SYNC-ING IN THE STREAM OF EXPERIENCE
TIME-CONSCIOUSNESS IN BROAD, HUSSERL, AND DAINTON
Department of Philosophy and Cognitive Science
Buffalo, New York 14208
Copyright (c) Shaun Gallagher 2003
PSYCHE, 9(10), April 2003
KEYWORDS: phenomenology, time-consciousness, overlap model, specious present.
COMMENTARY ON: Dainton, B. (2000). *Stream of Consciousness: Unity and
Continuity in Conscious Experience.* London: Routledge.
ABSTRACT: By examining Dainton's account of the temporality of
consciousness in the context of long-running debates about the
specious present and time consciousness in both the Jamesian and
the phenomenological traditions, I raise critical objections to
his overlap model. Dainton's interpretations of Broad and Husserl
are both insightful and problematic. In addition, there are
unresolved problems in Dainton's own analysis of conscious
experience. These problems involve ongoing content, lingering
content, and a lack of phenomenological clarity concerning the
central concept of overlapping experiences.
Barry Dainton provides a detailed and fascinating analysis of the
temporal structure of consciousness and the "specious present" in
Chapters 6 and 7 of *The Stream of Consciousness* (2000). As he rightly
notes, in agreement with Husserl, this is one of the most fundamental
problems for a phenomenology of consciousness. Trying to sort it out
has been a difficult and sometimes exasperating challenge for a series
of philosophers and psychologists, including (to name only a few from
what we might call the classical period in this regard) Lotze (1887),
James (1890), Stern (1897, 1898), McTaggart (1908), Brentano (1911),
Broad (1923, 1938), and Husserl (1927). Running alongside of these
original analyses there have been numerous scholarly commentaries that,
although sometimes helpful and important, often misconstrue and confuse
the issues. For example Mabbott's (1951) critique of the specious
present, Mundel's (1954) defense, and Plumber's (1985) rejoinder. I
will argue that Dainton's Chapter 6 belongs to this line of scholarly
commentary, and that it is in some respects helpful and in some respects
confusing. I will also suggest, however, that Chapter 7 should take its
rightful place in the line of original analyses. Like most of these,
however, it is also problematic.
2. DAINTON'S ANALYSIS OF BROAD
The majority of theorists that I listed above have wrestled with two
propositions that Lotze (1887) took as basic assumptions. To introduce
them, I'll call them Lotzean Assumptions, and I'll abbreviate them as
LA1 and LA2.
LA1: The perception of succession requires a momentary and
indivisible, and therefore durationless act of consciousness.
Dainton points out, for example, that Broad rejects LA1 in his early
account of the specious present, and maintains that an act of awareness,
and not just its speciously present contents, has some short duration.
In his later account, however, Broad accepts LA1 and treats an act of
awareness as momentary. Closely tied to LA1 is a second assumption.
LA2: A sequence or succession is represented by persisting
sensations or memory images that are simultaneous in present
Taken together these two assumptions inform William James's analysis of
the specious present. According to James (1890, p. 622), an act of
awareness is discrete and momentary (LA1). He also accepts LA2,
although to make sense of it he leaves the realm of phenomenology and
explains it in terms of brain processes: "the brain-processes of various
[successive] events must be active simultaneously, and in varying
strength, for a time-perception to be possible" (1890, p. 633n). It
would be interesting to discuss James's view in light of contemporary
research by people like Libet (1985, 1992) and Poeppel (1994, 1988), but
Dainton wants to stay with phenomenology, so we'll do the same.<1>
On the phenomenological level there seems to be a problem with the
specious present, although James actually takes it as a solution. Lotze
had already recognized the problem. To be aware of *successive* objects
consciousness needs to compare the earlier and later objects in an
operation that makes the earlier and later *simultaneous*. The problem
is how simultaneously presented objects can be sensed as successive
objects - how can they be both simultaneous and successive? Elsewhere,
and for reasons I won't rehearse here, I've called this problem, which
can be traced back as far as Augustine, the "cognitive paradox"
(Gallagher, 1998). Lotze decided to avoid this paradox by proposing a
theory of temporal signs according to which sensed content contains
markers derived from their objective temporal order. So although the
sensed content is simultaneous with the act of awareness, it is marked
in some way as belonging to the past. This, however, would not be a
direct perception of succession, and many of the psychologists that
James cited (for example, Herbart, Ward and Volkmann ) recognized this
paradox as a difficulty for any theory of the direct perception of
succession. James himself nonetheless takes this paradox to be the
solution, and in fact, an expression of the specious present. As he
later put it, "earlier and later are present to each other in an
experience that feels either only on condition of feeling both together"
(1894, p. 77).
Broad's account of the specious present also depends on the two Lotzean
assumptions. Even though Broad rejects LA1 in his early account, he
begins by accepting it in order to define the specious present, and then
later drops it. LA2, however, acts as an assumption throughout Broad's
account. Dainton's interpretation of Broad overlooks an important point
about this assumption. It has to do with the nature of the contents
that according to Broad and LA2 continue to persist even after the event
to which they correspond. Specifically, Broad maintains a sense-data
theory of consciousness, and he maintains LA2 precisely in terms of
sense-data that, taken together, constitute a sensory field. Within a
sensory field temporal change takes place, but "the qualitative
differences between its earlier and its later sections will be sensed
together ..." (Broad, 1923, p. 352). In effect, actual sense-data,
generated in a past experience, persist in present consciousness along
with actual sense-data generated in present experience. This is why
Broad, in contrast to James, limits the specious present to past and
present and excludes the future - the future has not yet generated
sense-data, so there is no content to the future. Past and present
sense-data have some existential value. They are really there in the
Any account of the specious present that includes LA1 and LA2 will run
into an array of problems, many of them identified by Mabbott (1951) and
taken up by Dainton. For example, if one starts by defining the
specious present in terms of LA1, a momentary act of awareness, and then
relaxes that assumption, one is led into the absurdity that "the
specious present of an act of awareness varies inversely with the
duration of the act" (Mabbott, 1951, pp. 158-159). Dainton refers to
this as the "ballooning of content" (p. 140).<2> Another problem,
mentioned by Mabbott and analyzed in detail by Dainton, is the problem
of repeated contents. Given an overlapping of specious presents, as
Broad's theory entails, if at t1 in an act of awareness e1 I hear a tone
A, and if in the specious present associated with the next act of
awareness e2 at t2 the sense datum for A is still present, then I hear A
again. On this model we seem to experience the same events multiple
times. Another problem that troubled James and Mabbott, but is only
mentioned by Dainton, concerns empirical evidence that the length of the
specious present may vary across different sense modalities. In other
words, the specious present for vision may be different than that for
audition. Thus, if I am watching a ballet, the music should appear to
be out of sync with the dancers' movements. The dancers would always be
a little behind or ahead of where I think they ought to be.
Are these problems resolved in Broad's later account? Broad fully
embraces LA1 - the idea that acts of awareness are momentary - in his
later model. He also adds the idea that contents have different degrees
of "presentedness," although this concept remains without clear
explanation. A lesser degree of presentedness had by a content in the
specious present means that it is already past to that degree. Dainton
construes this to mean that the content has a different phenomenal
characteristic (p. 145). Nonetheless, the content, even with a lesser
degree of presentedness, is still present, and this implies that LA2 is
still in effect. Maintaining LA1, however, resolves the problem of
ballooning contents, since there are no shorter or longer acts of
Dainton also argues that the problem of repeated contents is avoided by
Broad's later account. Even though a particular content is sensed by a
succession of momentary acts (and since we still have overlapping
specious presents), each successive instance of it appears with a
gradually diminishing degree of presentness. "Although every content is
apprehended by uncountably many different acts, no content appears in
two different acts under the same mode of presentation. So we do not
experience one and the same content repeated over and over; we
experience a single content sinking smoothly into the past" (p. 146).
It is not clear to me, however, why this is the case. For example, if a
sense-datum for tone A at t1, when it is sounded, falls into the
specious present of an act of awareness e1, and we thus hear tone A at
t1, and then at t2 the same sense-datum for tone A, now with a
diminished presentedness, persists in the specious present of an act of
awareness e2, do we not hear it again, this time, however, as a kind of
weaker reverberation from the past? As long as Broad conceives this in
terms of sense-data that are really present in consciousness, the
phenomenology would seem to involve hearing a tone and then continuing
to hear its reverberation, even as new tones are sounded. A clear
resolution to this problem would involve giving up the old theory of
sense-data in consciousness, and clarifying the concept of
presentedness. Ultimately this problem is responsible for sinking
Let me note one final aspect that comes into Dainton's interpretation of
Broad. Dainton defines 'phenomenal content' to include both phenomenal
objects (parts of experience) and phenomenal properties (features of
phenomenal objects) (p. 24). The term 'experience', of course, remains
somewhat ambiguous. Part of what is at stake in Dainton's work is to
clarify some of these concepts. I assume, however, that phenomenal
content is not equivalent to the notion of sense data as Broad used
it.<3> In the context of Dainton's discussion of temporal experience,
and in the context of what he calls the 'A theory' (which is not to be
confused with McTaggart's A-theory), I understand 'phenomenal object' to
mean something like *the object* (e.g., a musical tone) *as I experience
it*, rather than the sense data that stands for, represents, or in some
way participates in the generation of the appearance of the object
(musical tone). Dainton acknowledges something like this in his
analysis of the later Broad insofar as he takes issue with the latter's
anti-realism. To make sense of presentedness, Dainton suggests that
"instead of successive acts being apprehensions of numerically identical
contents, successive acts must be apprehensions of representations of
contents" (p. 147). Dainton, however, thinks this is a change from
Broad's earlier theory. The notion of presentedness is certainly new,
but Broad, insofar as he thinks of phenomenal content as sense-data, is
what Dainton defines as an anti-realist in both his early and late
accounts. Dainton, however, does not seem to notice an important
difference in this regard between accounts given by Broad and Husserl.
3. DAINTON'S INTERPRETATION OF HUSSERL
Dainton may have been influenced in his reading of Husserl by Miller's
(1984) interpretation. He appeals to Miller's "Principle of
Simultaneous Awareness" (PSA). It is possible, however, to distinguish
between a strong version of PSA and a weak version. The strong PSA is
specified as involving LA1 and LA2, the two assumptions that define the
cognitive paradox and that, as Husserl notes, derive from Herbart and
Lotze, and turn up again in Brentano's analysis. The weak PSA lacks
this specification. For example, Dainton offer a weak formulation of
PSA: "If we are directly aware of the immediate past, this awareness is
located in the present" (p. 133). Although Miller declines to give a
clear formulation of this principle, he considers a strong version of
PSA on the basis of Husserl's description of Brentano's position that he
(Husserl) wanted to reject. Husserl formulates these assumptions in the
In order to grasp a succession of representations (a and b for
example), it is necessary that the representations be the absolutely
simultaneous objects of a knowing that puts them in relation [=
LA2], and that embraces them quite indivisibly in a single and
indivisible act [= LA1] (1991, p. 21).
Husserl goes on to explain that for Brentano LA2 requires LA1. Notice,
however, that Husserl's formulation also describes Broad's later theory.
It is important to see that Husserl rejects Brentano's theory in a way
that implies that he would also reject Broad's theory. Specifically, he
rejects LA1 on phenomenological grounds and (although he must struggle
to do so) he also rejects LA2. In doing so, Husserl rejects the strong
version of PSA.
Both Miller and Dainton, however, suggest that Husserl accepts the
strong version of PSA, that is, Brentano's version. Miller writes:
"Like Brentano, Husserl regards PSA as a necessary condition for
temporal awareness" (1984, p. 120).<4> Dainton associates Husserl with
Broad, and makes the same claim: "Broad and Husserl both subscribed to
certain assumptions - specifically, a combination of an
awareness-content model and the Principle of Simultaneous Awareness, PSA
- and both found it hard to develop an unproblematic account of temporal
experience within this framework" (p. 136). Specifically both Miller
and Dainton claim that Husserl (like Brentano and Broad) accepted LA1.
But this is simply not true for either Husserl's early or later
accounts. He had rejected LA1 under the influence of Stern and in
working out his critique of Brentano. In 1905, and thus as part of his
early account, Husserl states unequivocally:
It is certainly evident that the perception of a temporal object
itself has temporality, that the perception of duration itself
presupposes the duration of perception, that the perception of any
temporal form itself has the phenomenological temporality that
belongs to its irreducible essence. (1991, p. 24).
I would argue that Husserl accepts the weak PSA, and specifically as it
is nicely and precisely formulated by Miller. "An awareness of
succession derives from simultaneous features of the structure of that
awareness .... A *continuous* awareness of a tone as enduring must
involve an awareness of (at least) some temporally extended part of the
tone at any given *instant* of that awareness" (Miller, 1984, p. 109).
This describes Husserl's position correctly. One should note that
awareness is continuous rather than momentary (versus LA1), and that,
most importantly, what is simultaneous has to do with structural
features of the act of awareness, rather than with the contents being
experienced (versus LA2). Husserl, in contrast to James, Broad, and
other theorists of the specious present, finds the answer to the problem
of temporal awareness (and an answer to the cognitive paradox), not by
accepting the present simultaneity of successive sensory contents, but
by affirming a retentional-protentional structure of awareness. For
him, the answers to all the problems associated with the specious
present are to be found by looking not to content but to noetic
structure (that is, the structure of the act of awareness). Of course
this means that Husserl still maintains the awareness-content model, but
the complexity introduced by retention and protention is on the side of
awareness rather than on the side of content.
It would be inappropriate here to try to present in all its detail
Husserl's complex analysis of time-consciousness, and my intention is
not primarily to defend Husserl's analysis against Dainton's critical
reading of it. Indeed, there are a variety of problems with Husserl's
analysis that I've addressed elsewhere (Gallagher, 1979; 1998). But I
do think Husserl has some important insights to offer, which Dainton's
presentation missed. To sort this out as efficiently as possible, I'll
briefly discuss three points.
3.1 EARLY VERSUS LATE ACCOUNTS
Dainton is right to distinguish between Husserl's early and late
accounts of time-consciousness. But he offers a misleadingly neat story
about the relationship between early and late accounts in Broad and
Husserl. He contends that Broad's early view is equivalent to Husserl's
later view, and Broad's later view is equivalent to Husserl's early
view. But this schema doesn't hold up if we consider that Broad
maintained LA1 in his later account whereas Husserl consistently
rejected it and maintained that the act of awareness is itself extended
in time. Broad also maintains LA2, specifically the idea that
successive sense-data are presented simultaneously, whereas Husserl
attempted to work his way free of this idea. This is one of the main
differences between Husserl's earlier and later accounts. In 1905
Husserl agreed with Stern's rejection of LA2, but it took him several
years to figure out how to break away from the strong current of this
assumption. The solution was implicit in his early theory, but he still
held to a conception of consciousness that he had developed in his
*Logical Investigations* (1970, original 1900-01). This conception of
consciousness requires that we distinguish between at least two types of
3.2 TWO KINDS OF CONTENT
There are two kinds of elements in consciousness, according to Husserl:
real (*reell*) elements and intentional elements. For example, I see
the cat. The fact that I am perceiving is something that is really
happening and as such it is a real event of my consciousness. The cat,
however, as Aristotle and the medieval theorists of intentionality
assure us, as if we needed assurance on this point, is not *really* in
my consciousness. The cat remains out there on the mat. The cat is an
*intentional*, but not a real, content of consciousness. The cat that
we see is the cat that is out there in the world. The cat that we
imagine or remember may not be out there in the world, but neither does
it have any *real* existence in my consciousness. In all of these cases
it has only an intentional status. But Husserl also thought that there
is *real *content that materializes in consciousness as part of the
process of perception. He started out calling this 'sensation', but in
the end settled on the term 'hyletic data'.<5> Hyletic data are the
pure uninterpreted sense impressions (of color, shape, smell, etc.) that
inform perception. According to Husserl, we are not normally aware of
hyletic data, but we can become conscious of them in phenomenological
reflection. When I perceive something, hyletic data are processed in a
nonconscious way in what Husserl calls the "apprehension - content
schema." So when I see the cat, I am aware of the cat (the intentional
content of consciousness), not the hyletic data, but this awareness is
generated on the basis of a processing of real (hyletic) content of
which I am not ordinarily conscious.
In his early analysis Husserl thought that the retentional aspect of
consciousness, which retains the just past moment of consciousness in
present experience, involved this apprehension - content schema. In
other words, retention depended upon something like a current
micro-processing of hyletic content that originated with the past event
but was in some way simultaneous with the current processing. In that
case, however, LA2 was still in effect. Some real content from the past
was still persisting in consciousness. Husserl finally came to reject
this idea, and it was this rejection that marked his move to the later
account. It is also this rejection that distinguishes his account from
accounts given by James and Broad. In 1909 he writes:
*Do we have a continuum of primary[hyletic] contents simultaneously
in the now-point and, in addition to this and simultaneous with it,
a continuum of 'apprehensions'?* ... [C]ertainly everything that
'really' [*reell*] belongs to this consciousness exists in it
simultaneously - that is to say, exists in it 'now' ... The primary
contents that spread out in the now, *are not able to switch their
temporal function:* the now cannot stand before me as not-now, the
not-now cannot stand before me as now. Indeed, if it were
otherwise, the whole continuum of contents could be viewed as now
and consequently as coexistent, and then again as successive. That
is evidently impossible. (1991, pp. 334-35).
Retention, according to his later theory, does not retain real contents;
it retains intentional contents. It retains the sense (the meaning
content) of what has just consciously passed. Although there is this
change in the status of the content that retention retains, it is
important to note that the structural status of retention does not
change. More generally, for understanding Husserl's analysis, it is
important to know that retention and protention, as they perform their
respective functions of retaining the past and anticipating the future,
are not contents themselves, either real or intentional. Rather, they
are part of the noetic structure of the act of awareness. Retentions
fall on the side of awareness, not on the side of content. This point
is often confused by commentators. Plumber, for example, who takes
retention and protention to be temporal intervals, suggests that
Husserl's concept of the specious present is similar to one proposed by
James and Broad, and involves "an *instant* flanked by intervals of
'retention' and 'protention'" (Plumber, 1985, p. 21n3). In other words,
he understands retentions to be part of what we are aware of, rather
than part of the structure of awareness. Dainton does something
similar. He equates retentions with a sequence of representations which
we simultaneously apprehend (p. 151). But retention is not something
that is apprehended; it is part of the structure of apprehending, if by
that we mean awareness. Even the notion of a retention of retention
does not mean an awareness of a previous retention or a real sequence of
retentions existing in the now of consciousness. We do not hear the
retention of a previous note, for example, we hear the present note as
following the previous one. The specious present, which for James and
Broad consists of a set of simultaneous sense-data paradoxically laid
out in a successive order, is for Husserl an intentional structure in
which the just-past is virtually (not really) retained. As we noted,
Broad excluded the immediate future from the specious present precisely
because he thought one needed an existing sense-datum to apprehend, and
future sense-data (or sense-data of the future) do not yet exist.
Husserl includes protention (anticipation) of the just-future, because
protention does not depend on the apprehension of non-existent
sense-data. Rather it is part of the structure of consciousness that
makes it anticipatory.
If we understand retention and protention in this way the problems of
lingering contents and the problem of the "clogging of consciousness,"
as Dainton identifies them (p. 156), do not apply. It's not that we
continue to *hear* the past note reverberate as the present one is
sounded (that would soon become an auditory jumble); we hear the present
note as following the one we have just heard and as preceding the one we
anticipate. Dainton writes: "If I snap my fingers, I hear the sound of
the snap and it is gone. The snap-sound does not linger on in my
immediate experience" (p. 156). Husserl's concept of retention, in
contrast to Broad's account, does not imply such lingering. Rather,
what is retained is the sense that I have just snapped my fingers.
Dainton writes: "If I turn my head to the right I will eventually lose
sight of the coffee cup to my left. But I do not experience the cup
*fading into the past*, rather I experience it moving to the left ....
When I lose sight of the cup, I do so completely and all at once. The
only 'fading' that occurs is due to the blurring of perception at the
peripheries of the visual field" (p. 156). But this is not an objection
to Husserl's concept of retention. It confuses retention with some
kind of faded image that supposedly would linger on in consciousness, as
if I were still seeing the cup but somewhat out of focus. Retention
retains the sense of my just-past experience of seeing the cup (and just
as clearly as I saw it), but it does not do so by keeping a faded image
in consciousness. The fading aspect of a fading image is not equivalent
to a temporal "fading" into the past.
If retention is thought to depend on the presence of sense-data, as in
Broad, and we add the complexity of Husserl's retentional-protentional
model, then it is understandable how one might think that consciousness
would be clogged with real content (Dainton, 2000, p. 157). In
contrast, on the purely intentional model of retention that Husserl
finally develops, there is no clogging of the works. What is "in"
consciousness in an intentional manner corresponds precisely to what we
experience in our temporal awareness - and nothing more than that. Even
if I am listening to a very fast melody I do not complain that the notes
are clogging my consciousness, at least in the sense that Dainton
suggests. Retentions do not get locked in some kind of traffic jam of
the present. The retentional aspect of consciousness at any one moment
opens up a unitary access to our just-past experience. Even if the
just-past is articulated into a specific sequence, retention provides
access to that temporal articulation without showing itself to be
articulated. Husserl's own phrases are sometimes misleading in this
way. When he speaks of a retention of a retention of a retention, and
so on, this may lead the reader to think that what we experience is a
series of retentions, and that all of these retentions are clogging
things up so that it is difficult to find the experience of the temporal
object itself. This issue leads directly to an important point that
Dainton helps to bring out.
3.3 PHENOMENOLOGICAL DESCRIPTION AND THE PROBLEM OF REIFICATION
Dainton points out the importance of Husserl's notion of the double
intentionality of retention. Retention is first of all an intentional
awareness of the just-past moments of consciousness (Husserl calls this
its "longitudinal" aspect), and secondly, on that basis, an awareness of
the just-past object (an indirect or "transverse intentionality"). Its
primary target is my just-past experience of the object, not the
persisting or changing object itself. Since my just-past awareness had
been an awareness of the just-past object, retention allows for the
continued awareness of the object, as just-past. The intentionality of
retention, which some phenomenologists refer to as functional
intentionality, is not the same as act-intentionality. It does not
thematize its object, as does a full-blown intentional act (e.g.,
perception or memory). It functions more on the order of working
memory, not just in terms of its short-term reach, but in terms of how
it "keeps hold of" the just-past (Husserl, 1966, p. 118).
This solves a problem that Broad had left unresolved. Mundel (1954)
posed the question in just the right way in regard to Broad's analysis.
If awareness is itself spread out in its own duration, and therefore
contains phases of its own, how are these phases synthesized to allow
for the continuity of experience? Husserl had asked precisely the same
question against Stern, approximately fifty years earlier. Stern, like
the early Broad, had maintained that the act of awareness, and not just
the speciously present content, is itself extended in time, but left
this insight unexplained. Husserl's notion that retention has a double
intentionality is an attempt to explain it, and in effect, to explain
the unity of consciousness and the fact that the stream of consciousness
Dainton raises a very good question about all of this: "is it possible
to detect in our own experience the postulated complexes of retentions,
primal impressions, and protentions performing their intricate dance?
It is by no means obvious that we can" (p. 156). He suggests that such
things are part of "a purely theoretical construction going far beyond
the phenomenological data" (p. 159).<6> Husserl claims to be doing
phenomenological description, on the basis of direct intuition; but if
we look closely at our own experience, we do not find retentions, primal
impressions, or protentions swimming around in the conscious stream. We
can push this a bit further. Husserl talks of momentary cross-sections
of consciousness that are structured by retentions, primal impressions,
and protentions. Do such cross-sections actually exist, or are they
simply abstractions, theoretical entities, posited by Husserl? Indeed,
this abstraction is precisely what leads many commentators, including
Miller and Dainton, to think that Husserl retained LA1: the
momentariness of the act of consciousness. As we have seen, Husserl
rejects LA1, but nonetheless (in a way that is similar to Broad, and
similar even to Dainton's attempt to map out his own overlap theory)
insists on analyzing things in terms of momentary cross-sections.
On the one hand, this problem, which is what Husserl calls the problem
of reification, reflects what Dainton identifies as a point made by G.
E. Moore. "[... T]he moment we try to fix our attention upon
consciousness, and to see what, distinctly, it is, it seems to vanish"
(Moore, 1922, p. 25; cited by Dainton, p. 43). This diaphanous nature
of consciousness is something that phenomenologists have to deal with if
they are to say anything other than consciousness is diaphanous. One of
the major (hermeneutical) objections to Husserl's notion of
phenomenological reduction is that in formulating descriptions we are
forced to use language. But language is not purified of theoretical
constructs (which, of course, were meant to be bracketed by the
reduction). Even if we could purify language in the right way, it would
still contain nouns (e.g., 'phase', 'retention'), which, in reference to
the stream of consciousness, might imply substantive parts rather than
transitive parts. Reflection itself may introduce distortions into what
we see in phenomenological intuition, as Husserl warns specifically in
regard to time-consciousness.
We must therefore distinguish: the prephenomenal being of
experiences, their being before we have turned towards them in
reflection, and their being as phenomena. When we turn towards the
experience attentively and grasp it, it takes on a new mode of
being; it becomes "differentiated," "singled out." And this
differentiating is precisely nothing other than the grasping [of the
experience]; and the differentiatedness is nothing other than the
being-grasped, being the object of our turning-towards. (Husserl,
1991, p. 132).
Elsewhere he advises that "One should not reify the structure of
consciousness, one should not falsify the modifications of consciousness
into modifications different in principle, etc." (1991, p. 337,
translation revised).<7> Lotze (1887) had warned against the kind of
reification of experience that takes the form of spatialization (also
see Dainton, pp. 18ff). Husserl points out that reflection tends to
freeze the flow of consciousness and to set it out in discrete parts.
To say that Husserl recognizes the problem is not to say that he avoids
it in all cases. To sort this out I propose that we distinguish between
a *descriptive abstraction* and a theoretical account. Given the task
of describing the stream of consciousness, a certain way of expressing
it either captures it or does not, sinks or swims, so to speak, and this
is something to be worked out, in part, in an intersubjective way. I
think the notion of retention is a phenomenologically legitimate
descriptive abstraction rather than a theoretical solution. The
phenomenology of listening to a piece of music is such that when a
series of notes in a melody are played, for example, I hear the melody
and not just one note and now another, and now another, etc. Previously
sounded notes are retained in the intentional experience so that as I
hear the note that is now being sounded I hear it as part of a
continuity of notes. This happens without persisting sense-data (or
hyletic data) and without my having to activate a memory of previous
notes. We find, in Husserl's texts, phenomenological descriptions of
just such experience, which seem to be very much on the mark. Based on
the experience and the descriptions of it, Husserl proposes the idea of
retention as an attempt to characterize just such aspects of experience.
It's a descriptive abstraction, and the only relevant question for the
phenomenologist is whether it is close enough to the experience or
introduces any distortions. I think Husserl does sometimes describe
things in a way that is too reified - the cross-section of
consciousness, and retentions and protentions as if they were elements
that we could directly experience. In such cases the task is to try to
pull such abstractions back closer to the experience by finding a more
appropriate way of putting it, or by introducing various qualifications.
This might be the beginning of a theorizing process, but it is one that
is phenomenologically generated. In any case, Husserl always intended
phenomenology to be an intersubjective enterprise, open to corrections.
In that spirit, he would welcome any improvements.
4. DAINTON'S OWN ACCOUNT AND THE PRINCIPLE OF PRESENTATIONAL
In the spirit of an intersubjective project, then, I would like to
consider Dainton's own account, which he calls the *overlap model*.
Given his critical remarks on Husserl, I will assume that he intends for
this not to be a theoretical model but a phenomenological account that
is close to experience.
The overlap model seeks to improve on Broad's account. As Dainton
points out, Broad fails to consider what happens to the specious present
when the act of awareness is taken to have its own duration. Like the
early Broad, and following John Foster's model, Dainton begins with the
supposition of a momentary cross-section of consciousness.<8> To
resolve Broad's problem of repeated contents in the overlapping specious
presents, Dainton appeals to Foster's solution: allow for an overlap in
the acts of awareness (which Foster terms 'presentations'). Dainton
provides a detailed analysis, summarized in *Figure 1* [all figures are
available from the html version of this article].
In the diagram e* represents an abstract momentary act of awareness, and
the solid lines of e*AC represent the momentary act with its specious
present as Broad conceived it. The innovative part of Dainton's account
is the idea of extended, overlapping acts of awareness. So e* is
extended as e1 (with the specious present of A-C). It overlaps with e2
and any other act in between. It is not a necessary requirement that
all acts (or all specious presents) have the same duration, but whatever
duration acts of awareness have, their specious presents have
proportional durations. Dainton actually makes this claim stronger by
defining the proportion as one of simultaneous equality. He adopts the
view "that acts of awareness and their contents exactly coincide in
time; they run concurrently" (p. 166). This is what, following Miller,
he calls the Principle of Presentational Concurrence (PPC). Acts of
awareness and their contents share the same temporality - there is no
temporal discordance between acts of awareness and phenomenal contents.
Furthermore, Dainton, following Foster, suggests that this common time
can be matched to objective time (p. 165).
PPC with overlapping awareness solves the problem of repeated contents.
But as it does so, I suggest, it generates a new problem - *the problem
of ongoing contents*. The same continuous content is seemingly present
across a number of overlapping acts. If content C is presented at the
end of e1, it is speciously present thoughout e1. Unless Dainton
appeals to some notion of protention, however, C's presentedness
throughout e1 is supposedly at a constant level, even though C does not
objectively occur until the end of e1. C is also speciously present
throughout e2, but unless we say first as protended, then in a primal
impression, and finally as retended, C will be heard, not only
throughout e2, but from the first moment of e1 until the last moment of
e3. Dainton, however, does not appeal to protention or retention to
sort this out, and I think the problem of ongoing contents remains
unresolved. If, as he contends, the common time of the act of awareness
and its content can be clocked with objective time, then consider what
happens if e1 is 2 seconds long. Two seconds before C occurs, I become
aware of C. My awareness of it continues up until C actually occurs and
for two seconds after C occurs. If we assume C is a momentary event
(e.g., the quickly dampened sounding of a musical note), or that it
lasts for approximately .5 secs (or any duration less than 4 secs),
unless we have some way to distinguish between our anticipation and
retention of it, then it would seem to last for 4 secs, it would be
heard before it actually sounded, and it would continue to be heard
after it was no longer being sounded (the problem of lingering contents
once again). Even if this is consistent with the logic of PPC, it is
not consistent with our phenomenology. What is missing here is some
account corresponding to what Husserl described as the retentional and
protentional structure of consciousness. Overlapping, by itself, just
doesn't capture the nuances of anticipation and retention in experience.
As we see from the diagram, Dainton's analysis starts out by positing
overlapping contents and overlapping acts of awareness. I don't think
he actually means overlapping contents (p. 164). An example of an
overlapping content would be if I am looking at someone as they tell me
their account of consciousness and the phone rings. I see them, I hear
them, and I hear the phone. This kind of thing happens constantly and
is phenomenologically unproblematic, although it may be pragmatically
problematic. It's not just one damn thing after another (as Whitehead
once said in regard to experience) it's too many damn things at once.
Rather than overlapping contents, I think Dainton means overlapping
specious presents, as we find them in Broad's diagram, or more precisely
in Mabbott's interpretation of Broad's diagram (looking at the same
diagram, Mundel, for example, doesn't find these overlaps at all).
Overlapping specious presents in Broad's model lead to some of the
problems that Dainton attempts to solve with overlapping acts of
What exactly are overlapping acts of awareness, however? The problem
here may be just mine. I'm not sure what overlapping acts of awareness
could mean for an individual subject. I can conceive of a temporal
overlap of two or more acts of awareness in the following way. I'm
sitting in my office looking at the ringing phone, for example. You
walk in, hand me a piece of paper, glance at the ringing phone, and walk
out. Your awareness of the ringing phone temporally and temporarily
overlapped with mine. Can something like this overlap happen in one
individual? For example, I am aware of the computer screen in front of
me and at the same time I am aware of the phone ringing on and off. I
would say this is just a phenomenologically unproblematic overlapping of
content, or with Dainton, this is a case of co-consciousness, but not
that I have two separate acts of awareness going at once. In the end,
trying to decide this for acts of awareness may be irrelevant since
Dainton wants to give up the A-thesis, which distinguishes between acts
and contents. Are these problems resolved by what Dainton calls the
If we give up the act-content schema, as Dainton wants to do, and adopt
the Simple Conception, this does not solve the phenomenological
problems. In fact, it makes things worse by adding a conceptual
problem. It robs PPC of any meaning. PPC states that "acts of
awareness and their contents exactly coincide in time" (p. 166). The
Simple Conception allows the distinction between acts and contents to
dissipate. Yet Dainton wants to maintain PPC - "although I shall still
refer to PPC, the latter principle should no longer be taken to imply
the validity of the act-object model" (p. 166). It is difficult to
understand what PPC could mean in this case.
The Simple Conception involves overlapping experiences or phases of
experience. Experiences in the stream of consciousness are not to be
individuated in terms of subjects (pp. 25, 220). Dainton suggests we
individuate experiences by differences in intrinsic or exact phenomenal
character (e.g., if one is of pain and the other is of smell),
differences in their time of occurrence, and/or differences in their
physical basis (p. 25). To stay with the phenomenology we can leave
physical basis aside. We can also eliminate differences in time of
occurrence, since we are trying to understand what overlapping
experiences are, and during the overlap period there is no difference in
time. That leaves exact phenomenal character as the criterion of
In this regard we run into another problem. If exact phenomenal
character refers to "what the experience is like, exactly like,
phenomenologically" (p. 23), this can change from moment to moment, and
experiences might seem to be momentary. More importantly, phenomenal
character seems to involve a holistic aspect of experience. Dainton
writes: "If my visual field were in any way different, it would have a
different phenomenal character. If my visual field had a different
phenomenal character, my overall consciousness would also have a
different character" (p. 24). This understanding of phenomenal
character helps to individuate experiences, but it undermines our
ability to speak of overlapping experiences. If e1, an ongoing
experience, is suddenly overlapped by e2, then my overall consciousness
would have a different character and it would not be a case of e2
overlapping with e1, but e1 being *replaced* by en (the effect of
combining e1 and e2). That is, a new experience rather than two
overlapping experiences would occur, because overlapping experiences
cannot retain their individual phenomenal characters. When two
experiences are in sync, that is, when they overlap, one sinks into the
other and something new surfaces.<9>
Let me conclude by considering the phenomenological adequacy of the
Simple solution to the problems of time-consciousness.<10>
There are two aspects of temporal experience that need to be explained:
the flow of temporal passage and the temporal order of appearances. For
Husserl the retentional-protentional structure of conscious acts
explains our temporal experience - the fact that the flow of experience
flows in a continuous fashion and that things appear in their proper
sequential order. I suggested that we may still need retention and
protention to solve the *problem of ongoing contents* in the overlap
model. Dainton, however, cannot appeal to retentional-protentional
structure since he pursues the Simple Conception and abandons the notion
of conscious acts. Rather, he proposes to solve all problems by
appealing to experiential content.
On Dainton's view, the flow of experience is no problem at all since
experience is intrinsically organized as a flow. This is not primarily
an ontological claim; it's a phenomenological claim. Consciousness just
is a flow of experience because it appears to be - and in phenomenology
appearance is all that counts. Content is not momentary, it endures,
and then it flows into the next content in a unidirectional fashion. My
experience of the content flows in sync with the content, and there is
no lack of coincidence between awareness and contents to worry about.
Furthermore, the phenomenal content of experience has an intrinsic
temporal pattern that presents itself as this unidirectional flow.
Thus, the problem of temporal order is also easily resolved. In the
overlap model, some element of content appears in two overlapping phases
of experience. In the first phase it is sequentially related to other
content in that phase which is not contained in the following phase
(e.g., B follows A), and in the second phase it is related to content in
the that phase which is not contained in the first one (e.g., B precedes
C). These relational differences make all the difference needed for
temporal order. Temporal order is a reflection of the relational
properties of these contents (Dainton, pp. 173-77).
One begins to wonder why Broad and Husserl were so exercised. The way
things seem to be is just the way they are. The flow of experience is
explained by the flow of experience; the temporal order of experience is
explained by the temporal order of overlapping experiences. The problem
of time-conscousness, as it was understood by James, Broad, Stern,
Husserl, and others, dissipates once the A-thesis is abandoned and the
Simple Conception is adopted.
Yet, as I have tried to show, there are some problems with the overlap
model that are left unresolved. Although logically and diagrammatically
we might be able to make sense of an overlap model, if I try to find
overlapping experiences phenomenologically, it seems just as problematic
as trying to find reified retentions and protentions appearing in the
flow. Dainton's criticism of Husserl seems to apply equally to his own
analysis in this regard. Perhaps my defense of Husserl would work
equally as a defense of Dainton: overlapping experiences are simply
descriptive abstractions. In contrast to Husserl's description,
however, according to which I can say that when I hear a piece of music
my experience is that I seem to retain the sense of previous notes in
the melody and anticipate what is to come next, I find it difficult to
say that when I hear a piece of music the current note seems to overlap
with previous and future notes. In the overlap model, for example, it's
not clear why, in a sequence of auditory experiences (or phases of
experience) a1- a2- a3-a4, the fact that experience a3 is just prior to
a4, or that there is an overlap between a2- a3 and a3- a4, explains or
describes anything about the phenomenal character of anticipating the
continuing melody at any moment of the experience.
As Dainton works out his overlap version of PPC, his phenomenological
account relies, at first, on the awareness-content model, the A-thesis.
As I have pointed out, overlapping acts of awareness are
phenomenologically suspect. One possibility is to think of this as
overlapping content of which one is co-conscious - I hear the phone ring
at the same time that I am looking at the computer screen. But
overlapping content doesn't explain the temporality of experience. Even
if we consider the notion of overlapping acts of awareness as a form of
descriptive abstraction, represented in the diagram (fig. 1) for
example, the question is then what happens to this notion in the Simple
model. Consider a diagram of the Simple model (fig. 2).
The overlapping acts of awareness seemingly sink into the overlapping
specious presents (sp1 overlapping with sp2, etc.) of content. One
needs to ask what work is being done by the notion of overlapping acts
of awareness in the first place, if that work can be taken over by
overlapping content. Furthermore, exactly what does a temporal
overlapping of content or experience mean in the absence of awareness?
Is it a real (<i>reell</i>) overlap or an intentional overlap? Or is it
something different? If it is an intentional overlap, how can this be
explained without noetic acts or the structural features of
retention-protention? If it is not an intentional feature, and it
depends on the real presence of overlapping content, then we are back
where we started, with Broad. Or worse, since if experience is now
simply running along the same line as the content, there is no good way
to explain what exactly the overlap is. Unless these issues are
resolved, the overlap model just won't float.
<1> Dainton does introduce Poeppel's analysis in Chapter 7, but only
briefly. For more on the relation between the phenomenology of the
specious present and the evidence from neuroscience, see Gallagher
<2> Mundel (1954) defends Broad against this charge, maintaining that,
for Broad, the duration of the specious present remains constant.
<3> I'm not sure about this, however. Dainton lists sense data, along
with phenomenal objects and qualia-patterns, as things that a
phenomenologist might encounter (p. 19).
<4> Miller goes further and is more specific. He cites Stern's rejection
of "the dogma of the momentariness of a whole of consciousness" (LA1)
and "the necessary isochronism of its members" (LA2), which he calls "a
rough formulation of ... *The Principle of Simultaneous Awareness*, or
PSA" (1984, p. 165). According to Miller, Stern rejected the strong
version of PSA for two reasons. He considered LA2 to be an "artificial"
assumption, but a necessary component of the principle. Stern also
rejected LA1 and introduced the notion of a *Praesenzzeit* - an extended
act of awareness as more realistic. Miller then goes on to claim that
although Husserl rejects LA2, he nonetheless maintained PSA. But this
is not quite the weak version of PSA since Miller thinks that Husserl
still maintained LA1 (and Dainton follows this reading). Miller writes:
"Stern's main reason for rejecting PSA is that the latter involves the
postulation of *instantaneous states of awareness* ('the dogma of the
momentariness of consciousness') [LA1], and he maintained, it seems,
that there are no such states of awareness. This objection, if sound,
is as much an objection to Husserl's theory as it is to Brentano's"
(1984, p. 165). But this is clearly not an objection to Husserl's
theory, since Husserl, following Stern, rejected LA1. If anything,
Husserl had difficulty freeing himself from LA2, although he always
meant to do so.
<5> In the phenomenological tradition there is a huge controversy over
the status of hyletic data, with people like Sartre and Merleau-Ponty
rejecting the very idea for different reasons. See Gallagher (1986) for
more detail on this.
<6> I appreciate Dainton's suggestion here. I have raised similar
concerns (in Gallagher, 1998, pp. 64ff) and I have also made the same
suggestion, that retentions, etc. may be theoretical constructs that
attempt to answer the question: What must the structure of consciousness
be like if it is to produce the kind of experience that we have? I did
this in a paper that I presented to an audience that included Natalie
Depraz and Francisco Varela. They raised convincing objections to this
suggestion, and I have revised my view accordingly (see below).
<7> He also suggests that it would be absurd "to doubt whether in the
end, the experiences [*Erlebnisse*], which pass into the [reflective]
glance, are not changed, precisely through that [glance] into something
totally different" (1962, par. 77).
<8> Since Dainton appeals to diagrams throughout his discussion of Broad
and Husserl, and he doesn't raise objections to their attempt to diagram
their accounts, I find it a bit disappointing that he doesn't offer a
diagram of his own account. So I'll try to provide one (see fig. 1). At
the start of his analysis, however, as I understand it, his diagram is
identical to Broad's.
<9> We could pursue this further by considering "phenomenal
interdependence" and Dainton's final two chapters, but that would
involve a much larger discussion. One might contend that there is a way
to work out the logic of overlapping experiences, distinguishing
sub-parts of total experiences and sorting through the complex kind of
analysis Dainton performs in these later chapters. The results might
show that there is no logical contradiction in this notion. What
remains unclear, however, is the phenomenology of overlapping
experiences. In this case we end up in a position similar to the one
Dainton criticizes in Husserl. To what extent have we reified
consciousness, or to what extent have we substituted theory for
<10> There are other interesting issues that one could pursue. Here are
several examples. First, is Dainton's critique of the A-thesis
adequate? I can think of versions of the A-thesis he does not consider
in section 2.6 - e.g., versions that do not begin with what he calls the
"common option" but nonetheless resemble the S1 A-theory, and that can
successfully satisfy Dainton's objections. Second, if we abandon the
A-thesis, and thereby abandon the notion of acts, how do we handle the
individuation of what Husserl calls "act-characters" - that is, what one
might call attitudinal distinctions between perceptual experiences, vs
memory or imagination, judgment, belief, emotion, etc.? Third, how does
the Simple Conception explain the phenomenal slowing or speeding up of
our experience of time, mentioned by James and many others? Related to
this is a question about the effect of content on temporal form, an
issue pursued by Merleau-Ponty and others. Dainton, in a complex and
extremely fascinating analysis of temporal modes of presentation,
suggests that "the temporal context of an experience *does* impact on
its [phenomenal] character, and so its identity" (p. 230). One could ask
if this could go the other way: could the phenomenal character of
content impact on the temporality of experience? Dainton points out
that a change in a later experience (e.g., a substitution of e4 for e3)
could impact on the phenomenal character of an earlier experience (e1)
even if that later experience is not part of the specious present of the
earlier one. But he then suggests "this result only arises when a
particularly stringent mode of individuating experience is adopted" (p.
232). This implies that it's possible in theory but perhaps not often
(if ever) encountered in experience. There are, however, actual
empirical-phenomenological studies that show that phenomenal character
and experienced temporal order can vary as an effect of variation in
phenomenal character. For example, if in a series of auditory tones
ABCDEF, the phenomenal qualities of later auditory tones, e.g., E and F,
are changed from a higher to a lower specific frequency, the tones that
precede E and F, e.g., CD, although objectively unchanged, may be
perceived in a different temporal order, e.g., ABDCEF, or may become
phenomenally indistinguishable, depending on the change in frequency
(see Bregman and Rudnicky, 1975).
Bregman, A. S. and Rudnicky. (1975). Auditory segregation: Stream or
streams? *Journal of Experimental Psychology: Human Perception and
Performance, 1,* 263-67.
Brentano, F. (1911). *Von der Klassifikation der psychischen Phänomene*.
Leipzig: Duncker und Humblot.
Broad, C. D. (1938). *An Examination of McTaggart's Philosophy*
(Cambridge: Cambridge University Press).
Broad, C. D. (1923). *Scientific Thought* (Paterson, NJ: Littlefield,
Dainton, B. (2000). *Stream of Consciousness: Unity and Continuity in
Conscious Experience*. London: Routledge.
Gallagher, S. (1998). *The Inordinance of Time*. Evanston:
Northwestern University Press.
Gallagher, S. (1986). Hyletic experience and the lived body. *Husserl
Studies 3,* 131-166.
Gallagher, S. (1979). Suggestions towards a revision of Husserl's
phenomenology of time-consciousness. *Man and World, 12,* 445-464.
Husserl, E. (1991). *On the Phenomenology of the Consciousness of
Internal Time (1893-1917)*, trans. J. Brough. Collected Works IV.
Dordrecht: Kluwer Academic; translation of Husserl (1966).
Husserl, E. (1970, original 1900-01). *Logical Investigations*, trans.
J. Findlay. London: Routledge and Kegan Paul.
Husserl, E. (1966). *Zur Phaenomenologie des inneren Zeitbewusstseins
(1893-1917)*, ed. R. Boehm. Husserliana X. Haag: Nijhoff.
Husserl, E. (1962). *Ideas: General Introduction to Pure
Phenomenology*, trans. W. R. Boyce Gibson. New York: Collier Books.
James, W. (1894). The knowing of things together," in *Essays in
Philosophy*. Cambridge, MA: Harvard University Press, 1978.
James, W. (1890). *The Principles of Psychology*. New York: Dover,
Libet, B. (1992). The neural time-factor in perception, volition and
free will. *Revue de Métaphysique et de Morale 2,* 255-272.
Libet, B. (1985). Subjective antedating of a sensory experience and
mind-brain theories. *Journal of Theoretical Biology 114,* 563-570.
Lotze, R. H. (1887). *Metaphysic in Three Books: Ontology, Cosmology,
and Psychology*, trans. B. Bosanquet, Second Edition. Oxford:
Mabbott, J. D. (1951). Our direct experience of time. *Mind 60,*
153-167. McTaggart, J. M. E. (1908). The unreality of time. *Mind
17* (New Series, no. 68), 457-474.
Miller, I. (1984). *Husserl, Perception, and Temporal Awareness*.
Cambridge: MIT Press).
Miller , I. (1982). Husserl's account of our temporal awareness. In H.
L. Dreyfus (Ed.), *Husserl, Intentionality, and Cognitive Science*
(pp. 125-46). Cambridge, MA: MIT Press.
Moore, G. E. (1922). The refutation of idealism. In *Philosophical
Studies*. London: Routledge and Kegan Paul.
Mundle, C. W. K. (1954). How specious is the 'specious present'?
*Mind, 63,* 26-48.
Plumer, G. (1985). The myth of the specious present. *Mind, 94,*
Poeppel, E. (1994). Temporal mechanisms in perception. *International
Review of Neurobiology, 37,* 185-202.
Poeppel, E. (1988.) *Mindworks: Time and Conscious Experience*. Boston:
Harcourt Brace Jovanovich.
Stern, W. (1898). *Psychologie der Veränderungsauffassung*. Breslau:
Preuss und Junger.
Stern, W. (1897). Psychische Präsenzzeit. *Zeitschrift für Psychologie
Ward, J. (1878-89). Psychology. *Encyclopaedia Britannica*, 9th ed.
New York: Scribners.