ArtsAutosBooksBusinessEducationEntertainmentFamilyFashionFoodGamesGenderHealthHolidaysHomeHubPagesPersonal FinancePetsPoliticsReligionSportsTechnologyTravel

Thin Particulars and Temporal Parts

Updated on January 29, 2012

D.M. Armstrong in A World of States of Affairs postulates the existence of fundamental particulars for many reasons, including to have something to instantiate universals. The problem is that the bare particular cannot exist by itself. We cannot know particulars without their properties. Although, we can abstractly think of a particular, called a thin particular, without its properties. Perhaps for this very reason some people like the bundle theories of particulars more appealing. As to what exactly a thin particular is, Armstrong does not have a satisfactory answer. If it cannot be demonstrated that particular’s exist, then how does this then affect his notion of temporal parts? It can be claimed the thick particular, with its properties, is partially identical to other temporal parts because the thin particular itself does not change, even when the arrangement of properties does.  For example, a green apple that rippens to a red apple is still the same apple, due to its thin particular, no matter the change in its properties, as in color. Even without the somewhat empty notion of particularity that Armstrong uses, we can still have an acceptable answer to the problem of change using Armstrong’s insights.

The Particularity of Particulars

Armstrong clearly recognized the problem of thin particulars, that is, of particulars thought of without their properties. Think of an object. Now strip away every property associated with that object, including form, mass, color, texture, nature. Now that sort of ‘blanknes’ you have left over is the thin particular. Armstrong states there are “epistemological difficulties thought to be involved in postulating particulars which are surplus to their properties and relations.” (p.95, Armstrong). The epistemological problems arise from the fact that particulars cannot be separated from their properties except by abstraction. Therefore, we cannot have any knowledge of what particulars are, except by a vague image of an empty bucket, without the bucket, that properties fill. It appears Armstrong wants to reject the idea that particularity as some sort of essence, since he clearly states, “There is no call to think of haecceity as a unique inner nature or essence possessed by each particular,” (p.109, Armstrong). Armstrong suggests that “It seems plausible to say that the particularity of particulars is given in experience.” (p.95, Armstrong). That is, it is given in our basic perceptions that “seem to take the form ‘This has certain properties’ and/or ‘This has certain relations to that.’” (p.96, Armstrong). Armstrong refers to the fact that visually we can understand an objects properties and its relations to other properties. From this argument he concludes, “Will it not be natural to take ‘this’ and the ‘that’ as referring to particulars, particulars which are not wholly constituted by their properties and relations?” (p. 110-111, Armstrong). Certainly theorizing particulars is very useful to Armstrong’s ontology, but the solution to epistemological argument is a little unsatisfactory given it is not really saying anything about the particularity of particulars.

The epistemological argument he suggests is odd to say the least and suggests particulars are given in our experience and not taken up by their properties and relations. For one thing, at our level of perception things appear quite different from the deeper reality of atoms and electrons. For instance, our basic perception tells us colors appear to be things, often leading to the misconception that they are universals. Armstrong often refers to the atomic level for most of his illustrations, which is not readily available to our perceptions. For example, we could have two electrons orbiting around two atoms. The two electrons, like all electrons, are identical in mass, spin and charge… their properties. We could say there is ‘this’ electron and there is ‘that’ electron, as a way of pointing to some vague particularity to each electron. But as can happen, electrons can switch places and orbit around the other atom instead, and this can happen because they are identical in structure. If the change happens without noticing, then you can still say there is ‘this’ electron and ‘that’ electron, even though you had no idea they had switched. This illustration demonstrates there is no fundamental difference to the observer, if we could observe such a thing, no individual electron essence or noteworthy affects of its particularity, since the electrons can switch places with no effect to the atom arrangement. It does not seem we are pointing to anything unique to an electron when we point to them and say ‘this’ one and ‘that’ one. This and that may simply refer to the electron as a whole and as a distinct bundle of tropes or properties. Such common terms do indeed point to objects of our perception, but not necessarily to its particularity.

If we indulge in a mental abstraction, and take the thin particular from its property and relations, I think we end up with a peculiar entity. If you take two thin particulars and compare them, what is the difference? They have no nature, because their nature is the properties that are instantiated in them. Indeed, “its emptiness of content all too easily translates itself in out thoughts into some unknowable inner constituent.” (p. 110, Armstrong). All that we can say is that “different particulars are numerically different, then we appear to have said all that can be said about the nature of particularity.” (p. 109, Armstrong). If they are the exact same, insofar as they both are empty of properties, then it seems there is nothing particular to each particular at all, save perhaps its location and its relation to others. Yet, why should we support a fundamental dualism of properties and particulars, when we do not need to? It may be far more reasonable to conclude that things are bundles of universals or bundles of properties. Yet, to Armstrong, they are fundamental, necessary and useful but in actuality we cannot say much about what a thin particular in fact is.

Temporal Parts

If we deny the existence of particulars in Armstrong’s ontology, this will affect his notion of change. If we are led to the conclusion that particulars in Armstrong’s ontology are insufficiently proven and we accept that bundles of universals have more problems (as discussed p. 96, Armstrong) then it seems that bundles of tropes are the logical solution. This however, alters how we would view objects persisting through time.

In Armstrong’s view of temporal parts we look at the thick particular that changes over time. What changes are the properties that the thick particular has. Armstrong states “It seems the particular at t1, cannot be (strictly) identical with the particular at t2. The identity can be loose only, a matter of two different temporal parts of the one particular.” (p.100, Armstrong). This is because a thick particular can be red at one time and green at another, and if there was strict identity, it would be both red and green which is impossible. With temporal parts the object is not both red and green, because each is a separate temporal part. These temporal parts are parts of a mereological whole. This is the same with spatial parts, “Where spatial parts of particulars are non-overlapping, the parts in question are wholly mereologically distinct particulars, although they may be nomically bound together.” (p.99, Armstrong).

It is worth mentioning that if one considers the thin particular only, then it appears not to change over time. Armstrong gives an example of an atom. The atom does not appear to change and therefore has strict identity. It appears then that thin particulars, which do not seem to have a nature abstracted from their properties and do not seem to change, have strict identity through time. However, when one considers the thick particular that alters in its properties then it can only be loose identity. Certainly properties are identical regardless of space and time, and so it seems with the thin particular.

For a principle of unity Armstrong thinks what is essential is that there “should be a causal, or at any rate nomic, relation between past and future temporal parts. A continuing thing must grow out of its past.” (p.105, Armstrong). As mentioned above, it does seem like there is unity in the fact that through time the thin particular does not change. What is important with this causal line is the idea that a thing through time is a thing replacing itself. Thus “We can think of it as a temporal part of a thing bringing about, causing, a later part of a thing.” (p. 105, Armstrong). And “If they are acted upon in one way, the line of their future will have a certain nature, if acted upon in another way their future will be systematically different,” (p.106, Armstrong).

If we take away particulars from this model, and refer to the thick particular as bundles of tropes instead, then we have problems with identity. If we have a bundle of tropes that through time changes from red to green, we certainly do not have strict identity, but we also seem to lose a hold on partial identity. This is because we are not referring to different temporal parts of one particular, but many particulars in different configurations and at certain times different components. There can no longer be temporal parts of one particular, because it is the bundle of particulars that is changing. Armstrong would agree that an apple that is red at one time and green at another are different temporal parts. With a bundle theory it seems that they are different as well, but not temporal parts. Once the bundle changes then it is a new bundle and thus a new thing.

It is still feasible if what connects the bundles in time is the same causal line that Armstrong uses. To reiterate when a thing changes its constitution it is bringing about, causing, a later part of that thing. For example, we have bundle (a,b,c) which persists as that bundle for a period of time. When the bundle is acted upon in a certain way it changes to (a,b,d) and then we have a completely different bundle that is causally connected, in the sense Armstrong uses, to the previous bundle. Change conceived of a bundle of particulars changes into a different bundle of particulars entirely, with no guarantee that any particular trope is going to remain constant. Change then is the creation of a new bundle, a new object, causally linked to a previous arrangement of particulars.

With Armstrong when the properties change you get a different temporal part, but the thin particular is strictly identical to all other temporal parts since it at least remains the same, as the whole, the thick particular does not. It is debatable whether Armstrong would agree that thin particulars do not change, since in regards to temporal parts he is referring to the thick particular, and really, the thin particular, is just an abstraction we make. Indeed each particular is numerically different, which is his only defining feature of a particular. I would say a thin particular thought of in abstraction does not change, since it has no parts to change, and therefore like an atom is strictly identical through time. After all the mereological whole is referring to one thick particular, it certainly seems the thin particular remains consistent.


We can see that the particularity of the particular is rather empty. Particulars cannot be known separate from their properties. Even if we speculate about what a thin or bare particular is we can discover nothing except that it makes things numerically different. It also seems that thin particulars might very well be identical with other thin particulars, except in location and relation to each other. Based on what we can postulate about thin particulars we can say that though time they do not change. Or as Armstrong suggests the thick particular through time changes properties, but that these are different temporal parts of the whole. If we are not willing to accept this empty version of thin particulars then we can reasonably accept a trope theory. With bundles of properties we do not have the strict identity through time because of a thin particular, but we still have objects that are causally linked. It is because they are causally linked that we can say that the apple that is now red was the apple that was green, even though they are different bundles of properties. In regards to change we do not need to accept the notion of thin particulars and separate properties.



Armstrong, D.M. A World of States of Affairs. Cambridge University Press, New York, 1977.

See also Philosophy: Exploring Existence- for an overview on Armstrong’s ontology for abstract and concrete objects.


    0 of 8192 characters used
    Post Comment

    • PhiMaths ATB profile image

      PhiMaths ATB 5 years ago from Charlotte, NC, USA

      It's structures all the way down...

      Actually, I'm not sure about this, but I don't think bundle theory or haeceitties are the way forward, leaving me I suppose, with structuralism.

      Also, your example of electrons as well as your statement about the indiscernability of bare particulars bring up the topic of PII and discernability and I was wondering if you were familiar with that literature...?

    • nmalbert profile image

      Nikki Albert 5 years ago from Canada

      No I'm not familiar with that literature... I'm curious though. Where would I find mention of that? I always found those bare particulars particularly odd. :)

      It wasn't an area I was too familiar in mind you, but all metaphysics is interesting to me.

    • PhiMaths ATB profile image

      PhiMaths ATB 5 years ago from Charlotte, NC, USA

      Ladymann, Linnebo, and Pettigrew: Identity and Discernability in Philosophy and Logic, in the review of symbolic logic is a good recent paper, though I'm biased as they all taught me...

      Perhaps I will write something about this myself.

    • nmalbert profile image

      Nikki Albert 5 years ago from Canada

      You should. It would be in interesting topic!

    • PhiMaths ATB profile image

      PhiMaths ATB 5 years ago from Charlotte, NC, USA - A brief overview of discernibility anyway.

    Click to Rate This Article