Universals and Generalities
Introduction
One of the most powerful characteristics of human cognition is its capacity for categorising perceptions and thoughts. This stems from our ability to perceive commonalities in the things brought to us by our senses. The first section of this article provides an outline of Bertrand Russell’s discussion of universals and particulars, where he looks at the shared properties of and the relations between physical phenomena. Observing how we use these features, he then considers their nature, dealing with the questions of whether they are independent of the things in which they are embodied and whether they are independent of our conceiving them. The second section of the article is a rather more sketchy outline of a paper by Peter Strawson. He notes that talk of universals and particulars leaves out those generalisations which lie somewhere in between. We make a wide variety of generalisations from particulars, all of which involve finding commonalities. These range from Russell’s shared properties and relations, at the most abstract, through to more concrete forms imbued with much of the character of the particulars from which they are taken. Strawson looks at how we distinguish particulars from generalities: whether it is the case that we can sense particulars but not general things, or that only general things have instances. He looks at the types of nouns we use to express generalities and what can be learnt from their usage, which leads on to an analysis of the characteristics of the instances of generalities. Both sections include my comments on the philosophers’ arguments.
Universals and particulars
Russell (1980: 53) says that we “speak of whatever is given in sensation, or is of the same nature as things given in sensation, as a particular; in contrast, a universal will be anything which may be shared by many particulars, and has those characteristics which […] distinguish justice and whiteness from just acts and white things.” He observes that most words in the dictionary stand for universals and that no sentence can be constructed without at least one universal. Proper nouns refer to particulars, whereas other substantives (nouns and constructions acting as nouns), adjectives, prepositions and verbs refer to universals. Pronouns are ambiguous: although they refer to particulars, the identity of these is only established by the context; similarly the adverb “now” refers to a particular, the present moment, but that moment for ever moves on.
In ordinary life, we are apt to think in terms of particulars. When we hear the sentence “Charles I’s head was cut off”, we think of Charles I, of Charles I’s head and the act of cutting off his head; all these are particulars. We do not dwell on the meaning of the word “head” or the meaning of the word “cut”, and both are universals. Even philosophers, while taking account of universals signified by nouns and adjectives, have often overlooked those signified by verbs and prepositions.. Generally, “adjectives and common nouns express qualities or properties of single things, whereas prepositions and verbs tend to express relations between two or more things” (ibid: 54). The neglect of relations has had a profound effect on philosophy. It led to the belief that propositions are about the properties of single things rather than the relations between two or more of them. It was supposed, since it was assumed that relations as universal entities did not exist or were of no account, that either there is only one thing in the universe, or there are many things which do not interact with each other (since any interaction is, or involves, a relation). The first view, espoused by Spinoza, is called monism, and the second, advocated by Leibniz, where each non-interacting thing is a monad, is called monadism.
Yet, despite the emphasis on properties, relations are if anything more secure. To those who would deny the existence of universals, we cannot strictly prove there are such entities as properties, but we can prove there are relations. Consider the universal whiteness: we might say things are white because they have have the quality of whiteness, but Berkeley and Hume vigorously opposed this view. They denied the existence of “abstract ideas”, instead contending that when we want to think of whiteness, we actually form an image of some particular white thing but take care not to deduce from it anything which is not true of any other white thing. As an account of our mental approach this may well be the case. For example if we want to prove a proposition relating to all triangles, we might draw one or a few triangles and reason about them, while being careful to avoid any features not shared with other triangles. However, as a denial of the existence of universals, it begs the question by what criteria an object is judged to be white or triangular. In order to avoid the universal of whiteness, one might choose a particular white patch and assert that something is white if it has a resemblance to this particular. But this resemblance must be a universal because there are many white things and as such all share this resemblance to the chosen particular. It would be an ineffective strategy to hold that there is a different resemblance between each white object and the referent particular, since, for there to be a common judgement of whiteness, these resemblances must resemble each other. Hence the relation of resemblance must be a universal and thus, having admitted the existence of universals, there is no point in constructing an elaborate argument to deny that properties such as whiteness are universals.
Having established there are such things as universals, it remains to be proved that they are not merely mental phenomena. By which Russell means that universals are independent of being apprehended (through a process of abstracting from perceptions) or of being thought about subsequently. He illustrates this independence with the proposition “Edinburgh is north of London”. Given the meaning of its terms, the proposition does not depend for its truth on its being known by anyone, but rather on the physical facts. These facts do not require anything mental, so the relation “is north of”, which is part of those facts, cannot require anything mental either. Although condensed to a few sentences here, Russell’s argument is compelling: the independence from minds of such universals seems undeniable, since the shared properties or qualities of physical objects and events, and the shared relations between objects and between events, are present whether or not we apprehend them or are capable of their apprehension. Humans, bonobos and aliens may discern quite different relations and shared properties in any given collection of particulars. If physical entities can embody universals then it is in principle possible for a brain to create, store and retrieve content that embodies or represents universals by using just physical states and processes.
The essence of a universal is that it is a shared feature, that is a property or relation, of many individual objects or instances. These instances, which are the individual manifestations, embodiments or exemplars of the universal, have until now been restricted to particulars, things of which we are sensible (either directly or indirectly through suitable instruments) or of the same nature as sensible objects. However, by dropping Russell’s sensory requirement, we can extend the notion of a universal, which is useful when we reflect that shared features may be found not only across particulars but also across universals. In the latter case, we have a universal that is a shared feature of instances which are themselves universals. For example, any triple of things (a particular) is an instance of the number three (a universal) and the number three is an instance of number (another, but wider, universal). Layers of universals (and instances), ever more remote from particulars, are found recursively in the shared properties of properties, shared relations between properties and shared relations between relations. Thus, while every particular is an instance, not every instance is a particular.
Russell argues that a relation (or property) does not exist in space-time (only its instantiation in a particular does): “[i]t is neither in space nor in time, neither material nor mental; yet it is something” (ibid: 56). This peculiar kind of being, Russell says, has led many people to characterise it as mental; however, “[i]n the strict sense, it is not whiteness that is in our mind, but the act of thinking of whiteness” (ibid: 56). Russell argues that conceiving universals as mental is to rob them of their universality, since one person’s thought at one time will be neither the same as their thought at another time nor the same as another person’s thought. He suggests that, for convenience, we should speak of things existing when they are in time, so thoughts, feelings and physical objects exist. Universals, though, are timeless and hence do not exist in this sense; rather we might say that they subsist or have being (personally, I find it easier to use exist for universals too, with the understanding that this denotes a very different sort of existence). The distinction between material and mental phenomena on the one hand and universals on the other does not mean we have go so far as to embrace a Platonic world of Forms, which is in some way separate from and more real than the world of sense. Indeed, by the foregoing discussion, a physical universe, in which entities share certain features, is sufficient for universals to be present. It might also be that without a physical universe to provide an ultimate basis for them, there would be no universals at all; so, whilst a universal may only have universals as instances, these would depend ultimately, at the bottom layer, on the shared features of particulars. If this is so, a physical universe would also be necessary for the presence of universals.
Generalities and their instances
Strawson (2012: 129-141) observes that the distinction between particular and universal leaves out general things which are not universals. All universals are general things but not every general thing is a universal. He counters the suggestion that general things cannot be perceived by the senses, by noting that we smell bacon, hear music, watch cricket. It is only with the things referenced by abstract nouns that this suggestion is plausible: we cannot see hope though we might see a facial expression indicating hope. On the other hand, some particulars cannot be sensed, for example the presence of some elementary particles can only be inferred. Another suggestion is that general things can be in several places at once, whereas particulars cannot, so that gold is in Africa and Australia, but some particulars can be spatially scattered. Similarly, with the idea that only particulars have dates or histories: general things can too, such as the history of dress. Could one instead say that (ibid: 131) “individuals, unlike general things, cannot have instances”? To understand more what is meant by the philosopher’s word instance we could examine the variety of qualifiers used to indicate an individual thing as an instance of a general thing: “a case of”, “an example of”, “a member of”, “a piece of”, “a quantity of”, etc. Using such qualifiers, there are expressions for an instance of an instance: “an example of Smith’s generosity”, “a lock of Smith’s hair”, even “an instance of itself”. So, on the basis of expressions used, it is not correct to say that only general things can have instances; so can non-general things.
It might be objected that the first two of Strawson’s expressions involve general things (generosity and hair), albeit confined to the individual, Smith. But the essence of an instance is its embodiment of a general thing. Yet, if “Smith’s generosity” and “Smith’s hair” are instances, then it must be admitted they are more general instances than the “example of Smith’s generosity” or the “lock of Smith’s hair”. The third expression is trivially true in the sense of “an example of”, or “a piece of” where the piece is the whole, but only true for certain classes as “a member of”. These considerations suggest that the boundary between what constitutes a generality and what constitutes an instance is not entirely clear cut and on occasion rather arbitrary.
Strawson explores three categories of nouns (re-ordered here in increasing generality):
- Articulative nouns: (a) man, (an) apple, (a) cat.
- Partitive nouns: gold, snow, music, water.
- Abstract nouns: redness or red, roundness, wisdom.
Those from (iii) are derived from adjectives and there is considerable latitude as the type of things which may be given as instances of them. “Thus an instance of wisdom may be a man, a remark or an action” (ibid: 132). By contrast there is no latitude for instances of those in (i). There is some latitude for those in (ii): instances of water range from a drop, puddle, splash, torrent to an expanse.
Many instances of the nouns in (iii) can be put in the form “the … of …”, for example: “the wisdom of Socrates”, “the redness of Smith’s face”; these are instances of properties or qualities. Speaking of an instance of φ where φ can be taken from any of the three groups, those from (i) differ in that here φ “can and does by itself function as an indefinite designation of an individual instance” (ibid: 132). This not the case for nouns from (ii) or (iii): “gold is not the same as a piece of gold”, wisdom is not the same a wise remark.
On what constitutes an individual or particular, Strawson remarks that the idea of an individual is that of an individual instance of something (usually general): “there is no such thing as a pure particular” (ibid: 132). Further, as an instance of φ, it can be distinguished from other instances of φ and it can be identified as the same instance at different times and places.
Strawson says it might seem that the difficulty attached to explaining the notion of an individual instance arises from the nature of the category distinction between general and individual; the distinction is so fundamental that there is no simpler idea in terms of which it could be analysed (ibid: 133). He challenges this view, citing what he calls “feature-placing” statements such as: “it has been raining”, “snow is falling”, “there is gold here”, “music can be heard in the distance”, which do not mention individual instances or presuppose statements which do mention them (ibid: 134). Yet the feature-placing statements contain the materials for instantiation. Compare the statement “there is gold here” with the phrase giving the instance “this piece of gold”: the former is a complete sentence, whereas the latter by itself is incomplete, yet the phrase implies what the sentence states explicitly. The instance denoted by the phrase is a complex which incorporates both the feature and its placing.
Regarding the relation of instances to the general, especially on the distinctiveness and identity of instances with their encapsulation of both placement and general features, we might develop Strawson’s ideas as follows. A general object possesses one or more properties, qualities and relations. An instance of a general object is one which has all the properties, qualities and relations of the general object plus some extra ones, which in combination and in the context, “place” it sufficiently as a distinct individual. A particular, since it exists in space-time, is also placed literally by the location of the space it occupies at a given time. The extra properties of an instance may also serve to enable identification of that instance at different times. The fewer the properties characterising the general object, the more varied are its instances. This is especially so where the general object is a single property or relation, in other words a simple universal. Conversely, the more properties characterising a general object, the more its instances are alike, since they have more in common.
Here, it is instructive to turn around Strawson’s observation that the most abstract generalities have many and varied instances, whereas those of more limited scope have fewer. We make a wide range of generalisations from particulars, all of which involve finding commonalities. The more numerous and varied the particulars, the fewer and more abstract the features they have in common: in other words, the more these features approach the properties and relations described by Russell. Conversely, the fewer and more alike the particulars, the more numerous the commonalities between them, which in aggregate take on some of the concrete character of those particulars.
As discussed previously, an instance, unlike Russell’s particulars, need not be perceptible by the senses, though in many cases both it and the general object of which it is in an instance may be sensed. In saying that a general object may be perceived by the senses, located in different places or existing over a time period, Strawson was aware of the objection that what we are really referencing are particular instances of the general object. The objection misses the point: in referring to the general object we are concentrating on those features which its instances share, not on those features which would isolate any one particular, and these shared features may relate to sense perceptions, locations and times. A general object may be an instance of another, more general object; again, as instance, it has all the features of this more general object as well as extra, distinguishing ones of its own.
The number of properties needed to identify an instance or distinguish one instance from another depend on the context, since that context determines the domain of relevant properties and relations. To specify or describe the instance adequately, many more properties may be required. If we ask what constitutes an instance, then in the case of particulars, properties and relations are, I suggest, not enough. A particular is a part of the material universe, in fact, it is at each moment a unique assemblage of matter or energy that is not shared with any other particular. This part has properties, qualities it shares with some other parts, including the property of being material, and it has relations with other parts, but the piece of material of which it is composed, being exclusive, is not a property.
Conclusion
Strawson points out that there are generalities which are not universals. Generalities lie on a continuum between concreteness and abstraction. Very similar particulars have in common, amongst other things, a general object that rather resembles them, whereas a large and highly varied collection of particulars may only have in common some abstract properties and relations. Turning this around, the less general an object, the less varied are its instances. In the light of this continuum, Russell’s assertion, that apart from proper nouns most words refer to universals, should be qualified by Strawson’s classification of nouns (articulative, partitive and abstract) covering a broader range of generality. Just as generalities include but are wider than universals, so instances include but are wider than particulars. Instances extend the notion of individual things to allow for universals being instances of a more general universal.
Russell stresses that universals comprise relations as well as properties. He uses the relation of resemblance to make a convincing argument that universals exist independently of the particulars which manifest them. He also argues that universals exist independently of our conceiving them. This leads me to suggest that a physical universe is sufficient for their existence and to speculate that it is necessary too. Yet, as Russell says, universals do not exist in space-time, although their embodiments in particulars do. Universals are timeless, so they have a different sort of existence or being; although not explored here, this has implications for physicalism.
Russell’s characterisation of a particular as “whatever is given in sensation” needs further qualification, given Strawson’s observation that some general things can be sensed: we can see gold or hear music, for example. Although universals exist out of time, some general things have histories, such as the history of dress, which means they change over time. However Strawson, echoing Russell’s argument about universals, denies that talk of generalities is really about their particulars. Universals and generalities exist independently of their particulars: they are not tied to any particular, but are rather what those particulars share.
Tackling the problem of how we distinguish between general things and instances, Strawson argues that it is not the case that only general things have instances: instances can have instances too. The notions of general and instance are not easy to disentangle, since instances carry with them the generalities that they are instances of. Yet, while an instance has all the features of the general, it also has about it aspects which place it as an individual.
An instance is placed by the uniqueness of its combination of properties and relations. A particular, in addition, is uniquely located in space-time. A general object has fewer properties and relations than its instances; a simple universal may just be one property or relation. Although many properties and relations are needed to specify an instance adequately, fewer may be required in a given context to identify it. If an instance is a particular, then its unique combination of properties and relations is not enough, I think, to constitute it, since, at each moment of its existence, its physical composition is a piece of matter or energy that is not shared with any other particular.
Bibliography
Russell, B. (1980 [1912]) ‘The Problems of Philosophy’, Oxford, Oxford University Press, 52-63.
Strawson, P. F. (2012) ‘Particular and General’ in Analytic Philosophy, eds A. P. Martinich and D. Sosa, Chichester, West Sussex, Wiley-Blackwell, 129-141.
© 2025 C P Blundred