Possibilia, Qualia, and Sensibilia

Possibilia, Qualia, e Sensibilia

Alberto Voltolini[a] alberto.voltolini@unito.it

Universidade de Turim (UniTo), Italia

Esta obra está bajo una Licencia Creative Commons Atribución 4.0 Internacional.

Como citar: VOLTOLINI, A. Possibilia, Qualia, and Sensibilia. Revista de Filosofia Aurora, Curitiba, v. 34, n. 63, p. 34-50, out./dez. 2022

From a semantic point of view, the departure point for showing that the objects of the Tractatus are possibilia is the Wittgensteinian thesis according to which objects are ineffable, or rather, it cannot be stated that these objects are (TLP 3.221). This ineffability is linked to the fact that, acknowledging the semantic theory of the Tractatus, it is meaningless to attribute to something an internal property, a property that that very something cannot fail to have (TLP 4.122). Given the bipolar nature of the proposition, so that a proposition must be able to be as true as it can be false^[2], all that can sensibly be said is what it might otherwise be, while existence is an internal property for objects, i.e., a property that objects cannot fail to have; in other words, objects exist necessarily.

At this point, the question immediately arises as to the nature of this property of existence that characterises objects necessarily. The first conjecture is that it is the very same property enjoyed by other (presumed) entities that constitute the ontological skeleton of the Tractatus, that is the complexes; the difference would simply be that whilst the complexes enjoy it contingently – indeed it makes sense to say that a complex exists – objects must enjoy it necessarily.

But this conjecture does not work. Indeed, implicitly in the Tractatus but explicitly in the Notes on Logic, and again in the Introduction by Russell to the first English translation of the Tractatus, we have a definition available concerning the relevant property when applied to complexes: to say that a complex . exists means none other than that its constituents are assembled in a certain way^[3]. From this definition, we immediately have that this cannot be the property that is enjoyed by objects necessarily. As is well known, for the Tractatus, objects are simple, i.e., entities that cannot be de-composed (TLP 2.02, 2.021); thus they cannot have constituents. And so, for an object, to exist cannot mean being such that its constituents are assembled in a certain way^[4].

Another conjecture can be advanced, though. Since objects exist necessarily, for them existing will mean occurring in some subsisting state of affairs. The possible states of affairs that actually subsist may well fail to subsist; but in a possible world in which these states of affairs do not subsist, the object that is their constituent will nonetheless be part of some other possible state of affairs subsisting there, and so on. For the supporters of this idea, it follows that the property of occurring in some subsisting state of affairs is necessary for an object; thus, it may well coincide with the property of existence which, for the sake of argument, it necessarily enjoys^[5].

But this conjecture does not work either because, as has rightly been pointed out^[6], it does not take into account the possibility of the empty logical space that Wittgenstein speaks of in TLP 2.013. The latter is the possibility that the world is such, that in it there is no possible state of affairs; in other words, it is the possible world in which no possible state of affairs subsists (expressed from a semantic instead of an ontological perspective, this is the possibility that all elementary propositions are false). In this possible world no object enjoys the property of occurring in a subsisting state of affairs, quite simply because in that world there is no possible yet subsisting state of affairs. Thus, this latter property is not a property that an object enjoys necessarily; a fortiori, it cannot be identical to the property of existence that the object necessarily enjoys. In other words: objects also exist in the possible world where no possible state of affairs subsists.

But the possibility of an empty logical space already indicates what the next alternative conjecture might be on the nature of the existence of objects. Since, in the world of empty logical space, an object cannot figure in any subsisting state of affairs, because no such state subsists in that world, the only thing it can do in that world is to belong to its domain. But then it can well be said that for any object existing amounts to the fact that, for every world, this object belongs to that world’s domain. It cannot be denied that objects do enjoy this property; to say that any object enjoys this property is tantamount to saying that the domain of all possible worlds is fixed, which is what Wittgenstein actually holds in TLP 2.022-2.023, 2.026-2.027-2.0271^[7]. Now, a property like this is certainly a property an object enjoys necessarily; ergo, it can well be identified with the property of existence we were looking for and which objects do enjoy necessarily.

Now, let me call this property the property of existence., to distinguish it from another property of existence, let’s say existence., which is precisely the property previously considered to occur in a subsisting state of affairs. A Tractarian object does in fact enjoy this second property but only in the worlds in which it actually figures as the constituent of a state of affairs subsisting there; since, as we already know, there is at least one world in which such an object does not enjoy them, the world of empty logical space, the object contingently enjoys this property. But therefore objects necessarily enjoy existence.and contingently existence.; this makes the objects possible, in the sense of entities that may exist as they may not exist (in the sense of existence.). Russell spoke of this very type of entity in 1903, when he argued that entities are such that they necessarily possess being, but contingently possess existence^[8]. To obtain that Tractarian objects and Russell’s possibilia are identical, it is sufficient to treat Russell’s being and existence exactly the same way as our existence¹ and existence².

Any reader of the Tractatus will be surprised to hear that Tractarian objects are possibilia.^[9] In the Tractatus Wittgenstein never explicitly advances such a thesis, which, by the way, independently of exegetical issues, appears in itself to be controversial on an ontological standpoint: can there be entities that may exist just as they may not exist, and therefore be entities that actually do not exist?

It should be noted that I wish to ascribe this thesis to Wittgenstein not as a purely ontological thesis, but as an onto-semantic one, i.e., regarding the nature of objects, to the extent of saying what objects should be like for language to be able to speak of them and make sense. In fact this argument can be ascribed to Wittgenstein as deriving from theses that in the final analysis are purely semantic: i) there must be objects for language to have a determined sense (TLP 3.23); ii) these objects must exist necessarily for the propositions to make sense, independently of how things stand (TLP 2.0211-2.0212); iii) the elementary propositions must be able to all be false, as required by the thesis of the logical independence of these propositions (TLP 5.134) – which, ontologically expressed, is none other than the supposition of empty logical space, or in other words, that there is a possible world without subsisting states of affairs. Taking these theses all together, it logically follows that Tractarian objects are possibilia; and as we know, at the time of the Tractatus Wittgenstein was only interested in objects from a logical point of view, in other words only from the perspective of how the world must be for language to have meaning^[10].

This does not mean that we cannot legitimately go on to wonder, from a purely ontological point of view, what the objects of the Tractatus might be (from a perspective that, by contrast, I might term onto-metaphysical). However, it must be remembered that any such investigation will have to agree with the logical prerequisites regarding the nature of objects; or, in Wittgenstein’s words, it is true that only the application of logic can tell us in practice what elementary propositions are like, and thus objects; but there must be no disagreement between logic and its application, in other words the application must conform to the general requirements posed by logic (TLP 5.557)^[11].

Recently, Pasquale Frascolla has argued that objects in the Tractatus are qualia, in the sense of universals, i.e., abstract, repeatable entities, of a phenomenological nature (like colours, places and phenomenological moments)^[12]. The interesting thing is that Frascolla does not want to confine himself to upholding the weak thesis that Tractarian objects are qualia from the point of view of applied logic, and thus from the perspective of a determined ontology compatible with the onto-semantics of the Tractatus, but the stronger thesis that these objects are qualia from the perspective of logic itself, thus from the perspective of the corpus of ontological theses advanced for giving an account of the good sense of our language; thus, to use the previous terminology, from an onto-semantic perspective rather than an onto-metaphysical one. Conceived of as qualia in the latter sense, in fact, according to Frascolla it is first of all possible to give a good explanation of why these objects exist in the world of empty logical space, too; in this world, objects will be there as qualitatively abstract universals, but we will not have there any subsisting state of affairs corresponding to the instantiation of several such universals, i.e., the effective combination between them of such universals (e.g. the possible state of affairs red now here, constituted by the combination of the universal chromatic-phenomenological red, the temporal-phenomenological one now, the spatial-phenomenological here, will not subsist there; but neither will there be the possible state of affairs red now there, which has as a constituent the different spatial-phenomenological universal there, nor the possible state of affairs red then here, one of whose constituents is the different temporal-phenomenological universal then, and so on for any possible state of affairs). Moreover, according to Frascolla, this conception of objects makes it possible to explain two ontological theses in the Tractatus; namely, TLP 2.0232, according to which objects are colorless, and TLP 2.0251, according to which space, time and chromaticity are forms of objects, whose compatibility has always been a brainteaser for interpreters. For Frascolla, since they are qualitative universals, objects are colours (phenomenological), places (phenomenological), moments (phenomenological). As such, however, they do not have, nor can they have, any (determined) colour, space and time; this is why they are colourless, and presumably also aspatial and atemporal. This does not mean, however, that they do not fall under the general formal concepts corresponding to colour, space, time. Indeed, for the Tractatus an object not only has formal or internal properties, the properties necessary to it, but also falls under the corresponding formal and highly general concept of object (TLP 4.126), and thus, according to whether a colour, a place or a phenomenological moment is involved, it will also fall under the formal, general concept of colour, space and time. This fact of coming under the respective general, formal concept allows a determined object to also have the respective formal quality, or, in other words, allows chromaticity, space and time to actually be forms of objects.

As I have said, Frascolla’s claim is that the conception of objects as phenomenological universals does not hold only at the onto-metaphysical level, but also specifically at the onto-semantic one; in other words, the fact that objects are qualitative universals must be part of what the world must be like, for language to have meaning. From this point of view, the claim seems to me to give rise to perplexity.

First of all, what need is there to consider Tractarian objects as abstract entities from the onto-semantic point of view? So doing, we are in fact led to suppose that from an onto-semantic point of view it is relevant to consider as important what is normally a distinction in onto-metaphysics, the distinction between abstract entities and concrete ones, according to which the former exist in a mode that is not spatio-temporal, whilst the latter exist in a spatio-temporal mode^[13].

Yet for Frascolla this is precisely how things stand: alongside simple objects taken as abstract entities (qualitative-type universals), and thus as non spatio-temporal, there are complexes, the (subsisting) states of affairs, which are concrete entities and thus spatio-temporal, since they are none other than instantiations of several abstract entities, of several qualitative universals.

But here the following question arises: in what sense can a subsisting state of affairs be seen as the instantiation of the universals that constitute it ex hypothesi? According to what Frascolla maintains, here we have to do with a many-one relationship: different universals (e.g. red, the phenomenological now, the phenomenological here) are ‘instantiated’ in one same subsisting state of affairs (red now here). But, as it has been traditionally conceived of, the instantiation relationship is a one-many relationship: one and the same abstract universal can be instantiated in a number of concrete examples. Thus, whatever ‘instantiation’ may be here, it does not seem possible that this is instantiation.

Naturally, Frascolla could resort to saying that the components of a subsisting state of affairs are not the universals themselves but their instantiations, in such a way as to realign with the traditional conception of the relation of instantiation as a relationship whose members are universals and their exemplifications. But this seems to go against the reading and, in my opinion, the spirit, too, of the Tractatus.

TLP 2.011 maintains that objects, they themselves, essentially have the property of being constituents of a possible state of affairs, or equivalently, the property of possibly being constituents of a subsisting state of affairs, since subsisting states of affairs are none other than possible states of affairs that subsist. Since necessarily enjoying that modal property implies contingently enjoying the corresponding demodalised property of being constituents of a subsisting state of affairs, they themselves, the objects, and not some Ersatz of them, constitute the subsisting states of affairs; as is, by the way, expressly suggested by various passages in the Tractatus (TLP 2.01, 2.0272-2.031).

Yet, Frascolla would retort, quite apart from the problem of the relationship between objects and subsisting states of affairs, objects cannot fail to be be abstract entities like universals, for they are described by Wittgenstein in TLP 2.0232 as colourless, and then also as aspatial and atemporal; if at all, it is the complexes that have colour, as well as space and time, and are therefore concrete.

In itself, however, the thesis according to which objects without colour (and thus also without spatial or temporal properties) can be explained without going into any onto-metaphysical theses regarding their nature. Having a determined colour (or a spatio-temporal position) is, as Frascolla himself would acknowledge, a property of complexes, what Wittgenstein in TLP 2.0231 calls a material property. Indeed, material properties, writes Wittgenstein in this case, are formed by the configuration of objects. This means that such a property is a property of complexes and not of objects. Thus, in itself, the thesis that objects are colourless is not problematic: objects are colourless because they are not complexes.

But, Frascolla would reply, the problem is, as we have already seen, how to combine TLP 2.0232 with TLP 2.0251, according to which space, time and chromaticity are forms of objects. The argument that objects are qualitative universals explains this brilliantly: objects do not have, nor can have, determined colours (they do not have, nor can have determined spatio-temporal positions) because they are colours (they are those positions), they are the red and the yellow (the now and the then, the here and the there); therefore, they fall under the formal concept of colour (as well as under those of space and time) and thus have chromaticity, space and time due to their form.

Nonetheless, TLP 2.0232 and TLP 2.0251 can be jointly explained anew without resorting to the thesis that objects are universals. Given that the form of an object in general is its possibility of occurring in (subsisting) states of affairs (TLP 2.0141), i.e., its possibility of combining in some way with other objects, and given that having form coincides with having formal or internal properties (TLP 2.0123-2.01231), an object also has chromaticity as form since it internally enjoys the property of possibly combining chromatically with other objects (and similarly, it has spatiality and temporality as forms since it has respectively the internal properties of possibly combining spatially and possibly combining temporally with other objects). Internally, or necessarily, enjoying such modal properties implies being able to enjoy the corresponding non-modalised properties, i.e., in certain possible worlds the external property of combining chromatically .spatially, temporally) with other objects (as Wittgenstein states in his general formulation, form is the possibility of the structure: TLP 2.033). But for an object enjoying such an external property does not mean enjoying the corresponding material property, let us say having a certain colour; simply, the fact that an object enjoys such an external property corresponds to the fact that a corresponding complex enjoys this material property. Thus, the object is colourless (aspatial, atemporal), although it can necessarily combine chromatically (spatially and temporally), i.e., it has chromaticity (space and time) as form. But none of what has been said in itself makes that object a universal. Of course, if an object is a (qualitative) universal, its form, let us say its chromatic form, i.e., its internal possibility of combining chromatically with other objects, can be described as the fact that this object can join with an universal of a different form, as Frascolla demands^[14]. But – precisely – if. Which is the same as saying that from an onto-semantic perspective objects are not required to be (qualitative) universals.

But there is more. The interpretative path Frascolla proposes, according to which objects are colourless (aspatial, atemporal) because they are colours (places, moments; i.e. qualitative universals), seems to be strewn with other dangers. Not only does it commit Frascolla to sustaining a thesis that is already controversial at an onto-metaphysical level, i.e. that for universals self-predication, the attribution of the property corresponding to them, does not hold; but risks, contrary to his own desiderata, to having to attribute it to them surreptitiously. We might in fact wonder if, from Frascolla’s point of view, a universal might not really have the property that characterises it. Let us see why.

First of all, Frascolla maintains that only the postulation of objects as universals identical in form but with a different internal nature, such as the phenomenological yellow and red, can explain the thesis of TLP 2.0233, according to which objects with an identical logical form can only be distinguished by the fact of being different.^[15] Here I shall not go into the aspect of how this explanation does not seem to be the only one possible; the simplest account of the passage in question is that Wittgenstein is arguing that nothing but their numerical difference distinguishes objects indiscernible in their logical form. The point is that if Frascolla’s explanation were correct, universals identical in logical form would seem to distinguish themselves by the very fact that they respectively have the properties that characterise them.

What is more, Frascolla seems to admit that conceiving of qualitative universals as Tractarian objects leads to violation of the thesis of the logical independence of states of affairs (TLP 1.21, 2.01-2.062, 5.135), since these universals stand in an internal relationship to one another; indeed, given the internal relationship, e.g. between the universal chromatic-phenomenological red and the universal chromatic-phenomenological yellow, the subsistence, e.g. of the state of affairs red now here including two temporal- and spatial- phenomenological universals and the first chromatic phenomenological universal, excludes the subsistence of the state of affairs yellow now here including the same temporal- and spatial- phenomenological universals but not the second chromatic-phenomenological universal.^[16]Now what is this internal relationship based on, if not on the fact that the two universals in question have two natures and thus in actual fact two chromatic properties that are incompatible with one another? In TLP 4.123 Wittgenstein describes two different shades of blue as two objects standing in relation to one another in terms of the internal relation of being respectively lighter and darker. How can the fact of these two shades standing in this internal relation to one another not depend on the fact of having different and incompatible chromatic properties?

As to the first problem, that of the difference in nature between two universals with the same logical form, Frascolla might argue that the nature of a universal is a primitive concept, impossible to analyse through that of properties. Thus, in the same way, with respect to the second problem, he might say that the internal relation existing between universals of a different nature should be considered as a primitive impossible to analyse further. This might lead Frascolla to note that from this point of view TLP 4.123 only poses a problem if the two singular terms “this blue colour” and “this other one” should properly denote Tractarian objects. But, he might continue, things might well not stand like this; Wittgenstein might be referring to two distinct complexes. In this sense, to speak here of an internal relation between two objects would be improper in the same sense in which it would be improper to say that an internal relation of exclusion exists between a state of affairs in which a point in the visual field has a certain colour in a certain time and another state of affairs in which that same point has another colour in the same time. The sentence “the point p, in the time t has the colour c and the point p in the time t has the colour c’” is not viewed by Wittgenstein as the conjunction of two elementary propositions respectively presenting two distinct states of affairs, but as a contradiction (TLP 6.3751).

Let us even suppose that this is a valid explanation of the passages in question and that therefore universal-objects stand in an internal relation to one another that cannot be further analysed. Nonetheless, if things stand like this, there remains what is for me a fundamental problem. If objects were qualitative universals, for the reasons we have previously seen above, the Tractatus would be defending an obvious contradiction, i.e., that the state of affairs are and are not independent of one another - the subsistence of the state of affairs red now here excludes the subsistence of the state of affairs yellow now here - and not more simply a thesis, that of the independence of states of affairs, that Wittgenstein’s subsequent reflection itself on sentences of the type “the point p in time t has the colour c” was to prove unsustainable. At the time of the Tractatus, Wittgenstein did not think that the impossibility of the truth of a sentence like “the point p has the colour c at time t and the point p has the colour c’ at time t” led him to argue that states of affairs are not logically independent of one another and thus to support in the text an evident contradiction on the question of the logical independence of states of affairs, to the extent that he did believe that the conjuncts of the above sentence were not elementary propositions, and therefore presentations of states of affairs.

Frascolla would presumably agree on this point. Indeed, according to him, what I consider a fundamental problem does not exist, because for Wittgenstein objects taken as qualitative universals would only apparently violate the principle of the logical independence of states of affairs; thus, at the time of the Tractatus, Wittgenstein would have thought that this apparent violation should be “analysed away” and only his subsequent considerations would take on the task of showing that this elimination was not actually possible^[17].

But at the time of the Tractatus could Wittgenstein have thought that this violation was only apparent without at the same time thinking that those universals were not really its objects? In the end, we saw previously that the way to acknowledge the attribution of an internal relationship between the two shades of colour, as this colour blue and that colour blue, which Wittgenstein speaks of in 4.123 is to argue that these shades are not objects but complexes. To think, therefore, that e.g. yellow and red do have an internal relationship but that this does not lead to the logical dependence of states of affairs, shouldn’t Wittgenstein have thought that yellow and red are not objects even at the time of the Tractatus?^[18]

Thus, the thesis that objects are qualitative universals raises perplexity in my mind. Nonetheless, Frascolla’s idea of accounting for some sections of the ontological part of the Tractatus as being fully part of its onto-semantics seems to me extremely stimulating; moreover, by mobilising phenomenological entities, it seems to me to be moving in the right direction. Is there therefore a way of saving this idea, giving up the accompanying thesis of objects as qualitative universals and instead integrating it with my previous thesis whereby holding to the onto-semantics of the Tractatus objects are possibilia?

There is a simple way to reply positively to this question and that is to see the objects of the Tractatus as sensibilia, i.e. as possible sense-data. As possible sense data, entities of this sort are entities that necessarily exist₁ but which may exist₂ as well as not exist₂, and therefore possibilia in all ways. Moreover, as we already know, the form of an object is generally the possibility it has to occur in (subsisting) states if affairs, i.e., to combine in certain ways with other objects. Thus, if space, time and chromaticity are specific forms of objects, these forms cannot but be respectively the possibility of an object to combine spatially, temporally and chromatically with other objects. But sensibilia are entities to which the properties of being able to combine chromatically, spatially and temporally with each other can quite well be assigned as internal properties. Lastly, objects are colourless (aspatial and atemporal) because, as I mentioned in the previous Section, as sensibilia, objects can well necessarily have internal, modal properties of combinability in which these specific forms consist, and contingently those external properties which are none other than those very same properties demodalised. Nevertheless, they will not have the internal chromatic, spatial and temporal properties which are material properties of those complexes that result from the effective combinations in chromatic, spatial and temporal modes, of these objects^[19].

Here, too, I cannot quote any textual evidence ascribing to Wittgenstein the awareness of this position. But I can say with certainty that this is a position Wittgenstein could have considered directly. First of all, I can advance a theoretical reason. Many people are ready ascribe to the Wittgenstein of the Tractatus a phenomenist position analogous to the one sometimes taken by Russell, according to which the world’s logical atoms, the entities that necessarily exist for language to make sense, briefly objects, are sense-data^[20]. Now, if sense data were not conceived of as possible, actualized sense-data, i.e., such that they actually, but contingently existed. as well as necessarily existed1, the position in question would not be defendable. Because, as Kripke has pointed out, for any such sense-data it is impossible not to exist only in an epistemic sense (its existence is undeniable), but not in a metaphysical sense (there certainly is a possible world in which it does not exist). And so, comments Kripke, the thesis of the necessary existence of sense-data is simply false, since it rests on the erroneous lack of distinction between epistemic necessity and metaphysical necessity. Thus, those necessary entities that are objects cannot be sense-data^[21]. But once we dispose of possible sense data-over and above the distinction between existence₁ and existence₂, then we can easily explain in what sense it is an epistemic impossibility and at the same time a metaphysical possibility for a certain sense-datum not to exist, i.e., in the sense of existing₂, while it is still a metaphysical impossibility that a sense-datum does not exist, i.e., it exists necessarily in the sense of existing₂; thus, if taken as a possible, actually existent., a sense-datum may well be a Tractarian object.

But as well as theoretical reasons, I can advance historical reasons in this case, though indirect ones. In 1914 Russell defended the very thesis that sensibilia, i.e. possible sense-data, were the ultimate constituents of the world.^[22] And this thesis is not only defended on an onto-metaphysical level but properly on the onto-semantical one; sensibilia are the ultimate constituents of the world since they are logical atoms, i.e., they are entities whose logically proper names refer directly to them, thus allowing language in general to have meaning. Indeed, Russell takes up Wittgenstein’s thesis here, according to which it is meaningless to say that a logical atom of the world, or a sensibile, does not exist^[23]; something which, as we have seen, in Wittgenstein amounts to the thesis that a logical atom exists necessarily, and thus ineffably (for reasons internal to the semantics of the Tractatus), in order to allow language to possess propositions that have sense^[24].