believes as equivalent to (and more basic than) (i), turns out to be too restrictive in the kinds of truthmakers it al- lows for, excluding parts and individual qualities from being (minimal) weak truthmak- ers and limiting therefore the explanation power of the weak truthmaking approach. We shall call this the restrictiveness problem. 3.2. Addressing the restrictiveness problem We think there is a solution to the restrictiveness problem, which relies on understanding the scope of the ’intrinsically’ modifier in the expression “x is intrinsically such that φ ”. There are indeed two scopes, which recall the de dicto vs. de re reading of modalities: 1. intrinsically x is such that φ at w (de dicto): The property of being such that φ holds intrinsically for x at w. This means that all duplicates of x are such that φ holds, i.e., φ holds in each world w where there is a duplicate of x. This is Parson’s interpretation. 2. x is intrinsically such that φ at w (de re): φ holds in virtue of the way x intrinsi- cally is at w (i.e., in virtue of a property that holds intrinsically for x at w). Let x be Rosie’s color, which is actually red. According to (1), it is not the case that intrinsically x is such that Rosie is red, since x may have a duplicate in a world where Rosie is not red. According to (2), x is intrinsically such that Rosie is red, since Rosie is red in virtue of the property being a red color that holds intrinsically for x10 . So, reading (2) of intrinsically such that would allow us to reconcile the two criteria (i) and (ii). 7 See Section 3.3 for an account of internal. 8 Keep in mind that wholly distinct from is not symmetric. In terms of duplicates, we may arrive to the same conclusion by observing that an exact copy of Socrates’ nose (attached to a different face or even isolated) may exist in a world where Socrates is not snub-nosed. 9 In terms of duplicates, we may say that being such that Rosie is Red does not hold intrinsically for q since an exactly resembling quality (a duplicate) may exist in a world where the rose is not red. 10 The fact that a certain quality has intrinsically its own value seems self-evident to us, but requires some care in understanding the tentative definition of intrinsicality we have suggested, i.e., (INT). One may observe that, since q is existentially dependent on Rosie, RedColor(q) implies the existence of something, Rosie, which is wholly distinct from q. Therefore, RedColor should not be considered as an intrinsic property of q. To avoid In the light of the above discussion, we may attempt to re-formulate weak truthmak- ing as follows: (WT1a) t is a weak truthmaker for the proposition that φ at w if and only if t is intrinsically such that φ at w. Or, equivalently, using just nominal variables11 : (WT1b) For every object x and every proposition p, x is a truthmaker of p at w iff there exists a property P such that P holds intrinsically for x at w iff p is true at w. Note that, although the notion of holding intrinsically is defined in terms of possible worlds, no quantification on worlds appears explicitly in the above definition, allowing something to be a weak truthmaker of p just in the actual world. For instance, consider the sentence ‘Socratie is snub-nosed’, where Socratie is a particular statue of Socrates. Suppose that Socratie”s nose (Nose1 ) is actually snub, satisfying therefore the above definition in the actual world w0 . Consider now the world w1 , where Nose1 has been replaced by a duplicate (Nose2 ). In such world, the truthmaker would be Nose2 , and not Nose1 . In conclusion, the above refined definition not only goes towards a reconciliation of interpretations (i) and (ii), capturing Parsons’ powerful intuition of truthmakers that depend on how something is and not on whether something exists, but it also allows us to fully exploit such intuition, including parts and qualities (and, in general, modes) of things as weak truthmakers accounting for the properties of such things: Socrates’ nose accounts for the fact that he is snub-nosed, the rose’s color accounts for the fact that it is red, and the emotional attitude John has for Mary accounts for the fact that he is in love with her12 . 3.3. The relevance problem There is however a further, subtle problem in Parsons’ account that also affects our re- fined definition. According to Parsons, the truthmaking relation is an internal relation in Russell’s sense, i.e., it is derivable from the intrinsic properties of its relata. This has as a consequence that, if t is a truthmaker of φ in the world w, everything that is duplicate of t in w is also truthmaker of φ . Parsons seems to be happy with this: “According to the truthmaker principle, as I”ve stated it, the truth of a sentence su- pervenes on the qualitative nature of its truthmaker. That nature is precisely what duplicates duplicate, so a duplicate of a truthmaker will do as well for truthmaking purposes as the original.” [8, p. 8] this, we may amend (INT) to clarify that an intrinsic property should be definitionally independent on the presence or absence of anything wholly distinct from its argument. This discussion is delicate, and we are not pursuing it further here. 11 Note that this version does not suffer of the problems discussed by Rami [15, p. 23-24] concerning the impossibility of expressing Parsons’ definition using just nominal variables. 12 Being in love with Mary requires that John’s emotional attitude towards her has to actually be a loving attitude. According to (INT), such property (being a loving attitude) holds intrinsically for John’s emotional attitude towards Mary since it holds independently of the presence or the absence of anything wholly distinct from such attitude. So, despite the fact that having a certain emotional attitude towards Mary is an extrinsic property for John, being a loving attitude is an intrinsic property of John’s emotional attitude towards Mary. For us, this is unacceptable, and indeed, again, it contradicts the intuition behind criterion (i). Suppose there are two exactly resembling roses, Rose1 and Rose2 , that are both red. According to Parsons, Rose2 , besides Rose1 , is also a truthmaker of ‘Rosie1 is red’. But of course it is not true that ‘Rose1 is red’ cannot become false without a qualitative change of Rose2 , so this contradicts (i), running indeed against commonsense. In conclusion, criterion (ii) is not enough accurate, since any duplicate of a truth- maker is also a truthmaker of the same proposition, so we have ‘irrelevant’ truthmakers. This is an instance of the general relevance problem, which also affects strong truthmak- ing. The strategy that we shall adopt to address this problem consists in a restriction of WT1b in the light of the following general principle, which in our view should be assumed for both strong and weak truthmaking: (CC) If a proposition p has arguments x1 , ..., xn , every truthmaker of p should be causally connected13 to at least one argument of p, where x is causally connected to y means that: x is existentially dependent on y (including the case where x = y), OR x is internal to y, i.e., it is either inhering in y, proper part of y, or participant to y. So, going back to the previous example, we can exclude Rose2 from being a weak truthmaker of ‘Rose1 is red’ since it is not causally connected to Rose1 . There is however a final problem concerning relevance: not only a truthmaker should be causally connected to the arguments of the proposition, but we also need to make sure that it is actually relevant to the meaning of such proposition. Consider for instance the case of an electron, e, whose charge and mass qualities, ce and me , have necessarily (i.e. in every possible world) certain values, vce and vmw . The two propositions ‘e has charge vce ’ and ‘e has mass mce ’ are then modally equivalent (i.e. true (false) in the same possible worlds), and so it turns out that both ce and me are weak truthmakers for each of the propositions, with no possibility to distinguish what accounts for what (whehter the mass or the charge). To avoid this, let us first assume that each proposition can be represented as a generic n-ary predicate r holding for a list of n arguments. We impose that the intrinsic property P that, according to (WT1b), characterizes a relevant truthmaker candidate should be constrained by the meaning of r. If r is a primitive, then the relevant intrinsic property should be r itself. Otherwise, if r is definable, that is, it is definitionally dependent [17] on other properties or relations, the relevant intrinsic property should be either such that r depends on it, or, in turn, definable in terms of properties or relations that occur in the definition of r. We shall use the predicate about(r, p) to say that the property p is relevant for the meaning of r (note that the first argument stands for a predicate of arbitrary arity, while the second one stands for a (unary) property). So, the property being a charge of value vce , which holds intrinsically for ce , is about the predicate has charge and not about has mass, since it is definable in terms of properties and relations that occur in the definition of has charge. Similarly, being a red color (holding for a color quality) is about being red, since being red is definable in terms of having a color quality that has 13 We are aware that this term may be not totally appropriate, but it will do for the scope of this article. The rationale for this constraint is to enforce a connection between the truthmaker and the arguments of the proposition at issue. the value ‘red’, and being a red quality is in turn definable as being a quality that has the value ‘red’. The relation about is in fact a way to associate a meaning to a predicate, by spec- ifying which entities are pertinent for its truth conditions. In this sense, it plays a role similar to a meaning postulate in the Carnapian sense, which specifies a constraint that holds in virtue of the meaning of the symbols involved, and is therefore stable across possible worlds (e.g. a bachelor is an unmarried male).14 We can then finally formulate a definition of truthmaking that takes into account our considerations regarding intrinsicality, restrictiveness and relevance (in terms of causal connectedness and aboutness): (WT1c) For every object x and every proposition p equivalent to a predicate r holding for arguments x1 , ..., xn at w, x is a truth-maker of p at w iff: 1. x is causally connected to at least one argument of r 2. there exists a property P such that r is about P, and P holds intrinsically for x at w iff p is true at w. 4. Conclusions In this paper, we elaborated on the innovative and non-standard philosophical notion of weak truthmaking proposed by Josh Parsons. We do this, firstly, by contrasting it with the standard account of (strong) truthmaking in the philosophical literature. We then conduct an in depth analysis of this notion and expose some difficulties in its original formulation. In particular, we demonstrate that Parsons’ reliance on David Lewis’ notion of intrinsically is in contrast with his own intuitions and that this renders his original formulation, on one hand, too restrictive w.r.t. the kinds of truthmakers that it enables and, on the other hand, not accurate enough, since any duplicate of a truthmaker of a proposition is also a truthmaker of the same proposition. This points manifests an instance of a general problem of relevance, which also affects strong truthmaking. We proposed a new precise understanding of weak truthmaking that not only addresses these issues but that is also in line with the ontological views of modern foundational theories such as DOLCE and UFO. In particular, we addressed: (i) the restrictiveness of Parsons’ approach through a new analysis of the scope of the modifier ‘intrinsically’. This is done by leveraging on the de re/de dicto distinction in philosophical logic. (ii) the issue of accuracy, by accounting for the relevance of properties in contributing to the meaning of sentences. This, in turn, is done by employing the notions of aboutness (which connects a predicate with the relevant properties that define it) and causal connectedness (which connects the truthmakers with particulars occurring in a sentence at hand). In our view, weak truthmaking should not be regarded as an alternative to strong truthmaking, but these two theories should be seen as complementing each other. In- deed, in a previous publication [7], we have shown how different truthmaking patterns combining the two theories can be of practical utility in concrete conceptual modelling scenarios. While in such previous work we where relying on an informal understanding 14 A similar move to associate predicative sentences with the pertinent ontological entities that are their truthmakers is discussed in [18]. A formalisation of the relevant implication for this purposes is provided in [19]. of the truthmaking notions, as future work, we shall present a formalization of these no- tions based on the account proposed here. In particular, we then intend to use this formal theory of weak and strong truthmaking to advance the formalization of UFO ([20, 21]) and specifically of the theory of relations in UFO, preliminarily addressed in [6]. Of course, many problems related to the notion of truthmaking are still there, in- cluding the problem of non-atomic sentences, which affect both strong and weak truth- making. We intend to address these problems in a future companion of this article. References [1] Fraser MacBride. Truthmakers. In Edward N. Zalta, editor, The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University, spring 2019 edition, 2019. [2] David Malet Armstrong. Truth and truthmakers. Cambridge University Press, 2004. [3] Friederike Moltmann. Events, tropes, and truthmaking. Philosophical Studies, 134(3):363–403, 2007. [4] Nicola Guarino and Giancarlo Guizzardi. Relationships and Events: Towards a General Theory of Reifi- cation and Truthmaking. In Proc.15th AI*IA, 2016. [5] Nicola Guarino and Giancarlo Guizzardi. “We need to discuss the Relationship”: Revisiting Relationships as Modeling Constructs. In Proc.27th CAiSE, 2015. [6] Fonseca, C. M, Porello, D., Guizzardi, G., Almeida, J. P., Guarino, N. Relations in Ontology-Driven Conceptual Modeling. In Proc.38th ER, 2019. [7] Guarino, N. et al. Reification and Truthmaking Patterns. In Proc.37th Int. Conf. on Conceptual Modeling (ER 2018), 2018. [8] Josh Parsons. There is no ‘’truthmaker” argument against nominalism. Australasian Journal of Philoso- phy, 77(3):325–334, 1999. [9] Josh Parsons. Truthmakers, the past, and the future. In Helen Beebee and Julian Dodd, editors, Truth- makers: The Contemporary Debate. Clarendon Press, 2005. [10] Claudio Masolo, Stefano Borgo, Aldo Gangemi, Nicola Guarino, and Alessandro Oltramari. Wonderweb deliverable d18. Technical report, CNR, 2003. [11] Giancarlo Guizzardi. Ontological Foundations for Structural Conceptual Models. Telematica Instituut / CTIT, 2005. [12] John Bigelow. The reality of numbers: A physicalist’s philosophy of mathematics. Oxford University Press, 1988. [13] Greg Restall. Truthmakers, entailment and necessity. Australasian journal of philosophy, 74(2):331–340, 1996. [14] Friederike Moltmann. Abstract objects and the semantics of natural language. OUP Oxford, 2013. [15] Adolf Rami. Introduction: Truth and truthmaking. In E J Lowe and Adolf Rami, editors, Truth and Truthmaking. Acumen, 2009. [16] Rae Langton and David Lewis. Defining ‘intrinsic’. Philosophy and phenomenological research, 58(2):333–345, 1998. [17] Kit Fine. Ontological dependence. In Proceedings of the Aristotelian society, volume 95, pages 269–290. JSTOR, 1995. [18] Claudio Masolo, Alessander Botti Benevides, and Daniele Porello. The interplay between models and observations. Applied Ontology, 13(1):41–71, 2018. [19] Claudio Masolo and Daniele Porello. A cognitive view of relevant implication. In Proceedings of the 3rd International Workshop on Artificial Intelligence and Cognition, Turin, Italy, September 28-29, 2015, pages 40–53, 2015. [20] Daniele Porello and Giancarlo Guizzardi. Towards a First-Order Modal Formalisation of the Unified Foundational Ontology. In Proc.3rd JOWO, 2017. [21] Guizzardi, G., Fonseca, C. M, Benevides, A., Almeida, J. P., Porello, D., Sales, T. P. Endurant Types in Ontology-Driven Conceptual Modeling: Towards OntoUML 2.0. In Proc.37th Int. Conf. on Conceptual Modeling (ER 2018), 2018.