Tag Archives: Baltzly

The Stoic theory of universals, as compared to Platonic and Aristotelian theories

By Tim Harding

The philosophical problem of universals has endured since ancient times, and can have metaphysical or epistemic connotations, depending upon the philosopher in question.  I intend to show in this essay that both Plato’s and the Stoics’ theories of universals were not only derived from, but were ‘in the grip’ of their epistemological and metaphysical philosophies respectively; and were thus vulnerable to methodological criticism.  I propose to first outline the three alternative theories of Plato, Aristotle and the Stoics; and then to suggest that Aristotle’s theory, whilst developed as a criticism of Plato’s theory, stands more robustly on its own merits.

According to the Oxford Companion to Philosophy, particulars are instances of universals, as a particular apple is an instance of the universal known as ‘apple’.  (An implication of a particular is that it can only be in one place at any one time, which presents a kind of paradox that will be discussed later in this essay).   Even the definition of the ‘problem of universals’ is somewhat disputed by philosophers, but the problem generally is about whether universals exist, and if so what is their nature and relationship to particulars (Honderich 1995: 646, 887).

Philosophers such as Plato and Aristotle who hold that universals exist are known as ‘realists’, although they have differences about the ontological relationships between universals and particulars, as discussed in this essay.  Those who deny the existence of universals are known as ‘nominalists’.  According to Long and Sedley (1987:181), the Stoics were a type of nominalist known as ‘conceptualists’, as I shall discuss later.

Plato’s theory of universals (although he does not actually use this term) stems from his theory of knowledge.  Indeed, it is difficult to separate Plato’s ontology from his epistemology (Copleston 1962: 142).  In his Socratic dialogue Timaeus, Plato draws a distinction between permanent knowledge gained by reason and temporary opinion gained from the senses.

That which is apprehended by intelligence and reason is always in the same state; but that which is conceived by opinion with the help of sensation and without reason, is always in a process of becoming and perishing and never really is (Plato Timaeus 28a).

According to Copleston (1962: 143-146), this argument is part of Plato’s challenge to Protagoras’ theory that knowledge is sense-perception.  Plato argues that sense-perception on its own is not knowledge.  Truth is derived from the mind’s reflection and judgement, rather than from bare sensations.  To give an example of what Plato means, we may have a bare sensation of two white surfaces, but in order to judge the similarity of the two sensations, the mind’s activity is required.

Plato argues that true knowledge must be infallible, unchanging and of what is real, rather than merely of what is perceived.  He thinks that the individual objects of sense-perception, or particulars, cannot meet the criteria for knowledge because they are always in a state of flux and indefinite in number (Copleston 1962: 149).  So what knowledge does meet Plato’s criteria?  The answer to this question leads us to the category of universals.  Copleston gives the example of the judgement ‘The Athenian Constitution is good’.  The Constitution itself is open to change, for better or worse, but what is stable in this judgement is the universal quality of goodness.  Hence, within Plato’s epistemological framework, true knowledge is knowledge of the universal rather than the particular (Copleston 1962: 150).

We now proceed from Plato’s epistemology to his ontology of universals and particulars.  In terms of his third criterion of true knowledge being what is real rather than perceived, the essence of Plato’s Forms is that each true universal concept corresponds to an objective reality (Copleston 1962: 151).  The universal is what is real, and particulars are copies or instances of the Form.  For example, particulars such as beautiful things are instances of the universal or Form of Beauty.

…nothing makes a thing beautiful but the presence and participation of beauty in whatever way or manner obtained; for as to the manner I am uncertain, but I stoutly contend that by beauty all beautiful things become beautiful (Plato Phaedo, 653).

Baltzly (2106: F5.2-6) puts the general structure of Plato’s argument this way:

What we understand when we understand what justice, beauty, or generally F-ness are, doesn’t ever change.

But the sensible F particulars that exhibit these features are always changing.

So there must be a non-sensible universal – the Form of F-ness – that we understand when we achieve episteme (true knowledge).

Plato’s explanation for where this knowledge of Forms comes from, if not from sense-perceptions, is our existence as unembodied souls prior to this life (Baltzly 2106: F5.2-6).  To me, this explanation sounds like a ‘retrofit’ to solve a consequential problem with Plato’s theory and is a methodological weakness of his account.

Turning now to Aristotle’s theory, whilst he shared Plato’s realism about the existence of universals, he had some fundamental differences about their ontological relationship to particulars.  In terms of Baltzly’s abovementioned description of Plato’s general argument, Plato thought that the universal, F-ness, could exist even if there were no F particulars.  In direct contrast, Aristotle held that there cannot be a universal, F-ness, unless there are some particulars that are F.  For example, Aristotle thought that the existence of the universal ‘humanity’ depends on there being actual instances of particular human beings (Baltzly 2106: F5.2-8).

As for the reality of universals, Aristotle agreed with Plato that the universal is the object of science.  For instance, the scientist is not concerned with discovering knowledge about particular pieces of gold, but with the essence or properties of gold as a universal.  It follows that if the universal is not real, if it has no objective reality, there is no scientific knowledge.  By Modus Tollens, there is scientific knowledge, and if scientific knowledge is knowledge of reality; then to be consistent, the universal must also be real (Copleston 1962: 301-302).  (Whilst it is outside the scope of this essay to discuss whether scientific knowledge describes reality, to deny that there is any scientific knowledge would have major implications for epistemic coherence).

This is not to say that universals have ‘substance’, meaning that they consist of matter and form.  Aristotle maintains that only particulars have substance, and that universals exist as properties of particulars (Russell 1961: 176).  Russell quotes Aristotle as saying:

It seems impossible that any universal term should be the name of a substance. For…the substance of each thing is that which is peculiar to it, which does not belong to anything else; but the universal is common, since that is called universal which is such as to belong to more than one thing.

In other words, Aristotle thinks that a universal cannot exist by itself, but only in particular things.  Russell attempts to illustrate Aristotle’s position using a football analogy.  The game of football (a universal) cannot exist without football players (particulars); but the football players would still exist even if they never actually played football (Russell 1961: 176).

In almost complete contrast to both Plato and Aristotle, the Stoics denied the existence of universals, regarding them as concepts or mere figments of the rational mind.  In this way, the Stoics anticipated the conceptualism of the British empirical philosophers, such as Locke (Long and Sedley 1987:181).

The Stoic position is complicated by their being on the one hand materialists, and on the other holding a belief that there are non-existent things which ‘subsist’, such as incorporeal things like time and fictional entities such as a Centaur.  Their ontological hierarchy starts with the notion of a ‘something’, which they thought of as a proper subject of thought and discourse, whether or not it exists.  ‘Somethings’ can be subdivided into material bodies or corporeals, which exist; and incorporeals and things that are neither corporeal or incorporeal such as fictional entities, which subsist (Long and Sedley 1987:163-164).  Long and Sedley (1987:164) provide colourful examples of the distinction between existing and subsisting by saying:

There’s such a thing as a rainbow, and such a character as Mickey Mouse, but they don’t actually exist.

A significant exclusion from the Stoic ontological hierarchy is universals.  Despite the subsistence of a fictional character like Mickey Mouse, the universal man neither exists nor subsists, which is a curious inconsistency.  Stoic universals are dubbed by the neo-Platonist philosopher Simplicius (Long and Sedley 1987:180) as ‘not somethings’:

(2) One must also take into account the usage of the Stoics about generically qualified things—how according to them cases are expressed, how in their school universals are called ‘not-somethings’ and how their ignorance of the fact that not every substance signifies a ‘this Something’ gives rise to the Not-someone sophism, which relies on the form of expression.

Long and Sedley (1987:164) surmise from this analysis that for the Stoics, to be a ‘something’ is to be a particular, whether existent or subsistent.  Stoic ontology is occupied exclusively by particulars without universals.  In this way, universals are relegated to a metaphysical limbo, as far as the Stoics are concerned.  Nevertheless, they recognise the concept of universals as being not just a linguistic convenience but as useful conceptions or ways of thinking.  For this reason, Long and Sedley (1987:181-182) classify the Stoic position on universals as ‘conceptualist’, rather than simply nominalist.  (Nominalists think of universals simply as names for things that particulars have in common).  In a separate paper, Sedley (1985: 89) makes the distinction between nominalism and conceptualism using the following example:

After all the universal man is not identical with my generic thought of man; he is what I am thinking about when I have that thought.

One of the implications of a particular is that it can only be in one place at any one time, which gives rise to what was referred to above by Simplicius as the ‘Not-someone sophism’.  Sedley (1985: 87-88) paraphrases this sophism in the following terms:

If you make the mistake of hypostatizing the universal man into a Platonic abstract individual-if, in other words you regard him as ‘someone’-you will be unable to resist the following evidently  fallacious syllogism.  ‘If someone  is in Athens, he is not in Megara.  But man is in Athens. Therefore man is not in Megara.’ The improper step  here is clearly  the substitution of ‘man’ in the minor premiss for ‘someone’ in the major premiss. But it can be remedied only by the denial that the  universal man  is ‘someone’.  Therefore the universal man is not-someone.

Baltlzly (2016: F5.2-15) makes that point that the same argument would serve to show that time is a not-something, yet the Stoics inconsistently accept that time subsists as an incorporeal something.

I have attempted to show above that Plato and the Stoics are locked into their theories about universals as a result of their prior philosophical positions.  Although to argue otherwise could make them vulnerable to criticisms of inconsistency, they at the same time have methodological weaknesses that place them on shakier ground than Aristotelian realism.  However, I am also of the view that apart from these methodological issues, Aristotelian Realism is substantively a better theory than Platonic Realism or Stoic Conceptualism or Nominalism.  In coming to this view, I have relied mainly on the work of the late Australian Philosophy Professor David Armstrong.

Armstrong argues that there are universals which exist independently of the classifying mind.  No universal is found except as either a property of a particular or as a relation between particulars.  He thus rejects both Platonic Realism and all varieties of Nominalism (Armstrong 1978: xiii).

Armstrong describes Aristotelian Realism as allowing that particulars have properties and that two different particulars may have the very same property.  However, Aristotelian Realism rejects any transcendent account of properties, that is, an account claiming that universals exist separated from particulars (Armstrong 1975: 146).  Armstrong argues that we cannot give an account of universality in terms of particularity, as the various types of Nominalism attempt to do.  Nor can we give an account of particulars in terms universals, as the Platonic Realists do.  He believes that ‘while universality and particularity cannot be reduced to each other, they are interdependent, so that properties are always properties of a particular, and whatever is a particular is a particular having certain properties’ (Armstrong 1975: 146).

According to Armstrong, what is a genuine property of particulars is to be decided by scientific investigation, rather than simply a linguistic or conceptual classification (Armstrong 1975: 149).  Baltzly (2016: F5.2-18) paraphrases Armstrong’s argument this way:

  1. There are causes and effects in nature.

  2. Whether one event c causes another event e is independent of the classifications we make.

  3. Whether c causes e or not depends on the properties had by the things that figure in the events.

  4. So properties are independent of the classifications that we make and if this is so, then predicate nominalism and conceptualism are false.

Baltzly (2016: F5.2-18, 19) provides an illustration of this argument based on one given by Armstrong (1978: 42-43).  The effect of throwing brick against a window will result from the physical properties of the brick and window, in terms of their relative weight and strength, independently of how we name or classify those properties.  So in this way, I would argue that the properties of particulars, that is universals, are ‘real’ rather than merely ‘figments of the mind’ as the Stoics would say.

As for Platonic Realism, Armstrong argues that if we reject it then we must reject the view that there are any uninstantiated properties (Armstrong 1975: 149); that is, the view that properties are transcendent beings that exist apart from their instances, such as in universals rather than particulars.  He provides an illustration of a hypothetical property of travelling faster than the speed of light.  It is a scientific fact that no such property exists, regardless of our concepts about it (Armstrong 1975: 149).  For this reason, Armstrong upholds ‘scientific realism’ over Platonic Realism, which he thinks is consistent with Aristotelian Realism – a position that I support.

In conclusion, I have attempted to show in this essay that the Aristotelian theory of universals is superior to the equivalent theories of both Plato and the Stoics.  I have argued this in terms of the relative methodologies as well as the substantive arguments.  I would choose the most compelling argument to be that of epistemic coherence regarding scientific knowledge, that is, that the universal is the object of science.  It follows that if the universal is not real, if it has no objective reality, then there is no scientific knowledge.  There is scientific knowledge, and if scientific knowledge is knowledge of reality; then to be consistent, the universal must also be real.


Armstrong, D.M. ‘Towards a Theory of Properties: Work in Progress on the Problem of Universals’ Philosophy, (1975), Vol.50 (192), pp.145-155.

Armstrong, D.M. ‘Nominalism and Realism’ Universals and Scientific Realism Volume 1, (1978) Cambridge: Cambridge University Press.

Baltzly, D. ATS3885: Stoic and Epicurean Philosophy Unit Reader (2016). Clayton: Faculty of Arts, Monash University.

Copleston, F. A History of Philosophy Volume 1: Greece and Rome (1962) New York: Doubleday.

Honderich, T. Oxford Companion to Philosophy (1995) Oxford: Oxford University Press.

Long A. A. and Sedley, D. N. The Hellenistic Philosophers, Volume 1 (1987). Cambridge: Cambridge University Press.

Plato, Phaedo in The Essential Plato trans. Benjamin Jowett, Book-of-the-Month Club (1999).

Plato, Timaeus in The Internet Classics Archive. http://classics.mit.edu//Plato/timaeus.html
Viewed 2 October 2016.

Russell, B. History of Western Philosophy. 2nd edition (1961) London: George Allen & Unwin.

Sedley, D. ‘The Stoic Theory of Universals’ The Southern Journal of Philosophy (1985) Vol. XXIII. Supplement.

Leave a comment

Filed under Essays and talks

Epicurean free will

by Tim Harding

Epicurus’ philosophy of mind is perhaps best explained in terms of Epicurean physics.  Epicurus was a materialist who thinks that the natural world is all that exists, so his physics is a general theory of what exists and its nature, including human bodies and minds (O’Keefe 2010: 11-12).

Epicureans thought that there are only two things that exist per se – atoms and void.  Atoms are the indivisible, most basic particles of matter, which move through void, which is empty space (O’Keefe 2010: 11-12).  Objects as we know them are compounds of atoms, and their various natures are explicable in terms of the different properties or attributes of their constituent atoms (Baltzly 2016: 02-1).

When Epicurus refers to the ‘soul’ he means what we today refer to as the mind, so ‘mind’ is the term I shall use here.  He identifies the mind with a compound of four types of atoms – air, heat, wind and a fourth nameless substance (Long and Sedley 1987: 14C).  Because the mind is composed of atoms, it must be corporeal – only the void is incorporeal (Long and Sedley 1987: 14A).  The mind is a part of the body (located in the chest), responsible for sensation, imagination, emotion and memory (Long and Sedley 1987: 14A, 14B, 15D).  Other functions belong to the ‘spirit’ which provides sensory input to, and carries out the instructions of the mind throughout the body (Long and Sedley 1987: 14B).

According to O’Keefe (2010: 62-63), another Epicurean argument for believing that mind is corporeal is as follows:

Premise 1: The mind moves the body and is moved by the body.

Premise 2: Only bodies can move and be moved by other bodies.

Conclusion: Therefore, the mind is a body.

Long and Sedley (1987:107) identify Epicurus as arguably the first philosopher to recognise what we now know as the philosophical Problem of Free Will.  This problem is if it has been causally necessary we should act as we do, it cannot be up to us, therefore we cannot be morally responsible for our actions (Long and Sedley 1987: 20A).  On the other hand, Epicurus notes that ‘we rebuke, oppose and reform each other as if the responsibility lay also in ourselves’ [Long and Sedley 1987: 20C(2)].

According to Cicero, ‘Epicurus thinks that necessity of fate is avoided by the swerve of atoms’ [Long and Sedley 1987: 20E(2)].  Baltzly explains this ‘atomic swerve’ as atoms moving a minimal distance sideways, apparently for no reason at all, from time to time.  This swerve from their natural downward motion results in atomic collisions (Baltzly 2016: F2.2-14).  Although this swerve is not explicitly mentioned by Epicurus himself, Cicero writes that:

‘Epicurus’ reason for introducing this theory was his fear that, if the atom’s motion was always the result of natural and necessary weight, we would have no freedom, since the mind would be moved in whatever way it was compelled by the movement of atoms’ [Long and Sedley 1987: 20E(3)].

Lucretius presents an argument that the atomic swerve enables free will (Long and Sedley 1987: 20F).  O’Keefe (2010: 74-75) states this argument in the following form:

Premise 1: If the atoms did not swerve, there would not be ‘free will’.

Premise 2: There is free will.

Conclusion: Therefore, atoms swerve.

This argument is logically valid, so if the premises are true the conclusion must be true.  Lucretius spends most of this passage trying to show that Premise 2 is true.  However, even if Premise 2 is true, we do not know that Premise 1 is true.  The atomic swerve introduces a slight element of indeterminacy, but this swerve does not necessarily entail free will, since no mechanism is given to explain the connection between these two concepts.  Indeed, Annas (1991: 87) argues that there is a fundamental problem in thinking of human motivation in terms of only the motion of atoms.  She thinks that occurrence of atomic swerves in ordinary macro-objects has no effect on them (Annas 1991: 96-97).  For this reason, I do not think that the introduction of random atomic swerves solves the Problem of Free Will.

Sedley (1987: 107) agrees that taken in isolation such a solution is ‘notoriously unsatisfactory’.  He offers an alternative explanation in terms of ‘development’ which contributes psychological autonomy and which is distinct from the atoms in a kind of differential or transcendent way (Long and Sedley 1987: 107-18).  In other words, these distinct developments are psychological rather than physical properties of the mind.  In particular, the development of consciousness which is an ‘emergent’ property of complex atomic systems like human beings (Baltzly 2016: F2.2 – 17).

In a later paper, Sedley provides some more detail on what he means by emergent properties:

‘I take Epicurus to be sketching some sort of theory of radically emergent properties.  Matter in certain complex states can, he holds, acquire entirely new , non-physical properties, not governed by the laws of physics’ (Sedley 1988: 323-324).

It is important to note that Sedley is attempting here to make a connection between free will and the atomic swerve.  As Baltzly (2016: F2.2 – 18) puts it, the swerve means that not every motion of the atoms which make up our bodies is determined by those atoms themselves.  Baltzly thinks that the swerve does not introduce an element of randomness or indeterminacy into our free choices:

‘Rather, the swerve leaves a gap where the psychological properties of my soul [mind] can cause something to happen where behaviour of the atoms that make up my soul [mind] leave it open what will happen’ (Baltzly 2016: F2.2 – 18).

My own view is that Sedley and Baltzly provide a plausible explanation of the connection between Epicurus’ atomic swerve and free will.  It is possible that consciousness is an emergent psychological property of the material mind.  Free will could be seen as a manifestation of consciousness.  Whilst we cannot yet fully explain what consciousness is and how is works, there is little doubt that consciousness exists.  If consciousness can exist, then so can free will.  However, where I part company with Sedley is that I find Epicurus’ theory of the atomic swerve unconvincing.  Neither Epicurus nor his followers provide any evidence for the existence of the atomic swerve.  It has been postulated as a kind of ‘retrofit’ in an attempt to solve the problem of free will by introducing an imaginary element of indeterminacy.  I think that Sedley’s idea of emergence could help to explain free will even in the absence of the Epicurean atomic swerve.

I would now like to draw towards a conclusion about Epicurus’ philosophy of mind, by comparing it with the theories of his competitors.  According to O’Keefe (2010: 80-83), these were mainly Carneades (214-129BCE) the head of the skeptical academy; and Chrysippus (c.280-206BCE) the third head of the Stoic school.

The most relevant criticism of Carneades is that positing a motion without a cause, like the atomic swerve, would be beside the point in solving the problem of free will (O’Keefe 2010: 82).  Carneades’ solution is to say that all events, including human actions, have causes   These actions are the result of ‘voluntary motions of the mind’ rather than external causes.  He thinks that there is no reason to posit, in addition, a fundamental indeterminism like the atomic swerve (O’Keefe 2010: 82).  In this way, Carneades was perhaps the forerunner of a compatibilist solution to the problem of free will, allowing both determinism and voluntary choices to co-exist.

Chrysippus criticises Epicurus from the opposite direction.  He shows that causal determinism does not make the future inevitable in a manner that renders action or deliberation futile.  In this way, determinism is compatible with human agency (O’Keefe 2010: 82).

In conclusion, I think that Sedley, Carneades and Chrysippus have pointed the way towards a compatibilist solution to the problem of free will, that does not depend on the dubious Epicurean postulation of the atomic swerve.  I therefore think that their approaches to this problem are more compelling than those of Epicurus.


Annas, J. ‘Epicurus’ Philosophy of Mind’ Companions to Ancient Thought: 2 Psychology, S. Everson, ed. (1991) Cambridge: Cambridge University Press.

Baltzly, D. ATS3885: Stoic and Epicurean Philosophy Unit Reader (2016). Clayton: Faculty of Arts, Monash University.

Long A. A. and Sedley, D. N. The Hellenistic Philosophers, Volume 1 (1987). Cambridge: Cambridge University Press.

O’Keefe, T. Epicureanism. (2010). Berkeley: University of California Press.

Sedley D. ‘Epicurean Anti-Reductionism’ in Jonathan Barnes Mario Mignucci (ed.), Matter and Metaphysics. Bibliopolis 295–327 (1988).

Follow me on Academia.edu

Copyright notice: © All rights reserved. Except for personal use or as permitted under the Australian Copyright Act, no part of this website may be reproduced, stored in a retrieval system, communicated or transmitted in any form or by any means without prior written permission. All inquiries should be made to the copyright owner, Tim Harding at tim.harding@yandoo.com, or as attributed on individual blog posts.

If you find the information on this blog useful, you might like to consider supporting us.

Make a Donation Button

Leave a comment

Filed under Essays and talks