1. James on Conscious Flow

In ‘The Stream of Thought’ in Principles of Psychology, James begins with arguing that psychology's foundational first data point is that ‘thought goes on’, and then defends several more specific theses about what that thought, or consciousness, is like (James Reference James1890: 224). One of his first goals is to argue that consciousness is ‘sensibly continuous’ (James Reference James1890: 226). By ‘continuous’, James says that he means ‘that which is without breach, crack, or division’ (James Reference James1890: 237). What it is for consciousness to be sensibly continuous is for consciousness to feel as though it is without breach, crack, or division.

He uses the ‘stream of consciousness’ metaphor to capture all these ideas. This is what that metaphor looks like: ‘[consciousness], then, does not appear to itself chopped up in bits. Such words as ‘chain’ or ‘train’ do not describe it fitly as it presents itself in the first instance. It is nothing jointed; it flows. A ‘river’ or a ‘stream’ are the metaphors by which it is most naturally described’ (James Reference James1890: 239). As we can see, this is where James introduces the language of ‘flow’ rather than ‘sensible continuity’. They mean the same thing for him, but ‘flow’ has simply become a more popular term to express these ideas in recent years.

James's metaphor has been very widely taken up in philosophy and psychology. Many of us seem to agree with James that there is just something right about the idea that our conscious lives do not appear ‘chopped up in bits’ and that our moments of conscious experience to seem to flow smoothly into each other, from one moment to the next. But James would have been surprised at how quickly some of us agree with him. He does not think the stream metaphor just obviously applies; he thinks that the claim that the metaphor applies is a substantive philosophical conclusion. This is a fact that is missed by readers, such as Aron Gurwitsch, who say that for James the interconnectedness of conscious states is something so obvious that it ‘need not be established’ (Gurwitsch Reference Gurwitsch1943: 449).

For James, there are two ways that consciousness can seem to have a breach or a gap that might worry us—gaps in time and gaps in content. First, there are time gaps of unconsciousness in our lives. According to James, what makes the moments before and after a state of unconsciousness feel like they belong to the same stream of consciousness is that they feel like they belong to the same self (James Reference James1890: 239). Secondly, this is how he describes the worry about gaps in content:

James goes on to defuse the second worry through arguing that closer introspective observation will reveal that no change in the stream of consciousness is truly totally abrupt. In his view, every two sequential moments of consciousness are somewhat phenomenologically alike, with the feelings in the first moment leaving at least some experiential traces in the feelings that are present in the second moment. This is how James thinks this style of reply applies to his objection about thunder: ‘[into] the awareness of the thunder itself the awareness of the previous silence creeps and continues; for what we hear when the thunder crashes is not thunder pure, but thunder-breaking-upon-silence-and-contrasting-with-it’ (James Reference James1890: 240). There would be a true breach in consciousness if the feeling silence was totally eradicated at the moment of thunder, but because traces of silence remain, there is no true breach.

James thinks this sort of moment-to-moment coloring, overlapping, or similarity is ubiquitous in consciousness. Here is one example of how he describes the phenomenon as it shows up during reading:

James thus rejects skepticism about the stream metaphor for consciousness by emphasizing his view about what it is for thought to be ‘sensibly continuous’. It is for any two moments in thought to overlap at least slightly in their phenomenology—for the feeling of silence to remain, at least somewhat, in the moment in which the feeling of thunder that follows begins, or for the feeling of ‘no more’ to remain, at least somewhat, when the feeling of the word ‘than’ arrives.

Moreover, James believes that in examining these two types of apparent discontinuity his defense of the stream metaphor has been exhaustive. As he writes before embarking on the defense of the metaphor, these two kinds of breaches in the stream of consciousness are’ [the] only breaches that can well be conceived to occur within the limits of a single mind’ (James Reference James1890: 237, emphasis mine). As I will argue in the next section of this paper, James's defense is sufficient to deal with the challenges raised by the phenomenology and neuroscience of binocular rivalry, but reading Buddhaghosa's work shows us that there may be a third way for consciousness to fail to flow, a way that James did not consider.

2. Binocular Rivalry and Flow

In Waking, Dreaming, Being, after Thompson reviews some literature on binocular rivalry, he raises the following provocative question: ‘But is the flow of conscious perception really discrete, like the frames of a movie or a series of snapshots? Maybe this way of thinking about perception comes from relying too much on the unusual situation of binocular rivalry’ (Thompson Reference Thompson2014: 34). In so doing, he is saying that binocular rivalry is a phenomenon that poses at least some counterexamples to the view that moments of consciousness always flow into one another—perhaps because of the phenomenology of binocular rivalry, perhaps because of what the neuroscientific study of binocular rivalry reveals about what goes on in our brains during binocular rivalry. I will argue that neither of these conclusions applies. As long as we keep in mind James's view of what is actually required for one moment of consciousness to flow into another, we can see that neither the phenomenology nor neuroscience of binocular rivalry poses a challenge to the idea that there is a stream of consciousness.

In binocular rivalry experiments, participants simultaneously view two images. One image is exposed only to the left eye, and one image is exposed only to the right eye. In visual consciousness, participants do not see both images at once. Instead, their experience alternates from an experience of one image to an experience the other image (Thompson Reference Thompson2014: 22). With the basic methodology in view, we can consider this question: ‘In binocular rivalry experiments, from the subject's perspective, does consciousness seem to fail to flow?’. What might motivate a ‘yes’ answer to this question is the fact that from moment to moment in these experiments, what is presented in the stream of consciousness can be dramatically different. A checkerboard may seem to change suddenly into a face, or a selection of maple leaves may seem to change suddenly into a butterfly, to mention two of Thompson's specific examples (Thompson Reference Thompson2014: 29, 33).

But this line of thought is structurally identical to James's suggestion that thunder breaking upon silence challenges the ‘stream’ metaphor, and it can be answered in a structurally identical way. As long as there is some trace of the previous moment's experience in the present moment, then the sensible continuity of consciousness has been maintained. And if we look at subjects reporting on the phenomenology of binocular rivalry, it seems as though some of these traces persist.

For instance, of his own experience, Thompson writes:

Part of what it is to have the experience gradually change from a checkerboard to a face is for some traces of the complete checkerboard experience to remain in the various stages of experience as the checkerboard changes to the face. These traces are what explain the feeling that something is changing before one's eyes. And even if there were not these gradual changes—even if the face changed completely and abruptly into the checkerboard—there would still be a felt difference between seeing a checkerboard that was a moment ago a face and seeing a checkerboard that was a moment ago a checkerboard. Experiential traces of the previous images explain this fact.

Not to mention that we can make use of a resource that James does not consider in his analysis of thunder breaking upon silence—namely, elements of the stream of consciousness outside of the dramatic change under consideration. Even if the switch from the checkerboard to the face is abrupt, part of what makes consciousness seem to flow from one moment to the next is that so much more stays constant in consciousness: the subject's auditory experiences of the quiet room the subject is in; the subject's experiences of their own body, the subject's experience of their internal monologue, and so on. At the level of phenomenological description, there is nothing about binocular rivalry that challenges James's thesis that every moment in the stream of consciousness is sensibly continuous with the next.Footnote ²

But perhaps Thompson thinks that something about the neuroscience of what goes on during binocular rivalry challenges the view that consciousness flows. I will argue that while the neuroscience does not teach us anything about conscious flow, it does teach us about another aspect of consciousness—how quickly it can change in terms of what it presents to a subject.

We can see this through looking to one of the studies that Thompson cites as relevant to the issue of conscious flow, namely, that of Doesburg and colleagues (2009, cited at Thompson Reference Thompson2014: 32). In the study, participants pressed a button to indicate when their experience shifted from one image to another image (in the context of binocular rivalry). Doesburg and colleagues used an EEG with the aim of understanding, from the perspective of a participant, what was going on in the brain when the view shifted from one image to another. Here is what they say about the relationship between their observations and conscious flow:

As this passage makes clear, what Doesburg and colleagues have supported is the claim that there is a certain rhythm at which the contents of consciousness can change (here I use ‘contents of consciousness’ just to mean ‘what consciousness seems to present to a subject’). It is not the case that every passing millisecond or nanosecond presents an opportunity for the contents of consciousness to change; instead, the rhythms that these neuroscientists are studying seem to give us a more accurate sense of when the contents of consciousness can change. This is an interesting result. Perhaps folk psychology tells us that consciousness is more sensitive and changeable than this. If so, we should use results like this one to temper our misplaced confidence in the high sensitivity and changeability of the stream of consciousness.

Doesburg and colleagues also say that this finding is in tension with the view that consciousness is a continuous and steady stream, and this is where they make a mistake. All that is required for flow is for the phenomenology of sequential moments of consciousness to overlap at least to some degree, and this is compatible with whatever the facts are about the rate of change of contents presented to us via the stream of consciousness. What this study actually supports is a conclusion about the rhythm at which the contents of consciousness can change.

At this point, one might raise the worry that nothing at all could falsify James's view about conscious flow. But this worry would be misplaced. James says that immediately sequential moments of consciousness always feel like they overlap slightly in their phenomenology and that moments of consciousness that are divided by a span of unconsciousness still feel like they belong to the same self. But if consciousness does not always feel like one of these two ways, then James's claim that consciousness always feels sensibly continuous would be mistaken. Perhaps there are experiences he did not consider that falsify his universal claim, or perhaps he did not have sufficient training to describe his experiences accurately, or perhaps he misdescribed his experiences because of some sort of theoretical bias. If any of these possibilities are the case, then James would be wrong about conscious flow. As we will see in the next section of the paper, we can use materials from Buddhaghosa's work to build a criticism along these lines.

3. Abhidharma and Flow

Here is how Thompson identifies a disagreement between James and some Abhidharma philosophers on conscious flow:

Consider the inference at work in the passage above: if it turns out that consciousness is actually made of ‘discontinuous and discrete moments of awareness’, then James's stream metaphor is inaccurate. Consciousness seems to flow, but in reality it does not. Whether this inference is correct, and whether it represents an objection to James grounded in the work of Abhidharma philosophers, depends on just how we interpret ‘discontinuous and discrete moments of awareness’.

If what we mean by ‘discontinuous and discrete moments awareness’ is something like ‘moments that exist very briefly and are metaphysically independent of each other’, then the inference does not make sense as an objection to the view that consciousness flows. This is certainly a claim that we can find in the Abhidharma and Buddhist philosophy more broadly (see von Rospatt Reference von Rospatt1995 for discussion). But that claim does not make sense as an objection to James because James's claim about flow is a claim about phenomenology, namely, the claim that the phenomenology of any two sequential moments of consciousness will always overlap to a certain extent or the claim that any two moments will feel like they belong to the same self. The claim is not about the metaphysical priority of any particular temporal unit of experience. Even if every moment of experience is brief and metaphysically independent of all the other moments of experience, these moments could fail to feel as though they are brief and independent of each other. It is possible that they could feel in a way that coheres entirely with James's description of the stream of consciousness.

Thus, if Thompson's inference above offers up a good objection to James, we need evidence that there are moments in streams of consciousness that actually feel discontinuous from each other. And if this truly is an objection to be found in Abhidharma philosophy, what we also need is textual evidence that this phenomenological judgment can be found somewhere in the works of the Abhidharma philosophers.

There is evidence in favor of both of these points, but the evidence is less straightforward than Thompson suggests. Looking at this evidence allows us to see that Thompson misdescribes the nature of the disagreement between James and the Abhidharma philosophers. The evidence I draw on comes from Buddhaghosa's Visuddhimagga: The Path of Purification (Reference Buddhaghosa and Ñāṇamoli2010) in particular. In Visuddhimagga we do not find an observation that simply clashes with James's analysis of conscious flow. Specifically, we do not find the claim that sometimes two immediately sequential moments of consciousness are so radically different in content that there is a real breach between them, which is James's account of how consciousness flows in most ordinary, day-to-day experience. What we find instead is a rejection of James's second sense of flow that James proposed to deal with the special case of flow that obtains across moments of unconsciousness. And even more interesting, what we find is the suggestion that there is a third way for consciousness to fail to flow. I will discuss these points in turn, focusing particularly on the latter point.

James's second sense of flow refers to a sense of self: he says that when the stream of consciousness stretches across a period of unconsciousness, flow still obtains because the states before and after the gap feel like they belong to the same self. In the Abhidharma and in Buddhist philosophy more broadly we can frequently find a rejection of the claim that the self, as normally construed, exists (see, e.g., Albahari Reference Albahari2006). This is a metaphysical thesis that is not directly in tension with James's second analysis of flow, which is a phenomenological claim.

But this metaphysical thesis also sometimes goes hand in hand with a phenomenological claim that does genuinely challenge James's second analysis of flow. For instance, in Buddhaghosa's Visuddhimagga, we can find the claim that we can ‘liberate’ ourselves from perceiving anything at all as our self after engaging in the right kind of contemplative practices. He writes that the meditator ‘delivers, liberates the mind . . . from the perception of self by means of the contemplation of not self’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 282/VIII.233). And if this is possible for some subjects, then their streams of consciousness will fail to flow in James's second sense: at least sometimes, they will not feel as though multiple moments of consciousness that are separated by a period of unconsciousness belong to the same self.

Before going on to consider the third sense in which consciousness can be discontinuous, we can consider the following question that challenges the line of thought above. Is my claim that according to Buddhaghosa we can learn to experience two subsequent moments as failing to belong to the same self actually incompatible with what Buddhaghosa says about a mental state called bhavaṅga (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 458/XIV.114)?

On Jonardon Ganeri's reading, bhavaṅga has an intentional content ‘roughly consisting in autobiographic semantic information’ (2017: 46). On Rupert Gethin's reading, ‘a being's bhavaṅga itself represents a kind of summing up of what he or she did in his or her previous life; in crude terms, it represents a kind of balance sheet carried over from the previous life detailing how one did’ (Reference Gethin, Pagel and Skorupski2005: 167). If these readings are correct, then bhavaṅga could be, at least sometimes, a part of the causal explanation of why some subjects experience sequential moments of consciousness as belonging to the same self. But my reading of Buddhaghosa is consistent with such a possibility. In my view, Buddhaghosa would say that while bhavaṅga might be a part of the causal explanation of why some ordinary, untrained subjects experience sequential moments of consciousness as belonging to the same self, part of the purpose of the meditative practices he describes in Visuddhimagga is to train subjects to respond differently to bhavaṅga and, more broadly, to respond differently to everything in the world, in a way that causes them to abandon the perception of self in all things (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 282/VIII.233, 632/XX.4, 655/XX.90).

Second, and even more interesting, in Visuddhimagga we can find the claim that the stream of consciousness can feel discontinuous because we can feel each moment of consciousness pass out of and into existence; that is, consciousness, like all things, appears to us ‘perpetually renewed’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 657/XX.104; see Gethin Reference Gethin2004: 216 for more on this particular passage as well as my references to Buddhaghosa immediately below). This is a kind of felt discontinuity that is compatible with moments in the stream of consciousness overlapping in content and with moments in the stream of consciousness feeling like they belong to the same self. It is a third kind of discontinuity that James did not consider in his analysis of how consciousness might fail to flow.

In these passages, Buddhaghosa describes a meditator who has first engaged in all the contemplative practices described earlier in Visuddhimagga going through a (roughly) three-step process that enables her to experience all things as impermanent and, moreover, as constantly destroyed and renewed. The first step is reflecting on the impermanence of materiality and as a result becoming able to experience materiality as impermanent. The second step is reflecting on the impermanence of consciousness and as a result becoming able to experience consciousness as impermanent. The third step is reflecting on the apparent constant creation and destruction of impermanent things and as a result becoming able to experience all impermanent things as constantly being created, destroyed, and renewed.

One of Buddhaghosa's suggested strategies for achieving the experience of materiality as impermanent is to reflect on how the materiality that shows up in various temporal subdivisions of a body's life seems like impermanent materiality. For instance, Buddhaghosa suggests that the meditator think of a body's life divided into three equal temporal spans; noticing that nothing of the materiality from the first third of that life persists in the second third of the life, the meditator should attribute impermanence to the materiality from the first third of life (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 645–46/XX.47–49). He suggests that the meditator continue to carry out this procedure with narrower and narrower time-slices of a body's life (e.g., the first ‘tender decade’, the second ‘sport decade’, and so on) (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 646/XX.51). He also recommends that the meditator carry out an analogous procedure with the materiality that occurs in many other temporal spans, like the span of a year or a season, and that as a result she will thereby eventually gain the ability to perceive the impermanence in materiality (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 647/XX.56–57).Footnote ³

Buddhaghosa says that after engaging in this process, the meditator can turn her conscious thinking upon the very consciousness with which she was thinking about the impermanence of materiality. He says that the meditator should be able to experience this consciousness that was directed at materiality as impermanent as well (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 652/XX.79). Moreover, he suggests that this process should be able occur recursively, that is, that the meditator should be able to experience her conscious thinking about her conscious thinking about the impermanence of materiality as impermanent and so on (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 653/XX.80–81).

After a meditator has engaged in both sets of contemplative exercises that are focused on developing the capacity to experience the impermanence of material things and of consciousness, Buddhaghosa says that she is in a position to contemplate change in impermanent things, which he asserts is equivalent to contemplating their constant apparent ‘rise and fall’, that is, their constant apparent creation and destruction (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 657/XX.93–95). The meditator is directed to compare all things to the sound of a lute, which does not come from any ‘store’ before it arises or go to any ‘store’ after it falls (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 657/XX.96). And after this final contemplative exercise, all things ‘appear to [the meditator] as perpetually renewed’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 657/XX.104).Footnote ⁴

Expanding on this idea with a set of similes, Buddhaghosa writes: ‘they are not only perpetually renewed, but they are also short-lived like dew-drops at sunrise, like a bubble on water, like a line drawn on water, like a mustard seed on an awl's point, like a lightning flash’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 657/XX.104). And if we really can experience consciousness in this way—always being created and destroyed—then Buddhaghosa has identified a real sense in which we can notice breaches in the stream of consciousness.Footnote ⁵ In James's language, the constant apparent destruction and renewal of consciousness could be a way that consciousness does manage to ‘appear to itself chopped up in bits’ (James Reference James1890: 239).

We can see some additional evidence for the idea that Buddhaghosa describes a third way in which consciousness might fail to flow in his analysis of the citta-vīthi, or cognitive series, and in his description of how we can put our knowledge of the citta-vīthi to use during meditative practice. On Ganeri's reading of Buddhaghosa, the cognitive series is a sequence of cognitive states that occurs as the mind ‘oscillates’ between taking up a task and occupying its resting state (Ganeri Reference Ganeri2017: 42–47). What Ganeri calls the resting state of the mind is a state that Buddhaghosa calls bhavaṅga, which he says, ‘goes on occurring endlessly, in periods of dreamless sleep, etc., like the current of a river’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 458/XIV.114). When we are interrupted from this resting state, a series of cognitive events, the citta-vīthi, eventually culminates in consciousness of some sort, which persists briefly and then lapses back into bhavaṅga.

In the case of visual consciousness, the cognitive series begins with a stimulus impinging on the eye, and if the stimulus is strong enough, it will ‘disturb’ bhavaṅga (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 463/XIV 115). After this initial disturbance, there will be a sequence of cognitive events that Buddhaghosa says perform functions including ‘adverting’, ‘receiving’, ‘investigating’, ‘determining’, ‘impulsion’, and ‘registration’ (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 463/XIV 114–24; see also Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 24/I.57). On Ganeri's reading, ‘registration’ corresponds to ‘having an object consciously in view’ (Ganeri Reference Ganeri2017: 58). More generally, on Ganeri's reading, all the steps of the cognitive series are performed by mental modules that are analogous to systems identified in contemporary cognitive science, e.g., the ‘early, middle, and late visual subsystems’ (Ganeri Reference Ganeri2017: 200).

Buddhaghosa draws on our knowledge of the cognitive series, as described above, in one of his descriptions of how a meditator can come to experience consciousness as impermanent. To put that idea into conversation with James, Buddhaghosa draws on our knowledge of the cognitive series in order to help explain how we can learn to experience a failure to flow in consciousness. In this part of the text, Buddhaghosa says that a meditator should reflect on how both material things and consciousness are generated. According to Buddhaghosa, as a result of focusing on the generation of material things and consciousness, the meditator will come to understand more deeply that both are things that pass out of existence (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 639–45/XX.22–45). In his discussion of reflecting on the generation of consciousness in particular (as opposed to the generation of materiality), Buddhaghosa explicitly invokes the cognitive series as it is defined earlier in the text. He says that the mediator should reflect on how each stage of the cognitive series leads to the generation of consciousness and that as a result the meditator will more deeply understand that consciousness is impermanent (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 644 / XX.44–45, 644, fn. 18).

This is one of the ways in which we can contemplate impermanence. And according to Buddhaghosa, contemplating impermanence in things is a part of what enables us really to experience them as impermanent (Buddhaghosa Reference Buddhaghosa and Ñāṇamoli2010: 632/XX.4). In his view, the meditator's theoretical knowledge about the cognitive series can be put to practical use in a contemplative exercise that will help enable the meditator to learn to experience consciousness as impermanent. Again, this is a way that Buddhaghosa believes that the right meditative practice can lead us to have experiences in which ‘consciousness appears to itself chopped into bits’ (James Reference James1890: 239).

This is why Buddhaghosa is best interpreted as challenging the way James understands his thesis about flow rather than merely as saying that James's thesis is false. To miss out on this fact is to underestimate the contribution Buddhaghosa makes to this discussion. He does not simply defend one option in logical space and disagree with James. Instead, he contends that James has misrepresented logical space and that the correct view about conscious flow occupies a region of logical space that was inconceivable to James.