Joseph Diekemper – Pastism and Events

Pastism and Events – Joseph Diekemper (Queen’s University Belfast)


Elsewhere I have developed a growing block theory of time according to which the only events that exist are past events (I call this theory ‘pastism’). There I argued that the existence (simpliciter) of past events can be accounted for in terms of the existence of thisnesses of past events. Elsewhere I have also developed an ontology of thisness which allows for their employment in this way. On this ontology, thisnesses have a two-step instantiation structure, according to which an abstract thisness property is instantiated by a trope-like, concrete thisness. The thisness instantiations are ‘trope-like’ in virtue of being necessarily unique, concrete, particularised properties; but they are unlike tropes in that they are both essential to their bearers and non-qualitative. In the current paper, I develop a theory of events which coheres with both of the earlier accounts and which also has independent motivation. Pastism requires that an event be identical to its thisness. How is this possible? Well, given that there are plausible trope theories of events extant, I claim that an event which has occurred at a past time t just is a concrete, trope-like instantiation of its abstract thisness property. The thought here is that we characterise an event’s having occurred in terms of a thisness property having been instantiated by its thisness trope. The thisness trope is, in turn, an instantiation of the abstract thisness property; and it is the latter which exists simpliciter. So, on this account, the occurrence of an event is the instantiation of an event thisness property by its thisness trope. Since, according to pastism, all events have occurred and exist simpliciter (inasmuch as there are no present or future events), all event thisness properties have been instantiated by their thisness tropes and exist simpliciter. I proceed by first by summarizing my earlier conclusions before further developing the account. I then go on to consider extant theories of events, in order to evaluate which of these is best suited for adapting and developing in a way which comports well with pastism.


Full paper here.

Because the papers are works in progress, they have been password protected. To register for the conference, and get the password, please contact Graeme A Forbes at


Skip to comment form

    • Nathan Wildman on August 27, 2015 at 11:54 am
    • Reply

    Thanks for this interesting paper. I’m totally unfamiliar with the ontology, but was wondering if you’ll end up with a problematic circle of dependence. Specifically, you say that thisness properties exist ‘simpliciter and tenselessly’, but ‘come into existence in virtue of their thisness modes being instantiated at past times’ (p. 3). Similarly, you identify thisness modes with events, and say that events depend for their identity – and hence also at least partially for their existence (no entity without identity, after all) – upon their thisness properties. So we’ve a circle: events depend upon their thisness properties, and their thisness properties depend upon events. Given that the standard assumption that dependence is anti-reflexive, this is problematic.

    Relatedly, what individuates the thisnesses? I’m not sure that there’s a good answer here. If we say that it’s their instantiation patterns, we’re got an even worse version of the above dependence circle (F*=F* because it is instantiated by e, but e=e because F*=F*!); if we add thisness-thisnesses, we’re off on a regress; and if we make it primitive, it’s unclear how there’s been any gain over e.g. ‘nakedly numerically distinct’ accounts. [Again, this is more a question about the underlying ontology than the paper topic itself, but the ontology is the driving force behind the paper, so thought it worth asking about.]

    More directly, you end by suggesting that ‘we need not claim that the event thisness mode was instantiated in anything, it simply was an instance’ (p. 13). But if we do this, what are we meant to make of the claim that thisness modes are ‘essential to their bearers’ (p. 2)? Put more bluntly, how can something be essential to its bearer if it doesn’t have a bearer?

    1. Thanks for your questions Nathan! My answers (the numbering corresponds to your three paragraphs):

      1) Given the identification of events and their thisnesses, and given that identity is reflexive, the circularity you indicate is benign. The ‘in virtue of’ relation in the quote above should be not be interpreted as ontological dependence, but as claiming that what it is for an event to occur is for its thisness property to be instantiated by its thisness mode.

      So what do events ontologically depend upon? I think events are a fundamental category of ontology, so I do not think they can be ontologically reduced; but I suggest that there is an non-reductive, ontological dependence of events upon concrete phases and the varying states of those phases over time. This is analogous to the dependence that an Aristotelian hylomorphic compound has upon its form and matter. Such a compound cannot be reduced to its hylomorphic constituents, but it certainly ontologically depends upon them.

      2) The individuation of primitive thisness most certainly is primitive! I suggested in the paper that if we had independent reasons for believing in primitive thisness (which I attempt to establish in my 2009 JPhil article), then relying on such entities would be preferable to the seemingly adhoc move of positing primitive determinable tropes.

      3) I probably need to spend more time in the paper explaining the relation between the generic ontology of thisness which I develop in my 2015 PPR article and the employment of that ontology with respect to certain kinds. The general ontology claims that thisnesses are essential to their bearers, but I should probably qualify this claim with ‘should they have bearers,’ since, as I argue in the 2015 article, the general ontology will need to be combined with a specific theory of the metaphysical kind which has it. So, for example, if persons have thisness, then the ontology of thisness must be combined with a specific theory of persons. So, too, for events, material substances, etc. (although I don’t believe material substances have thisness). And the point is that we would expect for the structure of thisness to vary in certain ways depending on the kinds in question. So event thisnesses don’t have bearers and are identical to the entities they individuate, but (presumably) person thisnesses do have bearers and are metaphysically distinct from the entities they individuate.

      Hope these responses help. Thanks again!

    • Emily Paul on August 27, 2015 at 1:59 pm
    • Reply

    Thanks very much for this, Joseph!
    I was just wondering whether you could help me with a few clarifications. I’m wondering about what exactly past times are. If all past events exist simpliciter and not “at” any times, then do past times themselves not exist, and the past only exists in virtue of past events? Linked to this, would you say that we can never talk about anything occurring (as opposed to existing) “at a time,” since no present events exist and past events only exist simpliciter?

    Also, when you talk about times t1-t3, are these past times that you’re talking about, or are they instantaneous present states? I suppose that they need to be past times, and then only after the more recent past time of t3 can we say that the event has occurred. If this is so, I was wondering when it would be true to say that a certain event is happening now. If I said at t2 “the event is occurring now,” would this only be made true after the terminal instant of the event? Or would it never be true, because events only ever exist once they’re completely over? Thanks again!

    1. Thanks for these questions Emily!

      1) I have a relational view of time, so it’s events that determine times and temporal relations. So, yes, the past and past times only exist in virtue of past events. We can talk about events having occurred as of a time (an event e which began at t1 and ended at t3 has occurred as of t3), but not as occurring (present tense) at a time, since there are no present events. Of course, I also include times in the analysis of concrete changes, so there is some circularity here (see my answer 1 to Nathan’s post above); but given that the analysis is non-reductive, and given that the primary constituents in the analysis are the concrete phases and their varying states, I don’t think this is problematic. I should definitely include more discussion of this in the paper though.

      2) Yes, the occurrence of an event can only be spoken of in terms of the past tense, so t1-t3 are past times. It is never true to say that an event is happening now, for the reason you suggest. I attempt to explain away the counter intuitiveness of this in my 2014 Synthese article (p. 1099 ff.).

      Hope these answers help! Thanks again!

        • Emily Paul on August 30, 2015 at 2:38 pm
        • Reply

        Thanks Joseph, they do help, and I’ll definitely re-read your 2014 paper. All the best!

    • David Sanson on August 27, 2015 at 6:22 pm
    • Reply

    Thanks for this paper, Joseph. I’m working to digest it. Rather than providing one extended comment, I am going to provide several shorter comments.

    First, a comment on individuals that are not events, like Socrates.

    If I understand your view, it is that, when Socrates came into existence, at least two things came into existence: an abstract thisness (let’s call it Socrateity), and a concrete instantiation of that thisness, Socrateity*. I’m not sure whether the view is that Socrateity* = Socrates, or the view is that Socrateity* = the principal part of Socrates responsible for his identity. The fact that you call it a concrete mode suggests the latter.

    Then, when Socrates goes out of existence, Socrateity* ceases to exist, but Socrateity continues to exist.

    But if that is right, then it does not seem right to say that, on your view, past concrete objects exist. Instead, on your view, the abstract thisnesses of past concrete objects exist.

    So, then, why call yourself a Pastist rather than a Presentist who thinks that abstract thisnesses, once they come into existence, never cease to exist?

    1. Thanks David for these questions! I’ll respond individually to your posts.

      On this question, first see my response to Nathan’s third question (above). Given that events and persons are different metaphysical kinds, there are going to be variations in the structure of thisness for each. So the *property* of being identical to Socrates (whether we are speaking of the abstract property or the mode) is NOT identical to Socrates; rather, Socrates’ abstract thisness property is instantiated *by* his concrete thisness mode *in* Socrates (so Socrates and his thisness are metaphysically distinct entities)

      In the case of events, however, an event e *is* identical to it’s thisness: as occurrent it was identical to it’s thisness mode, as existent it is (tenseless) identical to its abstract thisness property.

      The presentist claims that an event e exists [present tense] simpliciter just in case e is occurring [present tense]; whereas the pastist claims that an event e exists [tenseless] simpliciter just in case e has occurred [past tense] at t. Thus, the two ontologies cannot be reconciled.

        • David Sanson on August 28, 2015 at 4:54 pm
        • Reply

        Thanks! This helps but I am still confused.

        In the case of Socrates, you want to posit three distinct entities: the abstract thisness ≠ the concrete thisness mode ≠ Socrates. You think that, when Socrates comes into existence, all three of these entites come into existence together. You think that, when Socrates ceases to exist, all that remains is the abstract thisness. So far, this sounds to me like Presentism with abstract thisnesses that continue to exist simpliciter after their instances cease to exist.

        In the case of events, you want to posit two distinct entities: the abstract thisness ≠ the concrete thisness mode. And then you say that as occurrent, the event = the concrete thisness mode, but as existent, the event = the abstract thisness. I assume you think identity is transitive, so what you mean here is that:

        1. to say that event, e, occurred, over a span of time, t, is to say that the concrete thisness mode existed at that span.
        2. to say that event, e, exists is to say that its abstract thisness exists.

        From (2), I infer that really, e = its abstract thisness. A presentist who is willing to suppose that events are abstract thisnesses that need not be instantiated can say that same thing. They will join you in rejecting the claim that events exist just in case they are occurring.

        So I guess the ontological disagreement lies in (1). Here your claim is that the Presentist cannot say that concrete thisness modes ever exist, because the concrete thisness mode only exists at a span after that span has been completed, and so become past.

        Here Presentists do have to face a set of puzzles, about successive entities that never wholly exist at any given moment, whose existence is not guaranteed until the last moment. These puzzles are related to the imperfective paradox: there is no moment at which Socrates dies (at each moment, either he was still alive or already dead), and yet, Socrates is dead.

        Here I share Graeme’s thought that, while it is true that Presentist cannot say that a concrete thisness mode ever wholly exists at a given moment, she can say that it exists, part by part, in succession, over the span. And she can say that the earlier parts only count as parts of the whole in virtue of the later parts, and so, to borrow an example from Aristotle, if I am halfway through building a house, the truth of the claim that I am halfway through building a house might depend on still contingent truths about the future—e.g., that I will complete building a house.

        But I am also unclear on your proposal. You don’t think that the concrete thisness mode of a past event continues to exist. So you think that we can say, after the fact, that it existed over a certain span, and we can say, after the fact, that its abstract thisness exists. But isn’t that exactly what a Presentist, who believes in abstract thisnesses, will say as well?

        1. Some people think that there is a fundamental difference between events and substances: the former are occurrents and the latter are existents. I am embracing the fundamental distinction between the two predicates, without embracing the way they are predicated. On my view, both predicates apply in the case of events, because although events have occurred at a time, they exist tenselessly and at no times (they have occurred at times!). These two different ontological aspects, in the case of an event, are captured by an event’s abstract thisness property and its thisness mode. With these points in mind, I would revise your 1. and 2. in the following way:

          1. to say that an event e occurred over an interval of time t is to say that e’s concrete thisness mode instantiated its abstract thisness property at t
          2. to say that an event e exists is to say that its abstract thisness property exists [tenseless] (in virtue of 1)

          How can a presentist say that there are no present events and that all events exist tenselessly? I don’t see how this theory could possibly be misconstrued with presentism.

    • David Sanson on August 27, 2015 at 6:27 pm
    • Reply

    Second, a small comment on trope individuation. On p. 9 you say that “tropes are normally individuated in terms of either the material objects in which they inhere or the spatiotemporal locations in which they are instantiated.”

    Is that right? Many of the prominent 20th century trope theorists are also bundle theorists, and would, I thought, take the individuation of the tropes to be prior to the individuation of the bundles. Do these theorists then go on to individuate tropes by spatiotemporal location? (I admit that I have not spent a lot of time digging into this literature, so the answer might just be, “yes, David, they do!”) I don’t think they should, since they should allow that a given bundle of tropes (say, this chair), might have been located somewhere else right now (say, if I had moved it into a different room).

    So I guess I found this claim about how tropes are usually individuated surprising. I thought the standard view, among trope bundle theorists, at least, is that tropes are basic individuals.

    1. Maybe the sentence you quote from me was a careless generalization, since there are so many different views on tropes, and I don’t pretend to have investigated them all. But it certainly strikes me that trope theorists *should* individuate tropes in one of the two ways listed. I don’t think individuating in terms of spatiotemporal locations has the undesirable consequence you suggest (though I think it has others), since I think spatiotemporal history can individate material substances across possible worlds. Whereas, if tropes are primitively individuated, then it makes sense to say that the red trope in chair 1 could be swapped with the exactly resembling red trope in chair 2, but this would appear to be an empty possibility. What difference could such a swap make? But shouldn’t it make a difference? Of course, this objection would not apply to thisness modes, because they are not qualitative or phenomenal properties, as are the trope bundle theorist’s tropes.

        • David Sanson on August 28, 2015 at 4:59 pm
        • Reply

        I *think* many trope theorists reject the demand that tropes be individuated in either way, and instead take tropes to be basic individuals. It is true that this leads to distinct possibilities that are qualitatively exactly the same (mere haecceitistic differences). But anyone who accepts basic individuals into their ontology should accept that there are distinct possibilities of this sort (even Lewis accepts such possibilities—he just doesn’t identify them with possible worlds, but instead analyzes them counterpart-theoretically).

    • David Sanson on August 27, 2015 at 6:33 pm
    • Reply

    I don’t think I fully understand what is going on when you say that you want to identify an event with its thisness. Do you mean that the event is identical to its abstract thisness property? Or do you mean that it is identical to the concrete instantiation of that property (the “concrete mode”)?

    In the case of Socrates, if I believed in abstract thisness properties, I would be inclined to say that Socrates is identical the concrete instantiation of his thisness. But, as I mentioned in my first comment, I wasn’t sure whether that is what you want to say about Socrates. Maybe you want to say that the concrete instantiation of the thisness is instead a special part of Socrates—the concrete vehicle of his identity, say, which might be identified with a soul, for example.

    So if that is what you want to say about Socrates, then maybe what you are saying about events is that, for them, there is no distinction between event-as-a-whole and principle-part-of-event-that-serves-as-vehicle-for-its-identity?

    But I suspect that I just have everything completely wrong here.

    1. I think my first response may have answered this question. Let me know if not…

    • David Sanson on August 27, 2015 at 6:58 pm
    • Reply

    Let me see if I can correctly summarize your adaptation of Cleland’s view. The idea is that Cleland thinks that an event is, as you put it, “the time-ordered exemplification of differing determinate properties by the same determinable trope.”

    So, when a candle, C1, changes from straight to bent, for example, the event, e1, is a change in its determinable shape, S1, (understood as a trope rather than a universal), from straight to bent. The proper subject of the change is not C1 but S1.

    If the determinable shape, S1, was understood as a universal rather than a trope, this account would fail, because it would identify events that should be distinct: if two candles, C1 and C2, each with the same determinable shape, S1=S2, change from bent to straight, and e1 is the change, in S1, from straight to bent, and e2 is the change, in S2, from straight to bent, then e1=e2. But it is clear that there are two events here, not one, so the account would get that wrong.

    So we need to understand the shapes of the two candles, S1 and S2, as two distinct tropes, and we need an account of what individuates these tropes, so that we can say that when the C1 and C2 change (or, rather, when S1 and S2 change) there are two events, e1 and e2, not one event.

    (Aside: how does the account handle the case of a candle that changes from straight to bent to straight to bent to straight to bent…. In that case, we have a single determinable shape, S1, and at different times, it changes from straight to bent. It looks like the account handles this by building into each event not just the determinable shape, S1, and the before and after states (straight and bent), but also moments of time. So it looks like you are committed to saying that the same event could not occur at a different time. Is that correct? Or does your ulimate account remove this commitment because of the way it appeals to thisnesses? This relates to Emily’s question and Nathan’s question.)

    Cleland, you say, holds that tropes are basic individuals. But you instead want to account for the distinctness of the events by appealing to thisnesses. You don’t want to replace the trope, S1, with a thisness, because you don’t think thisnesses can change. (I’m not sure why you think this. If I believed in thisnesses, I would be inclined to say that, when Socrates changes (say, from straight to bent), there is a sense in which his thisness also changes. I would take that to be a derivative kind of change. But then, I think that when a candle, C1, changes, the sense in which its determinable shape, S1, changes, is equally derivative. So my head is probably not in the right place here.)

    So instead you propose identifying the event, e1, with the concrete change in S1. (On p. 11 you say that it is the concrete change in e, but I think that is a typo). And you say that F*—the concrete instantiation of e1’s thisness—is identical to the event, e1.

    Insofar as I can get my head into this, I think that sounds like the right thing to say. It should get you around the recurrence worry, and it should allow you to eliminate essential reference to moments of time in your account of the individuation of events.

    Translating down from the thisness talk, it sounds to me like you are simply saying that changes are events, and events, or changes, are basic individuals. Since they are basic individuals, we do not need to assume that they are individuated by which determinable tropes they involve, or by the before and after states they involve, or the interval of time in which they occur.

    I think I agree that a good theory of events should treat events as basic individuals in this way, even if I disagree with your further analysis of basic individuals in terms of abstract thisnesses.

    1. 1) Up to your ‘aside’, I think this is pretty close. The change that is an event has to be a concrete change if events are concrete, unrepeatable entities; and if the fundamental constituent of that change is a determinable property, then it needs to be a concrete particular (i.e. a trope).

      2) On your ‘aside’: I am 100% committed to the claim that events are unrepeatable, in spite of the differences between Cleland’s account and my adaptation. That is what my Assumption 1 amounted to.

      3) Thisnesses do not change because truths about identity do not change. But in the paper I never considered replacing the determinable trope with the thisness mode (that would definitely be a non starter). What I considered was the possibility that the thisness mode inhered in the determinable trope. I rejected this because it seemed to require the thisness mode to exist presently, which violates the pastist ontology.

      4) Sorry, by ‘concrete change in e’ I meant the concrete change that was e, not that there was a change in e; you’re absolutely right that the change was in the determinable trope. I need to make that less misleading.

      5) F* *was* identical to e1 (the tenses are easy to forgot, but are important here). What it is for e1 to have occurred is for F* to have instantiated F.

      6) The rest of what you say sounds right.

  1. You say on page 4 “We cannot logically allow into our ontology events which temporally straddle the boundary between existence and non-existence.”
    This is the reason why only past events exist.
    But can’t we make sense of present events as being those that exist, but are incomplete?
    If you have events which exist, are succeeded by no events and that will continue to be succeeded by no events for a bit, then you cann accept that present events exist. Even if you have problems with saying that events will continue, the fact that events are succeeded by nothing seems to be enough. Past events are thjose that stopped before now, present events are those that haven’t stopped.

      • David Sanson on August 28, 2015 at 5:01 pm
      • Reply

      Just to follow up on Graeme’s question: do you think that the same sort of argument shows that Presentists cannot allow into their ontology intervals of time?

    1. I think that the growing universe/block theorist cannot allow temporally extended but incomplete events, even if the presentist can. I also think that postulating both present and past events is what leads to incoherence. I spend a lot of time arguing for both of these claims in The Existence of the Past.

    • Natalja on August 29, 2015 at 8:25 pm
    • Reply

    Thanks for this! First, a brief follow-up to Emily’s question: do you think there are any entities of which we can ever truly say that they’re occurring or happening? And second, am I right in thinking that ‘as of t’ means something different from ‘at t’, and if so, is the distinction a technical one, or is it one that is intended to be familiar, pre-theoretically?

Leave a Reply to Joseph Diekemper Cancel reply

Your email address will not be published.