2. Remarks on occurrent mental states and the ontological category of occurrents

I said just above that philosophers have largely ignored the question of what occurrent mental states are. Some might find this claim surprising. For is there not an extant debate on the notion of occurrents?

Indeed there is (e.g. Simons Reference Simons2000 and Melia Reference Melia2000). However, it is a very general metaphysical debate which hardly touches on the mind at all, and which is thus not especially relevant to my project. While it is true that the category of occurrent mental states is historically linked to the general category of occurrents, there is now little relationship between them.

The term ‘occurrent’ first appeared in philosophy as a noun, paired with ‘continuant’ in W. E. Johnson’s magisterial three-volume Logic (Reference Johnson1921–1924). Johnson intended ‘continuant’ as a more neutral replacement for ‘substance.’ Exactly what he meant by ‘occurrent’ is less clear (cf. Broad Reference Broad1924a), but it seems to have denoted a temporary state of a continuant: ‘while we cannot say that a continuant occurs, we can say that a state occurs; and anything that may be said to occur will be called an ‘occurrent.’ And I lay it down that any occurrent must be referred to a continuant or to two or more connected continuants’ (Johnson Reference Johnson1921–1924, vol. 3, xx–xxi).

Today it is often assumed that objects are continuants and events are occurrents. Continuants do not have temporal parts; occurrents do. However, almost everything in this area is controversial. Stout (Reference Stout2016) denies that occurrents must have temporal parts. Melia (Reference Melia2000) denies that they even exist. Perhaps because there is so little agreement on the fundamentals of the continuant vs. occurrent distinction, it is no longer in common use.Footnote ¹

As for the now-common adjectival use of ‘occurrent’ to refer to a type of mental state: its precise origins are unclear, but Johnson’s occurrents were presumably the inspiration. The originator may have been Broad, who made a three-part critical study of Johnson’s Logic (Broad Reference Broad1922, Reference Broad1924a, Reference Broad1924b). The earliest use I have found of the phrase ‘occurrent state’ is in Broad (Reference Broad1945), where he briefly refers to anger and jealousy as ‘occurrent states’ (p. 135). The contrast was not with continuants but with dispositions. For by then the rise of behaviorism had prompted a vigorous debate over the existence of mental acts (e.g. Gallie, Sprott, and Mace Reference Gallie, Sprott and Mace1947), which began sometimes to be called ‘occurrent states’ (e.g. Gallie Reference Gallie1948).Footnote ²

The key point here is that the concept of an occurrent mental state, although inspired by Johnson’s concept of an occurrent, had from its inception a distinct purpose. Occurrent mental states were contrasted with dispositional states, yet Johnson would have seen both of these as occurrents.Footnote ³ The general ontological category of occurrents, then, has always been quite distinct from the mental category of occurrent states. So the (small) literature on the nature of the former is not particularly pertinent to my question about the nature of the latter.

3. The background in action theory, and some preliminaries

After its introduction in the 1940s, the adjectival use of ‘occurrent’ was ultimately crystallized into its contemporary form in 1970, by Alvin Goldman in A Theory of Human Action. Goldman was building on William Alston’s distinction between ‘aroused wants’ and ‘latent wants’ (Reference Alston and Edwards1967, 402). Aroused wants are those that are currently guiding the subject’s actions, while latent wants are currently inert. Goldman dubs them ‘occurrent wants’ and ‘standing wants, ’ respectively. He expounds the theoretical need for the distinction with the example of his omitting to buy cheese at the supermarket even though he wanted to buy some. The way to explain the omission, Goldman argues, is to say that while he had a standing want to buy cheese, the want did not (alas) become occurrent while he was in the supermarket. Goldman also notes that the same distinction is needed for beliefs, and for the same explanatory purpose. For example (my own, not Goldman’s), even though Clem believes he should bow if he meets the queen, when the queen is before him he may fail to bow if his belief does not become occurrent. Hence, Goldman concludes, ‘Standing wants and beliefs can affect action only by becoming activated, that is, by being manifested in occurrent wants and beliefs’ (p. 88).

I shall henceforth adopt Goldman’s terminology, and refer to occurrent (mental) states and standing (mental) states – usually dropping the qualifier ‘mental’, as is common. (§5 will make clear why I choose the term ‘standing’ over the more common ‘dispositional’.) For convenience, I shall also call the distinction between occurrent and standing states the o/s distinction.

Goldman (Reference Goldman1970) distinguished the occurrent forms only of beliefs and wants. My interest is broader. I am asking what it is for a mental state, in general, to be occurrent. As I said in §1, philosophers assume that many kinds of mental phenomena, not just propositional attitudes, are occurrent. For example, in a well-known paper Block and Fodor (Reference Block and Fodor1972) noted that we ‘draw a distinction between dispositional states (beliefs, desires, inclinations, and so on) and occurrent states (sensations, thoughts, feelings, and so on)’ (p. 168). And it is commonly acknowledged that emotions come in both occurrent and non-occurrent forms (e.g. de Sousa Reference de Sousa and Zalta2014). In fact, the o/s distinction appears to be treated as exhaustive, in that all mental phenomena are assumed to be either occurrent or standing. Here is H. H. Price in his 1960 Gifford Lectures:

Again, Goldman (Reference Goldman1970) adopts a much narrower focus. But it seems clear that his deployment of the term ‘occurrent’ is conceptually tied to the earlier usages I have mentioned. We can trace a path from the adjectival term’s introduction in the 1940s through to Goldman, via its appearances in the intervening decades – in Price’s lectures and elsewhere (e.g. Taylor Reference Taylor1960, 82–85).

I therefore believe that Goldman’s usage is not just a terminological coincidence, but that he applied the existing conceptual distinction in order to mark out two modes of existence of propositional attitudes. So I shall use his treatment as a starting point for my question: what, exactly, is that distinction? In this paper I shall consider three answers to this question, two of which I will reject. All three answers are visible in Goldman’s introduction of the o/s distinction:

Goldman writes as if this passage provides a single, unified account of the o/s distinction. Yet it actually suggests (at least) three different accounts. Firstly, calling an occurrent want ‘a mental event or mental process, ’ or ‘a ‘going on’ or ‘happening, ’ suggests that occurrent states are in some way active (cf. the passage from Price above). But Goldman then adds the phrase ‘in consciousness’ to the latter description, which suggests that occurrent states are conscious states. Finally, his description of a standing want as ‘a disposition or propensity to have an occurrent want’ suggests that occurrent states are manifestations of mental dispositions.Footnote ⁴ The passage cannot provide a unified account, then, unless active mental episodes, conscious states, and manifestations of mental dispositions are simply identical. I think they are not. I also think that the second and third accounts, though they are now very popularly assumed, are misguided. In the next two sections, I show why. I will then defend the first account.

I think that behind the rather messy usage of the idea of occurrent states lies a coherent and valuable concept. The core of the concept is that occurrent states, in contrast with standing states, are active. The account of the o/s distinction that I shall ultimately propose has, I think, been implicit in its use more or less from its origin in the 1940s. So I am not so much suggesting a new account as articulating an existing one. But a necessary first step is to dispose of two other accounts which are, I shall argue, unsatisfactory.

Before I proceed to that task, six preliminaries.

Firstly, I emphasize again that my topic is not beliefs, wants, or propositional attitudes. While I shall speak of ‘occurrent beliefs, ’ this is for convenience only. If you hold that there are no such things as occurrent beliefs, please take me to be referring to thoughts or judgments. Nothing in the paper turns on whether beliefs themselves can be occurrent.

Secondly (and relatedly), nor is my topic the theory of action explanation. While that is Goldman’s (Reference Goldman1970) topic, I take him only as a jumping-off point. As it turns out, the o/s distinction has become a foundational concept much less in action theory than in epistemology and philosophy of mind. Rarely does an introductory or survey text in either of these two fields fail to invoke it at some point (for two examples, see §4). By contrast, in the philosophy of action, the distinction has faded along with Goldman’s theory. His version of the causal theory of action is no longer influential (by comparison, especially, with that of Donald Davidson); and so, since the o/s distinction itself was not generally adopted by other theorists, it seldom appears in the literature (with the notable exception of the work of Alfred Mele – see §4). So while my arguments concerning occurrent states may have some implications for action theory, in my view the more significant impact will be on epistemology and (especially) the metaphysics of mind.Footnote ⁵

Thirdly: for simplicity I shall speak as if occurrent and standing states are categorically distinct, but nothing I say excludes the possibility of intermediate or even indeterminate cases. It is even possible that the o/s distinction is graded or quantitative rather than categorical. This would naturally entail some further complexities in the accounts I shall consider, but I do not think that the basic points I shall be making would be affected. (Indeed, I think that my own account, in §6, would be very readily adapted to a graded version of the o/s distinction. Whereas I suspect that gradability would create even further difficulties for the account I reject in §5.)

Fourthly: while it is routinely assumed that minds feature both occurrent and standing states, I do not require this assumption. Gertler (Reference Gertler, Gertler and Shapiro2007) suggests that only occurrent states are genuinely mental. This view does not in itself abandon the o/s distinction, however. Indeed, the distinction thereby becomes even more important, for Gertler uses it to distinguish the mental from the non-mental. By contrast, to abandon the o/s distinction is to take the eliminativist view that there is simply no distinction to be drawn.

Fifthly: what about the eliminativist view I’ve just mentioned? Might the states that we call ‘occurrent’ turn out to have nothing interesting in common? One might suspect, in particular, that the o/s distinction is just a misguided piece of folk psychology.

Here I take a conditional stance: if the o/s distinction is genuine, then the account I present in §6 is the correct one. So if you deem that account unsatisfactory, you should indeed favor elimination. But more positively, I doubt that the o/s distinction is folk-psychological at all. While non-philosophers readily grasp the contrast between occurrent and standing beliefs if given a standard exemplar, they do not deploy it themselves. The term ‘occurrent belief’ has next to no non-philosophical use. A web search for it yields almost exclusively philosophical sources. And again, my concern is not with occurrent beliefs, but with occurrent mental states more broadly; and the folk do not appear to put experiences, sensations, judgments, and thoughts into a single category in the way philosophers do. As with ‘occurrent belief, ’ a web search for ‘occurrent state’ yields almost exclusively philosophical sources. Philosophers are the creators and consumers of the concept of occurrent states, so I submit that we may analyze it as we see fit. Indeed, it seems incumbent on us to do so, so as to keep our conceptual house in order.

My sixth and final preliminary is to set some ground rules. I take it to be clear from existing usage – where standing states are sometimes simply called ‘non-occurrent’ states – that the o/s distinction is exclusive and exhaustive. So I shall assume the following basic principles:

Therefore, an account of the o/s distinction must posit some differentiating property F as being possessed by all occurrent states but no standing states, like this:

or vice versa, like this:

The proposals we consider will therefore be formulated in one of these two ways. Note, however, that [Schema 1] is a great deal more intuitive than [Schema 2] – as suggested by the fact (again) that standing states are sometimes just called ‘non-occurrent.’ There is a strong sense that it is occurrent states which possess something that standing states lack, rather than the reverse. This observation will be significant in §5.

4. Occurrent states as conscious states?

As I mentioned in §1, it is a platitude that all conscious states are occurrent. If we add to this platitude the converse claim that all occurrent states are conscious, we get the view that the occurrent states just are the conscious states. A narrower version of this view, concerning beliefs (or propositional attitudes in general), is commonly assumed. Consider these glossary entries from two relatively recent books – an introduction to philosophy of mind, and a survey of the epistemology and metaphysics of self-knowledge:

Both entries convey that to have an occurrent belief is to occupy some sort of conscious state: that is, to be ‘actively aware’ of or ‘currently entertaining’ the attitude in question. It is harder to find avowals of the more general view that occurrent states just are conscious states, but this is simply due to the rarity with which the general concept of occurrent states (as opposed to the more specific concept of occurrent beliefs) receives explicit attention.Footnote ⁷ The proposal, per [Schema 1], is this:

How does (C) fare as an account of the o/s distinction?

To begin with, notice that (C) renders talk of occurrent states redundant, by simply identifying them with conscious states. If we accept (C), then to talk of ‘occurrent’ and ‘standing’ states is to multiply words unnecessarily; we should replace those terms with ‘conscious’ and ‘unconscious, ’ respectively. However, such a wholesale replacement would be warranted only if there were no interestingly distinct notion of occurrent states. I shall suggest that there is indeed a notion of occurrent states that is distinct from the notion of conscious states.

There is a use of the term ‘occurrent’ that distinguishes it from the term ‘conscious.’ While it is (again) a platitude that all conscious states are occurrent, it is far from agreed that all occurrent states are conscious. Alston (Reference Alston and Edwards1967) – whose discussion of wants was a major influence on Goldman (Reference Goldman1970) – says that an aroused want is not ‘a kind of conscious state or content of consciousness’ (p. 402). And more recently, here is Alfred Mele criticizing Goldman’s (Reference Goldman1970) claim that occurrent wants are conscious wants:

Mele gives the example of a boy who speaks spitefully to his girlfriend out of an unconscious desire to hurt her feelings. The desire causes his utterance, so it is active, and thus, by Goldman’s own criterion, occurrent. So not all occurrent desires are conscious (see also Davis Reference Davis2005, 72). Mele does not discuss beliefs, but the same goes for them. Extending his example, suppose that what the boy says is, ‘I never liked that silly hat you always wear, ’ and suppose he says this because he knows that his girlfriend is especially fond of the hat in question. He may not have consciously thought at the time that she is especially fond of the hat, yet clearly that belief was active in producing his remark – so the belief was occurrent. Hence not all occurrent beliefs are conscious, so (C) is false.

In fact – though Mele (Reference Mele2003) appears not to notice this – even Goldman (Reference Goldman1970) himself evinces doubt about (C) for the very same reason. Later in his book he notes that since unconscious wants can cause behavior, they ‘cannot be dismissed as mere standing desires’ (p. 121). Are they, then, occurrent? Here Goldman wavers. He says that ‘an unconscious want is a nonstandard form of want, since it violates one of the normal criteria for wanting’ (p. 122) – the criterion of consciousness. Yet he still seems to allow that such wants are occurrent. His reason is that ‘unconscious wants, like normal wants, [tend] to cause other wants by practical inference’ (ibid.), where the latter wants, at least if they are conscious, in turn can cause actions. Goldman is less sure that unconscious occurrent wants can cause actions directly (i.e. unmediated by conscious wants), but he sees it as an open possibility.Footnote ⁸

There undoubtedly exists, then, a sense of ‘occurrent’ on which some occurrent states are unconscious. And while I cannot say how widespread this sense of the term is as compared to the sense expressed by (C), here is an argument that (C) does not capture what we have in mind when we speak of ‘occurrent states.’ The nature of conscious states is considered a deep puzzle. So if we think that occurrent states are identical with conscious states then we should be deeply puzzled about the nature of occurrent states. Yet no one ever expresses such puzzlement. Admittedly, this could be because (as I emphasized in §1) we hardly ever examine the notion of occurrent states. Perhaps if we did, we would converge on (C), and the puzzlement would duly ensue. But the conclusion of the argument is just that we are not presently committed to (C), so there is no special reason to think that we would converge on it; hence we should investigate other possible accounts of the o/s distinction.

I also think that a sort of cognitive illusion may exist here. When we think about occurrent beliefs or wants, conscious beliefs or wants are what immediately come to mind. But this may be because conscious occurrent attitudes are more salient than unconscious occurrent attitudes – precisely because the former are conscious. (There may also be an association of the word ‘occurrent’ with the meaning of its cognate verb in common phrases such as ‘A thought occurred to me’, which are used to report on one’s conscious states.) This does not show that unconscious occurrent attitudes do not exist. It does not even show that they are ‘nonstandard’ (Goldman Reference Goldman1970; see above). Indeed, since our behavior often seems to be driven by attitudes of which we are not aware (e.g. Nisbett and Wilson Reference Nisbett and Wilson1977; Wilson Reference Wilson2002), the unconscious occurrent attitudes could even be considered the standard form.

Thus there does exist a notion of an occurrent state that is distinct from the notion of a conscious state. Now it could still turn out that this distinct notion is not coherent. But until we establish that, we should not endorse (C). It seems to me that the distinct notion is coherent. I will attempt to describe it more carefully in §6. Before I turn to that task, however, there is one more alternative account to be examined and put aside.

5. Occurrent states as manifestations of dispositions?

In §2 I noted that as well as suggesting that occurrent states are conscious states, Goldman (Reference Goldman1970) hints at another reading of the o/s distinction: that standing states are dispositions, while occurrent states are manifestations of dispositions. Standing beliefs and wants are indeed very commonly called dispositional beliefs and wants; and as I noted in §2, the need for some distinction roughly along these lines seems to have been the original stimulus for the introduction of the concept of occurrent states in the behaviorist heyday of the 1940s.

However, the proposal cannot just be that standing states are merely dispositions whereas occurrent states are merely manifestations of dispositions. This would violate the principle of Exclusiveness. Dispositions and manifestations are not mutually exclusive, for a disposition can be the manifestation of another (second-order) disposition. Broad (Reference Broad1933) gave the example of magnetism, which is the manifestation of magnetizability. And on the assumption that standing beliefs are dispositions, examples involving mental states are very common. For instance, a gullible person has many standing – i.e. ‘dispositional’ – beliefs that are the manifestation of her gullibility. Such a belief, then, would be both a disposition and a manifestation of a disposition.

Another way to put the problem is that, on the imagined proposal, being a manifestation of a disposition is sufficient for being occurrent. Yet the gullible person’s standing belief that (say) Nostradamus predicted the rise of Hitler is a clear counterexample, for it is a manifestation of her gullibility, which is itself a disposition.

To respect Exclusiveness, we must use one of the two schemas formulated in §3. Either one is possible in this case. Let us first try [Schema 1]:

One immediate problem to note with (M) is that it still makes being a manifestation of a disposition sufficient for being occurrent. Again, a gullible person has many standing beliefs that are manifestations of her disposition of gullibility. (M) counts these standing beliefs as occurrent.

But not only is (M) too liberal in what states it counts as occurrent; it is also too conservative. It makes being a manifestation of a disposition necessary for being occurrent. Yet on any non-trivial notion of a disposition, many paradigmatically occurrent states are not the manifestation of any disposition. A person can have entirely new thoughts that are not the manifestation of a pre-existing attitude. Most sensory experiences are not the manifestation of a pre-existing disposition. In short, it seems patently false that all occurrent states are manifestations of dispositions.

So we might try [Schema 2] instead. Now as I noted in §3, the way we think about the o/s distinction makes [Schema 2] a good deal less intuitive than [Schema 1]. We think of the distinction as being built around the possession by occurrent states of some positive feature which standing states lack. [Schema 2] reverses that picture. Thus, in this case we get:

This does indeed seem odd. One wants to ask: but what are occurrent states, beyond just being states which are non-dispositional?

The oddness is borne out when we see that, contra (D), occurrent states are often dispositional. Indeed they are arguably always dispositional, for dispositions are ten a penny. All (or practically all) mental states are dispositions of some sort. (Functionalists erect an entire theory of mental states on this fact.) The problem with (D), then, is that a mental state’s being dispositional is not sufficient for its being a standing state. Dispositionality is not a differentiating property of standing states. For example, consider my fear of snakes. This fear comes in standing and occurrent forms. My standing fear of snakes is a disposition which, if I am exposed to a snake, tends to cause me to feel occurrent fear. Now we would normally not consider the occurrent fear to also be a disposition. But it is. It is a multitrack disposition which tends, in me at least, to cause (among other things) sweating, thoughts of fangs and biting, and a strong urge to flee.Footnote ⁹

You may object that surely the idea of (D) is that standing states are mere dispositions, while occurrent states are dispositions plus something else; and so my observation that occurrent states are dispositions is beside the point. But to the contrary, this objection actually reinforces the point, which is that (D) comes at the issue from entirely the wrong direction. For if both standing states and occurrent states are dispositions, but the latter also have some other, differentiating property, then what we need is a proposal modeled on [Schema 1], not [Schema 2]. This is why (D) seemed odd in the first place. What we want to say about the o/s distinction in regard to dispositions is that standing states are nothing but dispositions, while occurrent states are something more.Footnote ¹⁰ Thus we are brought back to [Schema 1] again, where we ask: what is it, exactly, that occurrent states possess which standing states do not? Yet with the failure of (M), it seems that the entire manifestation-disposition approach can give no answer to that question.

All in all, then, this approach to the o/s distinction appears to lead nowhere. I suspect that its initial attraction rests on an ambiguity in the word ‘disposition.’ In its original use in English, the word referred to an aspect of a person’s demeanor. Outside philosophy, this remains its dominant use. This use is reflected in philosophy by the common practice of referring to standing mental states as ‘dispositional states.’ However, in philosophy, ‘disposition’ has also acquired a technical use which, while inspired by the original meaning, has developed in an independent direction. In that usage, a disposition is characterized simply as a tendency to manifest certain results under certain conditions. As a result, there is no longer much relationship between the dispositions discussed by philosophers and the dispositions ascribed to a person’s demeanor.Footnote ¹¹ The fact that standing mental states are often called ‘dispositional states’ reflects the latter sense of the word ‘disposition, ’ and should not be taken to suggest that the o/s distinction has anything much to do with dispositions in the former, technical sense.

6. My proposal: occurrent states as active states

The last three sections have comprised an argument by elimination. In §3 I noted three ideas about occurrent states: that they are conscious states, that they are manifestations of dispositions, and that they are active states. I do not claim that no other accounts are possible, but I know of none that have been suggested. I argued in §§4–5 that neither of the first two ideas is adequate. I shall now pursue the third.

There remains, I remind the reader, the option of pursuing the argument by elimination all the way to the bitter end. Perhaps the so-called occurrent states have nothing in common, and we should simply eliminate the o/s distinction. This possibility cannot be entirely dismissed. I do not pretend to prove that the account I shall offer is correct. I do think, however, that it is the best account available, and that it may also offer sufficient explanatory power to earn its keep.

In what sense, then, is an occurrent state active?Footnote ¹² In my view, the word indicates an implicit appeal to a fairly ordinary contrastive notion. A thing is active (or is in an active state) when it is changing in certain salient properties over time; whereas if it is unchanging in those salient properties, it is inactive.

Many diverse kinds of things can be described in this contrastive way. Consider a volcano, a police investigation, an ant colony, a tumor. Certain changes in certain properties constitute the thing’s being in an active state, but the pertinent kinds of changes (and kinds of properties) will be very different from one case to the next. In a tumor, certain metabolic changes may count as making the tumor active, whereas no purely metabolic changes will count as making a police investigation active. By the same token, the sorts of changes that do make a police investigation active – which might include psychological, administrative, and legal events – will be wholly unrelated to what makes it the case that a volcano is active.

Crucially, when I say that occurrent states are active, I do not mean that they are causally active in producing action. There is a difference between a mental state’s being active and its affecting action. This difference is clear in Goldman’s (Reference Goldman1970) original presentation of the o/s distinction. He says that standing wants ‘can affect action only by becoming activated’ (Reference Goldman1970, 88) – that is, by becoming occurrent. Activity, then, is a necessary condition for a want to affect action. But it is not sufficient.Footnote ¹³ A want may be activated yet not issue in action. We do not act on all of our occurrent wants. You may occurrently want a third slice of peach pie, but dietary concerns or a sense of decorum may stay your hand. So the activity of an occurrent state, as I am speaking of it – that is, as involving change in certain salient properties – is a feature of the occurrent state itself, not of its effect on the subject’s actions. It is internal, not external. A parallel distinction can be made in the other sorts of cases I have mentioned. A volcano may be active (e.g. undergoing internal changes in pressure, temperature, or magnetization) even if it is not presently erupting.Footnote ¹⁴ An ant colony may be active even if none of the ants are out gathering food or otherwise engaged in the colony’s environment. A police investigation may be active even at a time when the case is merely being discussed inside the detective unit. Something’s being active in the relevant sense, then, does not entail its showing any external effects of that activity. The salient property changes will usually be an internal matter.

Nevertheless, the causal connection to action makes it obvious why Goldman (Reference Goldman1970) found it natural to speak of occurrent states as ‘active’ or ‘activated.’ Since action is itself a kind of activity (the intentional activity of an agent, roughly speaking), it is intuitive that mental activity is required to bring it about. (Again, this is paralleled in the other sorts of cases. A volcano, for example, cannot erupt without undergoing internal activity of particular kinds.) I do not intend this intuition about the causation of actions as an a priori claim. It is an empirical conjecture. Again, a mental state’s being active is not the same as its causing the subject to act, and the connection between the two is contingent. However, while I find it conceivable that some actions could be caused entirely by inactive states, I doubt that any actual actions are caused in this way. I would further hazard that if we knew of a case of an action which was caused entirely by inactive mental states, we would not call those inactive mental states ‘occurrent.’ So I suggest that occurrent states are active mental states, and that they are nomologically necessary for the production of action. These claims certainly require more defense, but here my purpose is only to offer a general sketch of an account of the o/s distinction, suggesting it as a better alternative to the two I have already canvassed. A full exploration of the account must wait on another occasion.

The general notion of activity at hand, then, simply entails some sort of fluctuation of properties internal to the entity in question. These fluctuations constitute an occurrent state. So here, finally, is my proposal regarding the o/s distinction:

Another way to put this, I think, is that occurrent states are processes, whereas standing states are not (cf. Bartlett Reference Bartlettunpublished). Recall Goldman: ‘An occurrent want is a mental event or mental process’ (Reference Goldman1970, 86). I shall continue to speak of activity and fluctuation rather than processes, however.

It must be admitted that (F) is somewhat vague. But this is unavoidable. It is largely an empirical matter what kinds of fluctuation, in what kinds of properties, make an occurrent state – just as it is an empirical matter what kinds of fluctuation, in what kinds of properties, make a tumor or a volcano active. In our own case, the obvious candidate is certain kinds of fluctuation in neurophysiological properties.Footnote ¹⁵ But I do not wish to foreclose the possibility that in other kinds of subjects (i.e. non-human, non-biological, and perhaps even non-physical), different kinds of properties and different kinds of fluctuation might be involved.

Moreover, (F)’s vagueness is not much of a strike against it, for the alternatives are no better in that respect. (C) says that occurrent states are conscious states, which only seems concise and informative until one remembers that we still don’t understand what consciousness is. (M) and (D) draw on the metaphysics of dispositions, which remains a subject of ongoing debate. And even putting that debate aside, both (M) and (D) are very vague. They do not tell us, for example, what kinds of dispositions are involved.

My point is not to criticize those accounts all over again, but to emphasize that (F), despite its evident vagueness, is no less informative than they are. All of these proposals would need significant amounts of elaboration. But in light of the problems with those earlier proposals, I suggest that (F) is the one to pursue.

While we are recalling the earlier proposals, we may note that (F) nicely explains their prima facie attractiveness. Consider first (M) and (D). While a disposition does not have to be inert, nor a manifestation active, the classic exemplars always have those features – as in a fragile glass shattering, or a soluble sugar cube dissolving. When we think of dispositions and manifestations, we think of inert states and active states, respectively. So if standing states just are a kind of stable state and occurrent states a kind of active state, one might easily come to think of standing states as dispositions and occurrent states as manifestations. As for (C), in a way it is very close to (F), as they appear to have similar motivations. (C) is what you get if you set out to identify occurrent states with whatever states are the proximate mental causes of actions, but become distracted by the most salient instances of those causes (see the penultimate paragraph of §4): conscious states. One gets from (C) to (F) by noticing that some unconscious states can also serve in that causal role.

I am suggesting, then, that as an account of the o/s distinction, (F) is superior to the two competing accounts that I have considered. There still remains, as I have said, the option of eliminating the distinction. As I said in §3, those who are tempted by this option may read my thesis as a conditional: if we retain the o/s distinction, then (F) is the account of it that we should accept. Of course, this conditional claim may be challenged. Perhaps one of the accounts that I dismissed in §§4–5 can be salvaged, or perhaps there is another account that I have not thought of. But I contend that if you agree with my rejection of those accounts, then absent some promising new account, you should see (F) – or at least something like it – as the only alternative to the elimination of occurrent mental states from our mental ontology.