Properties, Direction of Fit and an Argument for Property Realism

There is much debate about the existence of properties and how to consider both tropes and universals. While the antirealist contends that there are no true universal properties, the realist affirms that there are mind-independent universal properties. In this essay I present an argument for the realist approach to the problem of universals based off of the concept of direction of fit. After beginning with a discussion of properties, I discuss antirealist approaches to explaining the problem of universals, arguing that there are inherent problems with antirealism about properties that render it unsatisfactory. It becomes clear that universals must exist, since a mind-to-world direction of fit is necessary to enable the mind to interact


Introduction
In the debate about the nature of predicates, properties, and how they are related, some philosophers contend that there are no true universal properties (e.g. John Locke, Nelson Goodman, and W. V. Quine), while others accept the existence of mindindependent universals and explore the nature of their existence. The former group is antirealist, rejecting the existence of universal properties. The realist, in contrast, affirms that there are mind-independent universal properties. In this essay I present an argument for the realist approach to the problem of universals, which involves the concept of direction of fit. Direction of fit is repurposed from its typical application to mental states and is instead applied to properties and their existence relative to the mind. After beginning with a discussion of properties, I discuss antirealist approaches to explaining the problem of universals, arguing that there are inherent problems with antirealism about properties that render it unsatisfactory. Finally I will present my own argument for property realism. Ultimately it will become clear that universals must exist, since a mind-to-world direction of fit is necessary to enable the mind to interact with the world.

A discussion of properties
Many philosophers embrace the existence of properties in order to explain philosophical phenomena such as qualitative recurrence, e.g., instantiations of colors that are generally recognized and acknowledged such as 'red' or 'blue' (Swoyer, 2011). One popular consensus in defining properties seems to be that properties are entities which can be predicated (with correct and meaningful syntax) or attributed to things (Achinstein, 1974). In this case 'predicated' and 'attributed' can be considered synonymous. For instance, to say, "That is a yellow banana", of a yellow banana would be to attribute the properties of yellow and banana to the object. However, this raises the issue of whether or not properties can exist without predicates. A further discussion of this will be presented later. Another definition of properties proposes that they constitute objects, i.e., properties are things that come into being through the existence of objects and are instantiated by those objects. This definition implies that the formerly mentioned banana instantiates the property of being yellow and being a banana; being yellow, as well as being a banana, are properties that come into being through the existence of this object that we call a banana and other objects that we consider bananas. This goes beyond considering properties solely through predication by also associating them with their actual instantiation.
Both of the aforementioned definitions of properties really only touch on particulars rather than universals. At this point one may wonder whether properties act as universals or particulars, or perhaps both. Can a single property be instantiated by an indefinite number of distinct things, or are properties particular to each individual thing? In other words, can any banana instantiate the exact same color of yellow, or will no two yellows ever be the same?
To answer these questions we must first consider the concept of universals. Universals are mind-independent entities. Universals act both as characterizers and unifiers, since they characterize objects and unify these with other objects that are characterized by the same universal (Orilia, 2008). For instance, two blueberries would be unified by virtue of their sharing the properties of being blue, being the same kind of fruit, and so on. In this way, individual objects are explained as being similar to other objects due to their sharing a common universal. This 'sharing' of a universal indicates a commonality in the possession of the property in question, much like two people with brown hair 'share' the property of being brunette, or have brown hair in common. For instance, a sapphire and a blueberry with the same qualitative blueness would be said to share the universal property blue. By this thinking, it would be true that a single property could be instantiated by an indefinite number of different things.
Particulars properties (or "tropes"), on the other hand, are individual instances of properties that are thought to be unique compared to other instances of even a qualitatively identical property. Tropes characterize but do not unify, since these properties cannot be shared by distinct objects and therefore cannot be related to one another (Orilia, 2008). For instance, in this case the blue of a blueberry would be a distinct blue from the blue of a sapphire, or even another blueberry, even if these two instances of blue were qualitatively identical (Massin, 2008). In this case it would be true to say that properties are unique to each individual thing, and are not instantiated by an indefinite number of distinct things, in contrast with the case of universals.

A survey of antirealist views on properties
The antirealist about properties denies the existence of universals. Some antirealists, however, maintain that there are still properties that are tropes (e.g. Williams, 1953;Campbell, 1990;Ehring, 2011). There are two main anti-realist views: conceptualism and nominalism. Conceptualists contend that properties are really only features of the mind, and are thus mind-dependent entities. By the conceptualist belief there aren't really true properties, just concepts or our tendency as humans to categorize things. For instance, the aforementioned blueberry and sapphire would both be blue not because they share a universal property of blueness, or a particular instance of blue that happens to be qualitatively identical, but instead because the human concept of blue applies to both objects (MacLeod et al., n.d.). This will be expanded on during the discussion of predicate nominalism.
The other main antirealist belief is nominalism. Nominalists do recognize the existence of tropes while rejecting the existence of universals. There is a large variety of nominalisms, but most of these seem to stem from three main beliefs: resemblance nominalism, trope nominalism, and predicate nominalism (MacLeod et al., n.d.). Resemblance nominalists try to address commonality in nature by appealing to "resemblance classes". This type of tropist view is referred to as "resemblance nominalism". Orilia explains that an example of a resemblance class would be the colors "red", "crimson" and "scarlet" which are used ambiguously to refer to tropes of this color group that are similar to each other (Orilia, 2008). However, there is no good way to determine what instances would belong to a particular resemblance class, or how to determine the parameters of resemblance classes in general. For instance, would maroon belong to this resemblance group? Or fuchsia? There could be no effective way to objectively define which properties would belong to a particular group and which wouldn't, since decisions of this type are largely subjective and would thus lead to disagreement upon those deciding how to define a particular resemblance class. In addition, the concept of resemblance classes itself seems to be a way to acknowledge universals simply by a different name. To say that resemblance classes unify tropes by recognizing similar instances of a particular property is remarkably similar to saying that the same property can be instantiated by various objects, such as the instantiation of "red" by a ruby, a tomato, or a brick. In this way resemblance nominalism collapses into universal theory, and so we will move on from the notion of resemblance classes with only a consideration of tropes and their ability to characterize, but not unify.
Trope nominalists see individual objects as agglomerations of tropes, and they explain the qualities of objects in virtue of this (e.g. Williams, 1953;Campbell, 1990). This is also known as bundle theory, since objects or particulars are considered be collections or "bundles" of tropes (Simons, 1994). By this thinking a banana would have a yellow trope and a banana trope, as well as other particular properties that characterize it (a taste trope, a shape trope, etc.). Although tropes can resemble each other, each trope is unique and distinct from other instances of even qualitatively identical properties, as mentioned before. While trope nominalism seems an appealing alternative to many, recall that tropes only characterize but do not unify. How can we explain the obvious unity in nature without somehow also unifying instances of qualitatively identical properties? Despite the flaws in conceptualism and predicate nominalism, each of these approaches at least attempts to address this unity. Trope nominalism fails to do this.
Predicate nominalists paint a similar but slightly different picture. While properties refer to the way something is, predicates are linguistic expressions that express these properties. Predicates aim to provide information about the subject they describe, and are used to indicate that the property they express is instantiated by a particular object or thing. For instance, in the sentence "that is a yellow banana", "is a yellow banana" is a predicate that indicates that the subject ("that") instantiates both the property of yellowness and the property of being a banana. However, not all predicates are meaningful. For instance, the predicate "is a square circle" requires the combination of two properties that are contradictory. That being said, it is still a legitimate predicate (Heil, 2013). One can therefore observe that predicates can exist that do not express properties. Although truthfulness is not a requirement for predication, predicates are intended to be "truthmakers" and to have semantic value, i.e. to make a true claim about the object they refer to (Rauti, 2008), which is something that inconsistent predicates do not do. Now the question arises: can properties exist independently of predicates?
Predicate nominalists see properties as dependent on predication. They contend that both a blueberry and a sapphire are blue because the predicate "is blue" can truly be said of both objects. Predicate nominalists therefore also refer to commonality similarly to trope nominalists by making reference to individuals. However, unlike the trope nominalist, the predicate nominalist also stresses the semantic value of the linguistic expression that unites these similar individuals (Stout, 1923). Both predicate nominalists and conceptualists face the same crippling problem, similar to the aforementioned resemblance nominalists: how do we know what "being x" is? How can one determine which predicates apply and which do not? There is no good way to objectively determine what concepts apply to certain objects and not others. For instance, how would one determine that blue could be applied to two objects a and b, but not a third object, c? In order to explain this, the conceptualist and the predicate nominalist are forced to differentiate between the objects by affirming that a and b share some property that c lacks. This statement, however, turns both conceptualism and predicate nominalism back into realism by recognizing a common quality shared by the two distinct objects. To say that a and b both look similar to the observer but differ in some respect from c would be entirely arbitrary; any person could differ in their opinion of this because it is subjective. Even if one were able to identify a commonality between two objects that they could call x, there is no good way to objectively determine how the same x is manifested in distinct objects, e.g., how blue would be applied to the two objects a and b, but not the third object, c, without entering into infinite regress, i.e., without necessitating reason or evidence to justify each parameter and more reason and evidence to support that reason and evidence, etc. (Armstrong, 1973).

Direction of fit as an argument for property realism
At this point we have considered several of the more popular views on tropes and antirealism about properties. In contrast, the realist about properties accepts the existence of universals and uses them to explain qualitative recurrence (Armstrong, 1978). The realist appeals to our basic intuition that there is qualitative commonality in nature and not just coincidental, qualitatively identical tropes that compose objects; in other words, there are properties of distinct objects that both characterize them and unify them with other objects. We can look at this in more than one way: we can consider universal property that is instantiated by multiple objects, or we can consider just a particular instantiation that is unified by the concept of universals. Either way, the notion of universals is the only effective way to truly address duplicity and commonality in nature. One might argue that realism is unnecessary, because they believe that their ontological needs are met merely with a consideration of tropes (e.g. Keinänen, 2014;Goodman and Quine, 1947). However, I will argue that this is not the case. It seems bizarre to imagine that two qualitatively identical instantiations of a property would not, in fact, be identical. Even if only intuitively, it seems obvious that there is a commonality in nature that unifies distinct objects, and universals explain precisely this intuition. This would mean that two qualitatively identical objects could be identical by virtue of their sharing the same universal properties. For instance, in theory, two qualitatively identical Ford Tauruses could be identical, if they were to share the same properties of size, shape, color, and all of the other properties that define that car as a Ford Taurus (make, model, specific parts, etc.). However, this particular example may be only in theory, as methods of human production are not perfect, nor are our methods of measurement. The logic behind this thinking becomes especially clear when one realizes that there are properties that are mind (and, therefore, predicate)-independent. This can be shown by employing direction of fit. Gertrude Anscombe is credited with one of the clearest definitions of direction of fit in her book Intention, although the concept itself seems to have originated with J. L. Austin (Driver, 2011). Anscombe employs direction of fit to reference mental states (Anscombe, 1957). However, in this discussion direction of fit is repurposed, and instead addresses the way the one's mind (or, by extension, predication) interacts with the properties one sees in the world; we can either try to make our minds match the goings-on of the world, or we can try to make the world fit with the mental-happenings of our minds (Driver, 2011). This notion is particularly useful in philosophy of mind when talking about belief and desire. While beliefs involve a description or explanation of the world, desires aim to bring about change and affect the world. For instance, beliefs such as "it will likely be sunny tomorrow" align one's mind with the state of the world, and therefore exhibit a mindto-world direction of fit. In contrast, desires such as "alphabetize these books!" aim to affect the world in accordance with one's ideas, and therefore exhibit a world-to-mind direction of fit (Platts, 1979). Some philosophers, such as David Velleman, distinguish between these two directions by terming states with a mind-to-world direction of fit 'facta' (e.g. "it will likely be sunny tomorrow") and states with a world-to-mind direction of fit 'facienda' (e.g. "alphabetize these books!") (Velleman, 1992). Other examples of facta include perceptions and hypotheses, whereas facienda are mainly intentions and desires. Velleman also refers to facta as cognitive attitudes, and facienda as conative attitudes (Velleman, 1992).
As John P. Le Coq states in his essay The Function of Language, "Language proceeds from nature because it is generated by man who is a product of nature" (Le Coq, 1955). In the same vein of thought, I believe that generally it is the purpose of words to capture the essence of the world; it is for this purpose that language developed, to communicate goings-on, intentions, and articulate the properties or qualities of one's experience. Based on this logic, one can see that life is mostly composed of a mind-to-world (or word-toworld) direction of fit: we shape our words and beliefs in an effort to describe the world that we perceive and experience. This must be true, because the converse would involve knowing or understanding qualities that one has never experienced, or even predicating these qualities. This is obviously nonsensical. For instance, one would not know to identify or claim "that is blue" if they had not first seen what 'blue' looks like, or say "this tastes salty" if they had not first tasted salt. That said, regardless of the person knowing or proclaiming "this is salty" or "that is blue", both objects would continue being 'salty' and 'blue', respectively. For instance, whether or not I have ever seen a blueberry or the color blue instantiated, a blueberry will continue being a blueberry and that blueberry will continue being blue. The same would go for a saltine cracker; even if I had never eaten one before, the saltine would still taste 'salty' due to its having the substance we know to be table salt, i.e. sodium chloride. Were someone with a normally functioning sense of taste to begin to eat the saltine, they would experience (taste) its saltiness. These are few of many examples that show that properties can exist regardless of predication, and are thus mind-independent. In fact, the realm of predicates and the realm of properties are almost entirely distinct, and really only come to overlap through language. Simply because I do not refer to the fruit that we know as a yellow banana as a "yellow banana" doesn't make it any less yellow or any less of a banana. In fact, I could falsely predicate the object that we know to be a banana as a "plum", or a "brick", and it would continue being just as it is in reality.
Another example of this is in body dysmorphic disorder. Say a man were to be convinced that he was a woman. This man has all of the defining characteristics of being a man, but he is nonetheless convinced that he physically possesses the female sex organ, breasts, and other defining characteristics of females. Despite his beliefs, this man continues being a man. His beliefs about his appearance do not in fact change his appearance; he does not suddenly develop the characteristics of the opposite sex, and other observers would not suddenly observe them. Again, it is evident that actual properties exist regardless of one's beliefs. Just because one believes that the sun is cold, or even desires that the particles of the sun lose their thermal energy, that will not in any way affect the actual state of the sun. One therefore sees that a mind-to-world direction of fit truly predominates in respect to our relationship with properties; properties will continue being properties regardless of our mental goings-on or the existence of predication, and one must therefore strive to conform or alter their mind (including beliefs and desires) to the actual happenings of the world.
That said, I do believe that there are instances, albeit fewer, where a world-to-mind or world-to-word direction of fit is useful and perhaps even necessary. The "world" that each individual experiences is undoubtedly distinct from the world that someone else experiences. Words can act as a tool in trying to allow others to comprehend one's own particular experience of the world. Here we encounter the realm of communication, which is perhaps the essence of language and, by extension, predication. For instance, imagine a friend trying to explain a phenomenon that they have experienced but you never have. Perhaps what they're trying to explain to you is the sensation of being so happy that they could barely contain it, joyously enthusiastic to the point where they are almost unrestrained. The English language has a word for this sensation: exuberance. Your friend tells you they were feeling exuberant, and you look up the word and find its definition: "joyously unrestrained and enthusiastic" (Merriam Webster, 2011). You know what it is to be joyous, to be unrestrained, and to be enthusiastic, and therefore synthesize this information into an understanding of how your friend felt. In this way, we can strive to express to others our experience by condensing it into words, which we can share with someone else. Communication acts as a context for world-to-word (or world-to-mind) direction of fit. Inevitably, however, this relationship will again be reversed back to a word-to-world direction, and the word "exuberant" will be applied to capture their own feelings, or their impression of someone else, when this instantiation becomes part of their experience. In this way, world-to-word direction of fit acts as a tool, but ultimately becomes its own unique direction: world-to-mind-to-world.

Conclusion
It ultimately becomes clear that a mind-to-world direction of fit predominates and thus the presence of properties is independent of our own presence, or our making reference to them. Language truly is but a tool to express one's experience of the world. By reducing one's experience to words one can share their experience with others, who can in turn use that predication to capture their own experience. However, this will ultimately return back to a mind-to-world directionality (and we can thus consider this to have world-to-mind-to-world directionality). This evidence reaffirms the realist approach to properties, namely that there is commonality in nature which is explained by the presence of mind-independent universals.