Experiments About Whorfian Hypothesis Summary Essay

Rev.Phil.Psych. (2019) 10:133?154
https://doi.org/10.1007/s13164-018-0405-4
Early and Late Time Perception: on the Narrow Scope
of the Whorfian Hypothesis
Carlos Montemayor 1
Published online: 12 June 2018
# Springer Nature B.V. 2018
Abstract The Whorfian hypothesis has received support from recent findings in
psychology, linguistics, and anthropology. This evidence has been interpreted as
supporting the view that language modulates all stages of perception and cognition,
in accordance with Whorf?s original proposal. In light of a much broader body of
evidence on time perception, I propose to evaluate these findings with respect to their
scope. When assessed collectively, the entire body of evidence on time perception
shows that the Whorfian hypothesis has a limited scope and that it does not affect early
stages of time perception. In particular, all the available evidence shows that the scope
of language modulation is limited in the case of time perception, and that the most
important mechanisms for time perception are cognitive clocks and simultaneity
windows, which we use to perceive the temporal properties of events. Language
modulation has distorting effects, but only at later stages of processing or with respect
to specific categorization tasks. The paper explains what is the role of these effects in
the context of all the available evidence on time cognition and perception.
1 Introduction
Benjamin Lee Whorf (1956, 57) famously endorsed the claim that Bthe Hopi language
contains no reference to ?time,? either explicit or implicit.^ He interpreted this statement
in terms of what came to be known as the Blinguistic relativity^ thesis, or the view that
language determines how we conceive and represent reality.1 The Hopi, Whorf thought,
There are different readings of Bdetermines^ which are relevant to understand how strong this claim by Whorf
is. I focus on its scope and show that independently of how strongly one interprets this determination relation,
the scope of Whorf?s hypothesis is narrow, at least in the case of time perception and representation.
Establishing this narrow scope is the main goal of the paper. However, I briefly address the issue of linguisticdetermination in the introduction.
1
* Carlos Montemayor
[email protected]
1
Philosophy Department, San Francisco State University, Humanities 388, 1600 Holloway Avenue,
San Francisco, CA 94132, USA
134
C. Montemayor
must have no conception of time, or at the very least, none that a speaker of a language
that represents tense and time markers could possibly recognize. In particular, if
the Hopi language has no explicit or implicit representation of time then it
seems that time is not a fundamental aspect of how the Hopi represent reality.
Thus, the Whorfian hypothesis is the claim that time cognition is determined?
or robustly modulated?by language.
The claim that language determines thought, and through thought, reality, has a
venerable philosophical lineage, and it has been a central topic in psychology, anthropology and linguistics (see Athanosopoulos et al. 2016, 482?483). Stated this broadly,
linguistic relativity predicts language modulation at all levels of cognition and perception. Qualifications need to be in place in order to determine the scope of, nature, and
cognitive requirements for linguistic relativity, and to confirm linguistic modulation
with some degree of certainty. But even at this level of abstraction, the Whorfian
hypothesis has received enough support from recent findings in linguistic and psychophysical studies concerning categorization tasks in bilinguals (Niemeier and Dirven
2000; Athanasopoulos et al. 2015), to make it not only a theoretically insightful view of
language, thought, and cognition, but also a plausibly confirmed one.
What exactly does the linguistic modulation of thought and perception entail? With
respect to its scope, which is the focus of this paper (particularly with respect to time
perception), Blanguage modulation^ has a strong and a weak reading. On one version of
the strong interpretation, language determines how we represent reality in general,
including space and time. This has been an influential way of interpreting the Whorfian
hypothesis. For instance, Ludlow (1999) argues that Whorf is not only right in thinking
that there is a close connection, perhaps a constitutive relation, between language and
thought, but that Whorf?s most ambitious claim is also true, namely that there is an
equally strong connection between language and reality (Ludlow?s book is about the
metaphysics of time, based on the semantics of tense). Ludlow affirms that Whorf
exaggerated the differences between natural languages, thereby favoring a universalist
type of linguistic modulation that is in tension with Whorf?s relativistic view. But a key
aspect of the Whorfian view is preserved in Ludlow?s proposal?language determines,
and is deeply related to, thought and reality. This is a strong and wide-scope interpretation of the Whorfian hypothesis, specifically about the linguistic modulation of
thought and cognition, according to which it ranges over all types of thoughts about,
and representations of, reality.
This argument is theoretically plausible because tense (the distinction between past,
present and future) facilitates thinking in terms of possibilities and counterfactuals that
seem indispensable to how we structure reality. Here a critical assumption is that there
must be some linguistic format for representations of reality, either innate or acquired,
similar to a Blanguage of thought^ that is fundamentally structured in terms of tense. A
difficulty with this idea is that even Fodor (2008), who championed the Blanguage of
thought^ hypothesis, argued for the encapsulation of early perceptual processing
(Fodor 1983), which entails the view that language modulation is restricted in its
scope. So it is not obvious that a Blanguage of thought^ proposal must have the
consequence that linguistic modulation is pervasive.
The wide scope of the Whorfian hypothesis seems to be further weakened by
considerations of human uniqueness since, as reviewed below, there is strong evidence
for the continuity of human and animal time perception, but there is no clear evidence
Early and Late Time Perception: on the Narrow Scope of the Whorfian…
135
of continuity regarding the kind of syntactic language capacity related to tense and
counterfactuals that humans have and any non-human species (Berwick and Chomsky
2016). However, I will not pursue this syntax-based criticism concerning human
uniqueness here. In what follows, I focus on the scope of the Whorfian hypothesis in
time perception, independently of the specific nature of the mechanisms responsible for
linguistic modulation (whether they are innate or acquired, universal or relative,
syntactic, semantic or pragmatic, compositional or iconic, etc.; see Fodor 2008, for
discussion of the distinctions concerning the format of representations that are relevant
for the language of thought hypothesis).2
On a different version of the wide-scope reading, Boroditsky (2000, 2011) has
argued that time cognition and perception are structured through space cognition,
which is deeply influenced by metaphor and language (Boroditsky 2000). Boroditsky,
unlike Ludlow, endorses a strong version of Blinguistic relativity,^ namely, the view
that different languages entail differences in cognition and perception (i.e., they are
different windows to reality), which is the standard way of interpreting Whorf?s
hypothesis. She offers evidence in favor of mechanisms through which languages
and cultures modify and help construct our basic notions of time (Boroditsky 2011).
Most of the evidence she documents, which clarifies crucial aspects of time cognition,
concerns temporal representations in cultural traditions, and explicit judgments of
temporal order and duration. The influence of language on culture and judgments
about time is consistent with the argument presented here. Language modulates
semantic contents in time cognition, such as the categorization of events. Ludlow and
Boroditsky are right about the influence of language on time cognition and on thought
in general. I examine some of these effects of language on time cognition in terms of
what I call Bnarrative effects,^ particularly in section 4. But as the more comprehensive
body of evidence reviewed below shows, time perception is not modulated by language
categorizations or semantic influences at the early stages of processing, at least not in
any robust way.
More recently, Bylund and Athanasopoulos (2017) argued that time representation is
malleable and that language is a main source of its malleability. This is how I shall
interpret the Whorfian hypothesis, namely as the view that time perception and
cognition are influenced and modulated by language, in general, which includes all
stages of processing. Does this interpretation capture the proper scope of language?s
influence on time cognition and perception? This paper argues that the scope of
language?s influence on time representation is limited. Thus, the claim I defend here
is much more specific than an overall challenge to linguistic relativity as a general view
of the relation between language and cognition. I shall focus exclusively on time
perception, and show similarities and differences with other perceptual abilities,
Because of the innate nature of the Blanguage of thought,^ at least as presented by Fodor, this version of the
dependence of though on language is unlike Whorf?s original formulation, which postulates dependence of
thought on specific spoken languages, thereby making it incompatible with a universalist interpretation. The
debate on language-dependence and the Whorfian hypothesis rarely makes this distinction explicitly, which
corresponds to Chomsky?s (1986) distinction between I-language (internal representations of language) and Elanguage (external language), which correlates with the distinction between competence and performance. As
far as the debate is concerned, many of the findings on linguistic relativity may indicate aspects of
performance, rather than competence, but I shall not elaborate on this objection here because my main goal
is to show that language does not modulate early stages of time perception, regardless of the universality or
relativity of linguistic modulation.
2
136
C. Montemayor
particularly vision. But I will not argue against the overall plausibility of linguistic
relativity in other domains.
Based on all the available evidence, I shall argue that although language affects the
categorization of events and thoughts about time, it has no direct influence on the early
stages of time perception, which are determined by duration and simultaneity perception mechanisms that we share with many species (i.e., cognitive clocks and simultaneity windows). Early time perception is dedicated to encode and map the basic metric
features of the environment, in particular, the objective temporal structure of events, as
determined by simultaneity and duration. In reviewing the literature on late time
cognition, I shall interpret the influence of language in terms of the influence of explicit
judgment and symbolic representations on time representation, through categorization.
The influence of language produces distortions of accuracy and creates language-based
temporal structures, such as narratives. Evidence shows that the reliability of early time
perception is confirmed across species, unlike the flexible and distorting effects of
language on human time cognition and categorization, which cannot be decisively
confirmed in other species.
Thus, I propose to evaluate the available evidence for linguistic relativity in the
context of a broader body of findings, which include the vast literature on animal
cognition. I will not challenge the plausibility of the general hypothesis that language
modulates thinking, including thinking about time.3 Likewise, I shall not argue against
the plausibility of linguistic relativity in other domains, including the sense modalities
(although I draw an analogy with vision below). My purpose is to show that, in light of
all the available evidence on time perception, in humans and other species, there is a
compelling case to be made in favor of a narrow-scope reading of Whorf?s hypothesis. I
am neutral about whether or not linguistic relativity holds in many other domains.
A key problem one faces in trying to account for the mechanisms responsible for
linguistic modulation is that the literature on linguistic relativity has focused on
providing evidence, most recently in interdisciplinary studies on bilingualism
(Athanosopoulos et al. 2016), rather than on specifying mechanisms for modulation.
For instance, Pavlenko (2009) presents a theoretical framework for understanding a
large number of studies supporting linguistic relativity, showing that the sharp dichotomy between shared and non-shared concepts in language learning is unjustified. But
the specific mechanisms underlying linguistic relativity are poorly understood, even in
the context of conceptual cognition. Most of the debate has focused on the relevance of
concepts as guiding principles in a multiplicity of tasks, for instance by appealing to
some kind of labelling of stimuli relevant for applying prior knowledge of semantic
cues to sensorial stimuli (Lupyan 2012). However, this issue may be understood
entirely in terms of conceptual judgment and semantic categorization, which has a
clear impact on perceptual judgment, but need not entail the modulation of the entire
cognitive system?including the perceptual system?through the selective influence of
different languages.
3
As mentioned, Fodor, who defended the language of thought hypothesis, argued that the scope of language
modulation is restricted (1983, 2007, 2008). Fodor proposed a universalist view, but I think the issue of
modulation is independent of universalism or relativism and that this issue requires more careful analysis than
it has been given. In any case, the focus of this paper is the limited scope of language modulation, as proposed
by Whorf?s linguistic relativity view.
Early and Late Time Perception: on the Narrow Scope of the Whorfian…
137
Conceptual modulation is certainly a critical issue in language learning, concept
acquisition, and the semantic categorization of stimuli. But the valuable evidence
gathered in favor of linguistic relativity needs to be interpreted in terms of specific
mechanisms for determining or modulating early perceptual stages, and this is not well
established (see for instance Firestone and Scholl 2016, for multiple methodological
problems regarding experiments allegedly showing top-down perceptual modulation).
To ameliorate this problem concerning the lack of specificity concerning the mechanisms for linguistic modulation, I draw an analogy between time perception and visual
perception, in order to gain some insight into how modulation might occur in time
perception (whether the same process could apply in all cases of perception). I argue
that there is a reasonable possibility that the early stages of time perception are not
modulated by linguistic influences and that this might be different in other modalities,
including vision.
2 The Encapsulation Argument for Time Perception
In order to get clarity with respect to what linguistic modulation means in terms of
cognitive architecture, I propose an analogy with a much more debated and better
understood distinction concerning the nature of visual perception. In particular, I defend
the following argument, based on the architectural analogies between time perception
and visual perception, but with the qualification that the Whorfian hypothesis might be
more plausible in vision than in time perception, because the robust semantic features
of the contents of visual experiences, which drive visual attention for categorization,
might be susceptible to linguistic modulation:
(1) Time perception is analogous to visual perception in the sense that it is useful to
draw a distinction between early and late stages of processing.
(2) If (1) is true, then some forms of time perception must be more cognitively
encapsulated than others.
(3) Encapsulated time perception demonstrates that the influence of language on time
representation is limited to later stages of processing.
(4) Therefore, the Whorfian hypothesis has a narrow scope.
(2) and (3) are intensely debated with respect to visual perception, and it seems that
both parties to the debate have good reasons to defend either pervasive Bcognitive
penetration^ and language modulation, or encapsulation and a limited scope for topdown modulation (for review, see Montemayor and Haladjian 2017). But (1?3) are not
very controversial in the case of time perception, as I hope to show below. The main
reason for encapsulation in time perception is to preserve invariances that are
metric in nature?they concern the objective and invariant spatio-temporal
structure of external events in terms of Bmagnitude-based^ representations
(Gallistel 1990; Magnani and Musetti 2017; Montemayor and Balci 2007;
Walsh 2003). Encapsulation (i.e., the notion that processing is performed
automatically, and independently of other cognitive sources of information)
guarantees that early time perception is reliable. (4) is more surprising and
controversial, but it must be accepted once (1?3) are shown to be correct. I
138
C. Montemayor
shall first justify the relevance of the Bearly and late^ distinction for debates
about time perception.
It is important to emphasize that the analogy with vision is not perfect. As mentioned, visual perception is robustly determined by semantic contents and it may be,
although this is controversial, that even low stages of visual perception are susceptible
to the influence of linguistic modulation, inference or judgment. But time perception is
different because the early stages are entirely devoted to metric mappings. These are
mappings concerning the spatio-temporal features of the environment, independently of
conceptual or linguistic capacities, as verified by the findings on animal cognition,
particularly concerning navigation and decision-making skills (this evidence is
reviewed below). So I use this analogy in order to motivate the distinction between
early and late in the context of time perception (a distinction that has been
useful in vision science) and then show why time perception at low stages is
encapsulated in ways that other forms of perception might not be, including
vision. Time perception might be a type of perception in which the distinction
between early and late is uncontroversial.
The main function of early time perception is to satisfy a metric constraint for
reliable navigation and motor control. By the Bmetric constraint^ on early time
perception, I mean that perceived duration and simultaneity must reliably map the
essential structural features of the environment. The purpose of early time perception is
to satisfy the metric constraint by preserving and encoding temporal constancies
regarding the objective features of the environment. This is the main function (the
proper function) of early time perception.
Late time perception, by contrast, violates the metric constraint in the sense that its
main function is to integrate temporal structures in terms of linguistically based
representations and symbolic formats, paradigmatically, explicit judgments. By
Bjudgment,^ I mean the essentially linguistic capacity to epistemically evaluate semantic contents in a symbolic format of representation, which includes categorization and
inferential capacities. Several types of judgments allegedly demonstrate the influence of
language on time perception (this is what the Whorfian hypothesis predicts). I will
argue that such linguistic influences affect only either late time perception or categorization tasks that are more properly understood in terms of visual judgments or
inferences. Late time-perception is valuable because it allows us to create a meaningful
narrative of experiences and events, rather than to accurately map external features.
There is broad consensus that visual perception, one of the most studied types of
perception, occurs at different stages of processing. What is controversial is whether or
not there is Bcognitive penetration^ (i.e., the influence of later, semantic and epistemic
stages, on early, causally determined and more automatic stages). Regardless of which
view one favors, a plausible way of categorizing these stages is in terms of early and
late visual perception (Pylyshyn 1999). Although it is not entirely uncontroversial, this
distinction has helped clarify important aspects of perception, and has played a major
role in recent debates on the cognitive impenetrability of early sensorial processing
(Raftopoulos 2014). It is also a central assumption in debates concerning the distinction
between perception and cognition, iconic perceptual content and propositional content,
as well as the relation between the modularity of mind and concepts (Block 2014).
Language involves capacities for the formal formatting of sounds, symbols and
contents. I shall assume that syntax is the most distinctive aspect of human language,
Early and Late Time Perception: on the Narrow Scope of the Whorfian…
139
which hierarchically embeds contents according to systematically applied rules. But
nothing substantial depends on this characterization. Everyone agrees that what distinguishes language from perception is its unifying semantic role in all aspects of
cognition, through inference, judgment and symbolic manipulation. Language is, by
definition, a non-encapsulated system in the sense that it manipulates highly integrated
information, susceptible to all forms of conceptual influences. This is all that is required
to understand the Whorfian hypothesis for our purposes.4 This hypothesis claims that
language influences time perception and cognition, in general. Is the scope of the
Whorfian hypothesis global or limited when it comes to time perception and cognition,
and if limited, how limited?
Like visual perception, time perception involves multiple stages of processing. From
the millisecond range in which simultaneity and within-modality temporal order is
processed, to cross-modal coordination and suprasecond interval timing, time perception involves a vast array of cognitive processes. Analogously to the perceptual
processing of other magnitudes (e.g., space and rate), time perception requires cognitive
integration at different stages of processing and crucially, across perceptual modalities.
With respect to its essentially cross-modal character, time perception is analogous only
to space perception. This characteristic of time perception presents a difficulty that is
similar to the Bbinding problem^ in visual perception (Treisman 1996, 1998; Clark
2000), but with the essential involvement of cross-modal information. At early stages,
time perception is devoted to the accurate mapping of the temporal features of the
environment, in order to eventually Bbind^ them into events. At later stages, the
influences of intentional action, memory and emotions have an impact on time
perception. What about language?
Bylund and Athanasopoulos (2017, 911) report results that they claim Breveal the
malleable nature of human time representation as a part of a highly adaptive information processing system.^ As mentioned, the authors propose this as a confirmation of
the Whorfian hypothesis that language modulates time perception. Without qualification, this is an endorsement of the wide-scope reading. I shall argue that the Whorfian
hypothesis may be true with respect to late time perception, but that it cannot be true
about early temporal processing, which is required for reliable navigation and motor
control, thus defending a narrow-scope reading.
A potential worry is that what I define as Bearly time perception^ is not what authors
writing on this issue, such as Bylund and Athanasopoulos (2017), consider time
perception. Here the problem is that it is not clear what exactly is the scope of these
authors? claims, for instance regarding the Bhighly adaptive information processing
system^ constitutive of temporal representation. Focusing on this claim, and given that
the Whorfian hypothesis is typically interpreted with a wide-scope reading, it is
plausible to interpret Bylund and Athanasopoulos as endorsing this reading. Moreover,
they characterize time representation as an adaptive system, in general and without
qualification. If Bylund and Athanasopoulos mean something different, say, that only
later stages of processing are influenced by linguistically determined judgments and
4
A precise definition of language, as opposed to other forms of communication, is required to rigorously
assess the Whorfian hypothesis. Language, however, is notoriously difficult to define (i.e., is it a single
capacity or multiple capacities? Is syntax what characterizes it; if not, what characterizes it? Is it uniquely
human? Is it essentially representational? Is it essentially compositional (and/or conceptual) and if so, how?)
This cluster of problems presents other major difficulties that I shall not pursue here.
140
C. Montemayor
concepts?or that language modulation only applies to inferential and categorization
tasks?then they agree with the main claim of this paper, that the Whorfian hypothesis
has a narrow scope. This interpretation, however, would seem to be in tension with
their claim that the system for time representation is flexible and influenced by
language, in general. For this reason, I take their view to be an illustration of the
wide-scope reading.
The importance of the recent findings on bilingual subjects is that they allow us to
compare the type of tasks used in these experiments with the tasks used in more
traditional time perception experiments. The key issue is that the tasks used to verify
the Whorfian hypothesis invariably depend on forms of inference, categorization or
judgment, either implicitly or explicitly. The important findings showing linguistic
effects on duration estimation in bilinguals (Bylund and Athanasopoulos 2017), seem
to confirm linguistic relativity at the visual-temporal categorization level. But it
is clear that in such visual tasks, stimuli are categorized by the subjects in
terms of concrete semantic cues, albeit in an implicit manner. This is in stark
contrast with the traditional findings on time perception, which do not depend
on visual categorization tasks, or any type of categorical or symbolic inferences
at all, and which support the main claim of this paper concerning the narrow
scope of the Whorfian hypothesis, as I proceed to explain.
3 Time Perception: Early Processing
Crucial support for the distinction between early and late-time perception, as well as for
the encapsulation of early stages of time perception, comes from research on animal
cognition and on its verified continuity with human time perception. A very substantial
percentage of the findings on interval timing comes from animal research. To give a
very brief illustration, reliable behavior based on interval-timing skills has been
experimentally confirmed in invertebrates such as honey bees, (Renner 1960) and
bumble bees (Boisvert and Sherry 2006), a wide variety of vertebrates such as goldfish
(Drew et al. 2005), rats (Calvert et al. 2010), starlings (Brunner et al. 1992), pigeons
(Cheng and Roberts 1991; Mazur 1991), and primates (Platt and Ghazanfar 2010).
There are well-confirmed similarities concerning interval timing between humans
and other species, such as mice (Balci et al. 2009; Gallistel 1990). Some findings
suggest that interval timing (i.e., timing that resembles estimating an interval with a
stopwatch) may occur in invertebrates (Boisvert and Sherry 2006), but interval timing
may be very rare in invertebrate species (Craig et al. 2014). Unlike circadian rhythms,
which are ubiquitous in nature, interval timing may not be ubiquitous even among
vertebrates (Craig et al. 2014), although it has been identified in birds (Buhusi and
Meck 2005; Merchant and de Lafuente 2014). The vast evidence collected through a
wide variety of multiple experiments, in humans and non-human species, shows that
what unifies the interval and circadian systems is that both of them comply with the
metric constraint, because the purpose of these systems is to make reliable navigation
possible (for review see Gallistel 1990).
The differences between interval and circadian timing may involve distinct metric
formats for perceiving time, with non-conceptual contents at the earliest stages of
processing (see Peacocke 1992, for the notion of Bscenario content^). But their main
Early and Late Time Perception: on the Narrow Scope of the Whorfian…
141
function is the same, namely to reliably map the objective temporal features of external
events. It is this early type of time perception that is in charge of satisfying the metric
constraint. Both systems, for instance, depend on cognitive clocks, the circadian and
interval clocks (see Meck 1996; Montemayor 2013). The evidence strongly
suggests that interval timing operates independently of language in humans
because the interval clock works reliably and independently of language in
other species, and because there are robust similarities between interval timing
in humans and other species (Meck 1996).
Attention to perceptually represented magnitudes for motor-control and navigation
(e.g., duration, distance, or rate) differs from attention to the duration of sensations and
emotions, including experienced effort, as will be shown below. Among the evolutionarily oldest forms of early perception is the perception of magnitudes for navigation.
The language capacity cannot be assumed to be present in any species other than
humans. This strongly suggests that any effects of language on time cognition are likely
to be found in their full extent only in humans, but not at the early stages in which we
find strong continuities with other species. The vast amount of findings on time
perception for navigation, planning, decision-making, and motor control across species,
including humans, speaks in favor of the encapsulation of early processing, beyond the
reach of language.5 What unifies this reliable form of time perception across species is
the metric constraint.
Early stages of time perception are very reliable, and as mentioned, the clock and
memory mechanisms involved are remarkably similar across species (Gallistel 1990).
Such reliability indicates that early time perception is not susceptible to late, or topdown influences?it is encapsulated in the sense that it operates independently
of other cognitive systems (see Raftopoulos 2014). Time perception at early
stages includes optimal responses to complex stimuli involving comparisons of
intervals, addition and subtraction of intervals, spatial calculations involving
temporal representations, and combinations of rate and probability assessments
(Mazur 1991, 2000, 2007; Balci et al. 2009).
With respect to cognitive architecture, it has been proposed that subsecond interval
timing is performed by intrinsic and local mechanisms, and that a modally independent
mechanism for interval timing is dedicated to register suprasecond intervals (Ivry and
Schlerf 2008; Coull et al. 2004). Evidence shows that there is a core mechanism for timing, which receives information from modally-specific areas
(Merchant et al. 2013). Besides the importance of the circadian clock in time
perception for navigation and planning (Gallistel 1990), an interval clock is
devoted to timing shorter events (Meck 1996).
The modular and cross-modal mechanisms for time perception operate automatically
and reliably. First, very brief intervals of time between perceptual events are perceived
as simultaneous in different modalities, and with slightly different durations, called
Bsimultaneity windows^ (P?ppel 1988). Then there is a cross-modal window of
simultaneity that integrates information from the within-modality simultaneity
5
For more on this issue see (Montemayor 2010, 2013; and Montemayor and Haladjian 2015). In particular,
the evolutionary approach to early and late-time perception can allow for important new insights concerning
cognit