GREEK REALISTS
BRITISH CONCEPTUALISTS AND NOMINALISTS
WITTENSTEIN'S ANTI-UNIVERSALISM

GREEK REALISTS

Plato was concerned with the meaning of universals, and spoke of them as abstract objects of thought, rather than sensed objects. Obviously universals are not what we see. But we could not recognize things for what they are without universals, or without generalities, either as nouns, verbs, adjectives and adverbs. So, what we perceive are particulars, but we know things to be depend upon either the recognition of or the application of universals. All knowledge is about `kinds' of things, qualities or events, and this `kind' is a classification based upon common elements or relations between particulars. Generality is the essence of any language, since all words, except proper names, are general.

Plato sought for a universal definition of each virtue, such as Goodness, Truth, Beauty, Courage, Piety, etc. He was asking how do we know this action to be of a certain virtue, or this object to be what we call it. What is being recognized in each instance or in each object that we correctly call them by the same name? What ARE we recognizing in these slightly different actions or things? And how do we know if we are correct? It seemed inadequate to just point out instances of these virtues from our experiences, because there should be some reason, or a fulfillment of some premise, for one to point out the right (or deserving) instances. Also, Plato could not accept the Sophist theory that these definitions of virtue are always conventional, that is, what is merely socially agreed as examples or definitions of virtues.

Thus, Plato's theory of universals places them ontologically prior and metaphysically independent of this world of expression, so Plato's universals become definite Ideals to be sought, or Perfect Pattern-Forms to guide action in this world. Plato's universal forms, then, are an unchanging, world-independent knowledge, of which particulars participate in epistemically and metaphysically. Particulars find their meaning and recognition in these forms, as well as their source of becoming truly and fully what they essentially are. The universal forms remain the same, which is necessary for a stable epistemology, while particulars of the same universal kind might vary somewhat in their differences and in how various forms are combined through them, and in their degrees of perfection or in their being close to being the most perfect exemplar of that apriori universal form.

Aristotle argued against Plato's ontological priority of universals and his giving them a metaphysical existence independent of this sensed world. Aristotle saw this fixing of universals outside of the world as unnecessary, and saw as wrong the implication of universals as metaphysical substances patterning the world. Instead of particulars participating universal qualities, such as trees being green-like; Aristotle proposed that particulars possess certain universal predicate qualities, such as trees having a green nature.

For Aristotle universals are predicate properties or essences of particular substances, or generic classifications of substances. For Plato universal forms are intelligibly-known individual substances having certain properties or definable characteristics; while for Aristotle universal forms are the definable properties or common essences of individually sensed substances. For Aristotle, whiteness or courage is the name of a common property which all white things or courageous acts possess. Plato might say that these things or acts resemble to some degree those universal qualities. We classify things according to the essential properties they hold in common, and it is these common properties which govern our use of universal terms.

Still though, Aristotle maintained that universals must logically and necessarily be epistemologically stable. The universal is what is predicated of the individual object or event, because that universal is shared by all similar objects or events. Particulars experienced in the world can be grouped according to their common features. These common features, such as green or leafy or noisy or rational or whatever, are themselves universals qualities, and what they are common to, such as trees or ducks or man, form universal species. So, universals can be attributes or natural kinds.

Aristotle maintains a realism here, because these attributes and species do actually exist, mind-independent, that is, there are similar attributes and there are similar things in the world. Those things which can rightly be said to be a certain universal kind DO in fact have common essential properties or functions, and those individual properties and functions of these things ARE in fact similar, though not the same. What makes the similarity are the common essential features. I must note here the inherent problem of infinite essences, since properties would need to have similar features, of which would require similar features themselves, so when do we really intelligibly know the universal if the requisite is infinitely analyzable?

For Aristotle the universals are "found" in things, or he would say that they are intuited in a kind of inductive logical construction from our common experiences with various particulars. They are understood out of experience. Plato apprehended universals by way of sensed objects, but the universal did not require sensory experience for its formation. For Plato I am reminded of the universal when sensing a particular which adequately exemplifies that universal. But for Aristotle I must first experience many similar objects or actions having common characteristics before I can adequately classify them as similar and give them the same name. Thus, for Aristotle the universal is as much formulated by a process of the intellect (or induction), as is apprehended (or intuited).

Natural kinds are apprehended as they really are, in the sense that the common elements are in fact relevantly perceived as similar, so that there are things of various kinds. Yet, natural kinds are conceptually constructed by the intellect, in the sense that common elements are logically and relevantly organized together, or we could say that things are classified into conceptual kinds according to essential, necessary, common features. Aristotle is certainly not sensing natural kinds, anymore than Plato, for the kinds are only known intelligibly or through conceptual constructs. Though the properties are sensed, yet not without some kind of intelligible recognition. While Plato is sort of a rational realist, Aristotle is a sort of empirical realist. Universals, for Aristotle, are not real as in "out there somewhere", but concepts, though ideally referring to real, "out there" substances.

BRITISH CONCEPTUALISTS AND NOMINALISTS

The conceptualist theory of universals is usually contrasted with a realist theory. The realist universal is thought of as mind-independent and as existing prior to any private awareness. Obviously though, the universal itself is a concept, but the realist says this universal can refer adequately to real essences of things and real properties, while a radical conceptualist would doubt this as even possible. The conceptualist maintains that universals are not public givens, that universals are private constructs vs. universally accessible facts. The radical conceptualist would argue that there is no necessary correspondence between the conceptual universal and the actual world.

The realist, or Aristotle, might say that a necessary essence, or group of essences, is present in all things designated as of the same kind. But this means that essential characteristics are `a priori' present, which are then recognized conceptually to correspond to that certain kind of thing that it is. Thus, Aristotle is as much a conceptualist as a realist. Whether the universal is in the mind or out there depends upon what we mean by the question or how the answer is conveyed. The essential characteristics, or universals, are partly discovered and partly made by the intellect and language. Maybe this doesn't help explain the relation between "discovered" and "made", and Aristotle could not really say more than that an intuitive induction or Nous was at work.

Still, a mystery remains as to how general concepts are formed (or "found") from the data of particular experiences. This was a concern for the British empiricists. Locke seemed to bring up the questions, and he wavered between conceptualism and nominalism. The conceptualist position is that there are general or abstract concepts "in the mind" which correspond to the universal words we use. These concepts could adequately refer to real things, properties or similarities/differences in the world, which is basically Aristotelian. Or, from a radical position, these concepts have no necessary correspondence (or no possible one), but one would have to question here what exactly makes certain concepts useful or relevant or non-arbitrary. Even if the answer is pragmatic, there still must be some real correspondence to experience.

Locke seemed to hold a kind of representational theory of conceptualism, which is that abstract images or concepts in the mind represent classes of things. He said that we form general ideas by a process of abstraction, separating out all that is not common to all those things named. This might be a kind of image but not necessarily one ever experienced. It might be a rough outline of a man with no distinct features, except skeletal ones, to represent all men. Or it might be a pond with some indefinite number of fish to represent all fishing ponds. Berkeley wondered how there could be such a mental entity as general or abstract images, or even abstract concepts. But Locke elsewhere suggested that certain particular impressions, such as the whiteness of snow or of milk, might become the mental representative of universal qualities.

Berkeley maintained that particular ideas could become general ideas when they somehow represent a set of particular ideas of the same sort. Locke suggested basically the same, in Book III, that some particular ideas acquire a general nature by the capacity they are put into for representing many particulars. If this means that one particular image or impression is always the representative, or always accompanies any use of a general idea, then it seems to be a false notion, since a general idea, say of justice, or beauty, or white, or dog, may conjure up any number of particular memories or images at different times. Yet, there seems to be value in having certain distinct impressions or archetypal images as representing general ideas or classes of similar things. This seems less useful with virtues, since I do not rigidly use one image of beauty or of love or of courage to represent those virtues or act as the norm in comparisons. A general definition would seem to be more useful, but even this is doubtfully useful all the time.

Hume also thought that general ideas were particular ideas in a representative capacity. But he argued that a particular idea (or impression) was able to represent many others by being directly associated with a general word. Thus, Hume tried to explain how this psychological process worked, that is, how particular ideas could be representative: the general term can be associated with a set of particular ideas - the word has the capacity to stimulate any of those particular ideas and vice versa. The general term will call to mind one associated particular idea at a time, and this one idea will then represent the others, but any of the other particular ideas could just as well be the mind's representative at that time.

There is, then, a definite set of particular ideas, any of which could come to mind, given the stimulus of the general word. What holds them all together in one set is the associative power (or dispositional power) of the general term. This kind of behavioristic explanation, or nominalism, that particular ideas are linked together by their each being associated with the same general word naming them, does not explain how they all got named the same in the first place, or how they all were seen to resemble one another enough to be considered as that general kind.

Nominalism can be of an extreme form or a moderate form. The extreme position is that there is nothing in common to a class of particulars called by the same name (other than the same name), but then we might question how such an arbitrary use of language could work for us. We do need some kind of correspondence or reason for making classifications, whether these classifications are concepts or mere words. Nominalism also eliminates any need for mental concepts interposing between words and what they supposedly refer to -- but the nominalist cannot just claim that general words correspond to real particulars - this would be an ultra-realist position. Extreme nominalism is radical anti-realism, maintaining that only words are general, that the name is all that the different things that we call by the same name have in common. The name is arbitrary in that it is not based upon a reality outside the mind. Only names are universal and all things of the world are uniquely individual. What is named (as a generality) does not actually exist.

Again, what does this mean? Aristotle could agree that all substances are unique in matter, but classifiable into universal forms or names. Real properties are present in things, but it is not the same material property. It is the same formal property which is not sensed but intellectually recognized.

One could argue that objects and properties are not divided by nature into classes, that minds distinguish things in these ways for practical reasons, that minds decide what is sufficiently different or similar in objects. This would not deny some outward correspondence and realistic reason for the mind's choice of classification, but the classifications or universal concepts could change according to different purposes, or various objects could be classified in any number of ways or into any number of sets. This kind of conceptualism concerning universals, which could be attributed to Locke, is a progressive step out of the naive realism which assumes nature to present itself with pre-fixed classifications or universal sets. It also recognizes the epistemological status of the mind in actively "forming" reality as well as receptively experiencing it.

There may be a defining essence of properties for certain kinds of things, the lack of which would necessitate that it is not of that kind, but we would have no way of knowing this complete set of essential properties, as argued by Locke. All we can know are the limited properties given by experience, and we can only pick out certain of these properties constantly found together to be the defining characteristics.

Let us say that we make the definition that anything which has the properties a, b, c is going to be called X. The defining properties of X is this group of properties, called the nominal essence of X, "nominal" because it is chosen out of many possible choices. We could of chosen other common properties to group together. We do in fact always find these particular properties together when perceiving that impression called X, so in this sense it is a realistic position. But something not called X, not known as gold for instance, could also have the properties of a, b, c, but also require the property d in its nominal essence. So one class of things, one universal, requires a, b, c, while a different one requires a, b, c, d. I could then mistaken identities if I overlook certain properties. So, the defining characteristics of a universal might not be enough of a verification for correct application. In our ignorance of the real essence of this kind of thing, we could have easily made an arbitrary collection of properties into a named universal.

So, Locke suggests a moderate realist view, or a moderate nominalism, that classes of objects have some resemblance or similarity. Not that all individual objects of the same name share the same essential properties, but that significant properties of them are similar or resemble one another. Things may resemble each other in many different ways, but we choose only some of these resemblances in our grouping of things together. Nature doesn't provide us with these classifications or grouping of resemblances; the mind arranges many simple impressions into a few complex general ideas. The principles of classification are mind-made, not nature-made. Here, Aristotle would argue that the mind's principle, which is logical, reflects nature's principle, which is onto-logical. But Locke denies the Aristotelian theory that objects could share a common property, such as whiteness or triangularness. At the most, these objects exhibit properties which "resemble" one another in the mind or to perception. But what does this mean? They either all resemble some normative form or share something in common.

The question is in what way are they similar, or how do they resemble one another. If they are similar "in some respect", that "respect" must be what is common. Two properties resemble one another, or are similar, either because there is commonality in the properties themselves or they resemble some platonic norm. What else? Maybe resemblance itself is the mystical universal quality shared, a quality of relation? And if resemblance is a universal, are there other relational universals? So, this weaker sense of nominalism doesn't really go much further than Aristotle (or it goes back to Plato). It reduces itself to resemblance theory, since names are not just arbitrarily differentiated, but universal words are different according to the differences in what they refer to.

Russell argued that resemblance must be admitted as a relational universal to avoid an infinite regress of analyzing the resemblance between objects. I don't know if this is valid, but since the question of resemblance remains as the last unavoidable problem in all arguments concerning universals, we might just have to just assume the universality of resemblance. If there were no resemblances whatsoever there could be no coherent language or speaking about experience. Some sort of resemblance remains in any theory, except extreme nominalism which is extreme arbitrariness. I suggest two lines of argument for universals. One is that realistically there are resemblances between objects (or properties) in the world; and the other is that we need to perceive resemblances in the world in order to describe our experience with a limited number of useable words, meaning that there is a conceptual and communicational necessity for universals to represent classes of things.

WITTENSTEIN'S ANTI-UNIVERSALISM

Wittgenstein had a different kind of solution to the problem of universals. But first, a brief introduction to his latter works. In his earlier work of the TRACTATUS Wittgenstein claimed that language stated facts. It consisted of sentences containing word/names which pictured objects of experience. The meaning of these words were derived from the objects they stood for. Truth was correctly picturing the facts of experience, and the structure of language was as fixed and definite as mathematical logic. In his later BLUEBOOK Wittgenstein rejected these assumptions. Now he claimed that language had many pragmatic uses, other than just description. There is no ideal language, nor can be. Instead, there are a multitude of "language games", each connected with various activities or needs of expression, and each having their own particular rules. Words are like tools, each having the capacity to be used in many different ways, and in order to know the meaning of a word we must first be acquainted with the specific language game in which it is being used.

The understanding of words, or signs, is not contained in just those signs or in other, hidden signs or definitions. The understanding of words cannot be a mere translation of words into other words, or the translation of words into definitions. The signs form the inorganic part of meaning, while the organic part, or life of the signs, is in our actual use of them with other signs in actual communication. Neither is the meaning some occult conception or non-linguistic thought within the mind. Neither is it a picture or representation of reality, since all of these can be interpreted in multiple ways. Thus, the only reliable sort of meaning is in how we use the words within the living context of the question or purpose of speech. This living context is the language game. A term or expression finds its use in a particular language game. The meaning of a word is in its use in that language game. When we compare language games we find the different kinds of word-use. One must first master the language game before having the capacity to understand how words are used within it, or what the words mean. The circumstances of its use must be known before knowing the meaning of the word used.

The meaning of a general term can only be fully known by how it is used within the rules of a specific language game; it cannot be independently known. The meaning is not hidden from view in some museum of definitions, nor mysteriously locked up in a hidden process of thinking. The meaning lies on the surface context of its use. Neither are the rules of the games hidden, but are learned through the activity of hearing and speaking. The rules cannot be determined `a priori', but can only be known through familiarity with the language at work. Communicational and philosophical confusions arise from confusing one language game with another, or interpreting sentences by rules of a wrong language game. And the work of the philosopher is much to do with discovering the rules of the various games and bringing linguistic confusions out in the open.

Now about universals, Wittgenstein says that our "craving for generality is the resultant of a number of tendencies connected with particular philosophical confusions." First, there is "The tendency to look for something in common to all the entities which we commonly subsume under a general term," as though such common properties were like ingredients of things, as though beauty were an ingredient of all beautiful things. "We think there must be something common to all games, say, and this common property is the justification for applying the general term `game' to the various games." He claims that games form a family the members of which have family likenesses.. and these likenesses overlap (BLUEBOOK 17)." This is Wittgenstein's rebuttal to Plato and Aristotle.

Secondly, there is a tendency to think that the understanding of a general term, say of leaf, is "a kind of general picture of a leaf, as opposed to pictures of particular leafs," that the general idea, or meaning of the word, is like an image only containing what is common to all leafs. This is Wittgenstein's rebuttal to the British empiricists.

Thirdly, we tend to think that the meaning is found in some mental state, or that we can be conscious of this intentional object called meaning, as opposed to possessing no such `thing' in consciousness. Instead, the "meaning" would be in the use, or found on the surface of the behavioral activity of its expression. This is Wittgenstein's rebuttal to the conceptualists.

Fourthly, our craving for generality has its source in our scientific preoccupation with the "method of reducing the explanation of natural phenomena to the smallest possible number of primitive natural laws," and unifying topics by using generalizations. This seems to be Wittgenstein's rebuttal to phenomenologists. Wittgenstein sees metaphysics as a philosophical mistake to ask and answer questions by this method (BLUEBOOK 18).

Wittgenstein argued that there is not always something common to all instances covered by a general word, and that even if there are some resemblances between recognized instances these are more like "family resemblances", where something is common to some of the instances, while some other things are common to other instances, but there is no one thing in common to all, although there may be plenty of overlapping commonalities.

What is it, then, that justifies the use of a general word in each case? Let's consider expectation. What "makes" us use this word, or what conditions justify its use as an expression in a sentence? Is there a general rule of application? Wittgenstein says no. If we look at many of the different cases of applying this word, we will find no set of common features. He denies the essentialist claim that meaning is tied to a necessary set of right conditions.

Each time we use the word we might see different conditions. Nothing necessarily remains the same. The conditions belonging to my expectation at one time are not the same as those belonging to another time. So the meaning of the word cannot be determined by the conditions surrounding it, or hidden within it. I might be expecting someone to tea today, while yesterday I was expecting my paycheck. Obviously "someone to tea" is different than a "paycheck". But dear Wittgenstein, why do you think the necessary conditions are things or events? I use the word `expectation' in reference to either a belief about the future, in the sense of "believing" someone is coming by or my paycheck is arriving, or in reference to a state of mind/body, in the sense of being in a state of anxious anticipation or the "waiting of a cat".

Yet, Wittgenstein might claim that even if a definition could be made and defended, this could be meant or understood in a variety of ways. The definition itself could have multiple or vague meaning, and its understanding could have multiple interpretations. So, Wittgenstein could argue that not only are there no necessary common essences, but there is nothing about language that could compel us to understand them correctly if there were.

There are no necessary and sufficient conditions for the right application of the word. There is no `meaning' like a blue-print or right definition, which can show us the right use. For Wittgenstein, to think that we need to know such a definition or formula is to confuse general terms with mathematical terms which do require strict definition.

First of all, what might justify one case could not possibly justify another, since there might be nothing at all in common between two cases. Second, even if there were necessary conditions, we do not necessarily see all these supposed necessary conditions. Some might be hidden from view, or neglected, and one would doubt that each person using general terms always perceives the presumed set of necessary conditions. This is similar to one of Locke's arguments. Thirdly, even if we were to find common properties or conditions in all the different cases, how would we know these to be sufficient conditions?

Or fourthly, Could any properties be sufficient conditions? We could apply the same word to a number of objects between which there is no common feature. Wittgenstein uses the example of games. And even if they all did have a certain property X in common out of all of their many properties, say this is the property of being manipulated in some way, this would not be sufficient to call any of them by that name, game, because other things or instances, other than games, also share this property X, that of being manipulated in some way. And yet, we could say, in a metaphorical way, that all things or instances having the property of being manipulated in some way ARE games of a sort. Which, is really the origin of the confused vagueness in many universals.

Wittgenstein admits of a complicated network of similarities, or family resemblances, which do justify the use of a certain general term. This places him in a moderate realist position and not in the extreme nominalist position. But we cannot grasp the meaning of these general words via one swooping definition or essential similarity, because this itself could be misunderstood, or understood in a variety of ways. So, the understanding is not complete without reference to particular cases of its use. If there is a right meaning, it could only be learned by comparing the many [right] applications. But of course, then we would first need to know which are the right applications, vs. wrong use by family and friends. Then again, if there is no `right' application, communication would be very confusing.

Wittgenstein does admit of a right application of words, but the understanding of this, and how to interpret correctly, depends upon how the term is used in the sentence and upon the context of the sentence itself. We need to look and listen carefully to examples of use, and over time, or over the repetition of resembling uses, we will learn how to use the term. There is nothing to examine, except for the examples of use: no outside formal or form definition of meaning.

But then, as I see it, we need to recognize the resemblances of contextual use, or what is common to different cases of use. Back to the resemblance problem of universals? Wittgenstein might reply that resemblances only occur within the same language game, and since there can be many language games resemblance cannot be the ultimate criteria, though useful. Knowing the language game, or knowing what is being asked, is the ultimate criteria, not resemblance or similarity. It is the language game that determines the kind of meaning meant. So here, Wittgenstein seems to base meaning not on mind-independent reality but on conventional use, so he seems to move toward nominalism. Yet, actual resemblances seem to keep the different language-game-uses from being nominally arbitrary.

The realist holds the premise that there must be some outside justification for the use of such general terms, while the nominalist holds the premise that there can be no common elements. So the realist concludes there to be at least some common element for justification, while the nominalist concludes there to be no outside justification since no common elements. Wittgenstein rejects both premises. He believes the family resemblance is enough justification. So he says there are no necessary or sufficient elements, except family resemblance. And yet resemblances is a generality implying something in common beyond the mere naming.

Wittgenstein rejects the notion that the something in common between particulars is exterior, or in addition, to those particulars. He neatly rejects generalities constituted by [other] common elements. What is common to all games is on equal status as what is common to a color, say blue. What can you say is common to all blues? Of course this may be a physiological puzzle. What is common to all bachelors? That they are unmarried? But is this a common property underneath being unmarried? Or is it a mere tautology or synonym? What else could you say? Yet, as I see it, we now have three kinds of examples. The term `bachelor' can only be described analytically or in a simple definition. The term `blue' can only be described according to the division necessarily made physiologically, or in terms of a "mental state." And the term `game' can be described both analytically and empirically as those things or activities which can be manipulated or "played with."

If all that Wittgenstein can say of games is that they all share in being games, then this seems to be the extreme nominalist position that there is nothing in common except that they are all called games, unless he means [which he doesn't want to] that they all share the essential properties of gamehood. What else could it mean that all games share in being games, except an obvious tautology. I suppose Wittgenstein could reply that they all share a common use in language, but this seems too simple an answer. If I only said that all tables have in common their being tables you might ask why I did not state that they all have a common [Aristotelian] functional use which is to serve as objects for placing things on, just as chairs are objects to be seated in, or books are objects intended to be read, or games are things or circumstances to be played. And why not look at the functional use, or the pragmatic use that "makes a difference", as the significant kind of essential element?

Let's consider games as an example. If I learn that `this' is a game, how can I recognize other games latter on? Is there anything that can be said for guidance? It seems that I would need to know the essence of all games, or the respect in which all games are alike. How can I recognize games in my experience if there is no essential criteria for all games or if there is no resemblances between all games? But of course there is no essential resemblance between all games. They often look different and they are played differently. The problem here is in looking at the games themselves for resemblances or essential properties.

What IS common to all games is that one plays them. At least that is the sense in which one should use the word. If it is not something to be played, whether it be monopoly, poker, business, or life, then it is not a game. But then we are confronted with the problematic meaning of the word `play'. To play at something, or with something, is to be in some degree of command over the choice of strategy in a game with different possible moves. Granted, one could use the word `game' or `play' without this kind of stated meaning, but then one is using the words incorrectly or carelessly or obscurely, even if it is used metaphorically. In fact, the metaphorical use of words is often the beginning of obscurity and vagueness. And we cannot not deny that some words are used in an overly vague or general way, or without any thought to their right (or even metaphorical) meaning.

Wittgenstein uses vague terms, such as games and family, for his examples. We must admit that any language can have some vague, overly general terms, due to expanded and metaphorical uses over time. Yet, these are just the kind of terms which are interesting in this study. I could claim here that Wittgenstein's examples are just those general terms which do not have as much resemblance among their particulars as other examples might, say of white or stars or trees or computers or beer or even tables. But to argue that his examples are just those without much resemblance or those used in vague, unclear ways seems to easy of a rebuttal.

To give him deserved credit, Wittgenstein is saying something significant about what could be going on with such vague terms. It is not that there is no meaning to these terms, or that they have no use in communication, but that the meaning is shifting according to use. The meaning of such terms cannot be simply defined for all cases of use. Even if there were a simple definition, or an essential meaning, which I believe there is, this is not enough to understand its use, and not enough to use it correctly in any case, because the definition of an overly vague term is overly simple, and we need more information, so to speak. In accordance with Wittgenstein, we need to know how the term is being used, we need to know its linguistic contextual framework.

I think that language is too complex to state general theories made by realists, nominalists, or Wittgenstein. Some universals can be justified by common properties, others justified by common functional uses, and others can only clarified by the use-context. Of course, Wittgenstein's arguments need to be looked at more carefully, and with more time, but my intention here has been to present his radical solution to the problem, which undercuts the main theories before him.