Awareness: More Than Meets the Eye
A Review of Consciousness Lost and Found: A Neuropsychological Exploration by Lawrence Weiskrantz
Trevor Chong
Faculty of Medicine
Monash University
Clayton VIC 3168
AUSTRALIA
tcho5@mfs06.cc.monash.edu.au
Copyright (c) Trevor Chong 2001
PSYCHE 7(01), January 2001
Previously held: http://psyche.cs.monash.edu.au/v7/psyche-7-01-chong.html
KEYWORDS: awareness, blindsight, amnesic syndrome, double dissociations, covert processing, commentary, primary visual cortex.
REVIEW OF: Larry Weiskrantz (1997). Consciousness Lost and Found: A Neuropsychological Exploration. Oxford: Oxford University Press.
ISBN 0198523017. 272pp. US$30.00 hbk.
1. Introduction
Larry Weiskrantz has made a substantial contribution to the field of
neuropsychology over the last 30 years. In particular, his seminal work
on blindsight and the amnesic syndrome has led to the gradual acceptance
by the scientific community of what were previously dismissed as mere
artifacts of poor experimental design. Today there is an explosion of
interest in the covert processing found in such neuropsychological
deficits in the absence of awareness. It is rather timely, therefore,
that a review has been written to summarise the major and rapid advances
in research on blindsight and amnesia. Although this is essentially
what this latest monograph of Weiskrantz's (appropriately subtitled, "A
Neuropsychological Exploration") represents, the book also contains a
statement of the author's views on awareness and speculations on how
research in this area should continue. In some parts, this book reads
like an updated version of Weiskrantz's 1986 monograph on blindsight
(Blindsight: A Case Study and Implications), only vastly simplified in
order to be accessible to a non-specialist. In particular, the central
theme of this book (commentaries as being crucial for awareness) appears
to be an elaboration of the final chapter in the 1986 monograph (on
commentaries).
Weiskrantz confronts a number of formidable issues of consciousness in
this book, but frames them all in the context of patient performance in
neuropsychological deficits which demonstrate a double dissociation.
For example, blindsight subjects may actually be able to detect and
discriminate visual stimuli of which they are unaware in a two-
alternative forced-choice paradigm, while amnesic patients may be able
to retain information which they do not realise is in memory. Hence,
the title: by "lost", he means the loss of awareness in these
neuropsychological syndromes, and by "found", he refers to the retention
of a capacity for covert processing.
His review of previous research, mainly in blindsight and amnesia,
culminates in an argument on the ontological nature of awareness itself.
Essentially, the argument is that the ability to render a parallel
acknowledged commentary is a sine qua non for consciousness.
Weiskrantz uses the many examples of perceptual or attentional deficits
in the literature to bolster this argument and, based on neuroimaging
techniques, even speculates on sites which may be integral to the
generation of awareness and the possible experiments which might reveal
these sites. He also broaches a number of subsidiary but nevertheless
relevant issues, namely those of animal awareness and the evolutionary
value of awareness.
The organisation of the book sees Weiskrantz switch from topic to topic,
with some degree of repetition between chapters (which was deemed
necessary for each chapter to be intelligible on its own). As such,
this review will highlight features of the book according to their
conceptual significance, rather than in the strict order in which they
arise.
2. Surveying the Dissociative Disorders
The first two chapters are spent orienting the reader to the general
neuropsychological deficits which reflect "performance without
awareness." Of course, today it is acknowledged that covert processing
occurs in a great many, if not all, perceptual and attentional deficits.
Unsurprisingly, Weiskrantz chooses to head the list with the amnesic
syndrome and blindsight, on which Weiskrantz is arguably the foremost
authority. Other deficits are not ignored, however. There are
references in the text to such disorders as unilateral neglect, blind
touch (numbsense), deaf hearing, prosopagnosia, anosagnosia, acquired
dyslexia, aphasia and achromatopsia; and the performance of
commissurotomy (split-brain) patients is also addressed briefly in the
introductory chapters. It is commendable that such a wide range of
disorders are sampled in the text, albeit in considerably less detail
than the author's two main preoccupations (blindsight plays an
unquestionably dominant role in the book). But this was all to
Weiskrantz's advantage, for the data gleaned from those who suffer from
these other deficits only offer additional support for the argument
which he later advances.
Having said that, it is rather unfortunate that the author chose not to
elaborate on some of the syndromes mentioned. For example, some
researchers of unilateral neglect (which is essentially an attentional
problem) suggest that it may offer significant insight into the neural
basis of visual awareness, especially when contrasted with a
perceptual deficit such as blindsight (e.g., Driver & Mattingley,
1998). Weiskrantz does briefly allude to the differences between the
two, when he describes blindsight as a disconnection between visual
inputs and a commentary system, and unilateral neglect as a bias within
the commentary system itself (p. 220). However, further explicit and
somewhat less abstract contrasts might have reinforced some of the
propositions that he advances (e.g., that V1 is necessary but not
sufficient for visual awareness). It would be unfair, however, to
criticise the author for negligence in having to deal with so many
deficits. Clearly blindsight and amnesia were chosen to play central
roles in the book and Weiskrantz does offer his apologies in advance for
any omissions or over-simplifications of his colleagues' research
(p. 5).
As far as the centrepieces of this monograph are concerned, the author's
reviews are succinct and accurately encapsulate the more significant
research in each area. The amnesic syndrome was probably the first to
be seriously considered in the context of performance in the absence of
awareness, with some amnesics (the famous HM, for one) now known to show
excellent residual retention of motor skills (such as following a track
on a pursuit rotor) and learnt pictorial or verbal material, as well as
the ability to be primed and classically conditioned. The observation
of a syndrome with such double dissociations has had a major impact on
theories of memory, leading to the highly influential theory of multiple
memory systems.
A similar double dissociation was found in the celebrated blindsight
subject, who is capable of detecting and discriminating certain aspects
of a visual stimulus in the absence of awareness. Much has been learnt
about blindsight since the publication of the author's 1986 monograph.
Indeed, the extensively tested subject of that book (DB) would appear to
have disappeared from the blindsight scene, to be replaced by the
long-suffering and much sought-after "professional subject," GY. The
review of the blindsight literature synthesises some of the older
research with more recent findings into a coherent whole, giving due
credit to several well-established researchers in the area, such as Alan
Cowey, Petra Stoerig and Tony Marcel, to name but a few. In his broad
survey of the literature, Weiskrantz describes the residual visual
abilities of blindsight subjects (e.g., in spatial localisation;
orientation, colour and form discrimination; motion detection and
direction discrimination), as well as the results of numerous animal
studies by Gazzaniga and Nakamura and Mishkin, for example.
The fact that blindsight commands most of the author's attention is
unsurprising and largely warranted. After all, the literature
demonstrates that blindsight is an example par excellence of a double
dissociation. Coupled with the fact that the visual system is one of
the most extensively studied sensory modalities in humans and other
primates, blindsight has a great potential to reveal something about
visual awareness; in particular, the neuroanatomical structures that
give rise to it. In this monograph, blindsight data facilitate a
discussion on neural pathways of awareness, while amnesia provides
insight into the nature of the patients' commentaries.
However, as Weiskrantz would be the first to point out, neither the
amnesic syndrome nor blindsight has always been accepted as valid
instances for covert processing in the absence of awareness. Indeed,
there have been (and continue to be) numerous attacks on the legitimacy
of blindsight as an example of performance without awareness (e.g.,
Campion, Latto, & Smith, 1983; Fendrich, Wessinger, & Gazzaniga, 1992).
For example, it has been claimed that the above-chance performances of
blindsight subjects represent nothing more than the artifactual effect
of stray light; spared islands of visual cortex; or differential
response criteria. Blindsight has also been described as nothing more
than degraded normal performance.
We see in this monograph some evidence of the exasperation that has
arisen from a lifetime of having to defend against such criticisms and
of having to dispel the skepticism and frank disbelief initially
expressed by the scientific community. For instance, he writes "These
were healthy responses... although they were pursued in some instances
with unseemly zeal and persistence beyond the point of positive
relevance" (p. 37). In fact, to digress slightly, there are many
examples in the book of facetiously snide remarks towards his critics or
to those whose opinions he challenges. Witness, for instance, the
author's rebuff of Dennett's hypothesis of multiple drafts: "Nor will
the blindsight subject wish to have visual awareness dismissed as a by-
product of multiple drafts.... He is likely to be a multiple draft-
dodger" (p. 202).
To return to Weiskrantz's main line of thought, both blindsight and
amnesia (along with the spectrum of other neuropsychological deficits)
are deficits which have resulted in a loss of some aspect of awareness.
Blindsight subjects are unable to acknowledge awareness of a visual
event, while amnesic subjects are unable to acknowledge awareness that
retrieved information can be accessed. Therefore, these subjects may be
able to perform in forced-choice tasks at a better-than-chance level,
but they are unable to provide a parallel acknowledged commentary of
their performance. This is the hallmark of all of the syndromes
considered in the book. We are thus brought one step closer to the core
of Weiskrantz's argument.
3. The Commentary and Its Relation to Consciousness
Weiskrantz argues that commentaries (or the lack of them) are crucial in
studies on consciousness because they provide the means by which we
decide whether or not a subject is aware of an event. As such, the
commentary-key paradigm figures prominently in a number of chapters.
This paradigm has been instrumental in research on blindsight and it is
appropriate that its success be described here by the man who first
proposed it (Weiskrantz, 1986). In this paradigm, the subject is told,
as usual, to guess which of two alternatives was presented (e.g.,
horizontal vs. non-horizontal movement of a visual stimulus). The
innovation involves a second set of response keys ("commentary keys"),
through which the subject indicates his awareness of the visual event
(indicating "yes" for any experience whatsoever of the stimulus and "no"
only if he had no experience at all).
This paradigm has allowed numerous advances to be made in understanding
the differences between the processing of visual information of which
the blindsight subject is and is not aware. Responses to moving stimuli
are particularly useful in this regard, for it has been shown that
subjects such as GY can actually report some conscious experience of a
sharply transient transient or rapidly moving visual stimulus. By
applying the commentary-key paradigm to moving stimuli (whose stimulus
parameters may be adjusted), it has been possible to precisely compare
GY's performance in the aware mode (involving high velocity visual
stimuli) with his performance in the unaware mode (involving low
velocity stimuli). Doing so has confirmed that GY can perform very
accurately even when he reports being completely oblivious of the
stimuli that had been presented.
The versatility of the commentary-key paradigm is also highlighted as
Weiskrantz takes a chapter-long foray into the slippery area of animal
consciousness. The investigation of animal consciousness hinges on
behavioural studies, but the problem of such studies is usually that
awareness does not always accompany a behavioural response. Here,
again, it is the commentary key to the rescue. Since first proposing a
gedanken experiment involving apparently blindsighted monkeys and the
commentary key in his 1986 book, such an experiment has actually been
conducted and it demonstrates the close parallels between human and
animal blindsight. While variations of the commentary-key paradigm have
been applied to other neuropsychological syndromes such as amnesia (both
in animals and humans), the remaining syndromes remain largely
unexplored in animals, due mostly to a difficulty in finding the
appropriate animal correlates to test.
Despite the seemingly diverse range of neuropsychological deficits in
which covert processing has been found to occur, Weiskrantz proposes
that they all reflect the fact that commentaries are a sine qua non of
consciousness. This is the central argument around which the entire
monograph revolves. In both blindsight and amnesia, the ability to
render a commentary is lost. Consider memory, for example. Weiskrantz
argues that it is the interaction between current and stored information
that yields a commentary in normal healthy subjects. In the amnesic
syndrome, when the ability to provide such a commentary is lost, the
amnesic may be able to access stored information, but not in remembered
form.
Similarly, in blindsight, the fact that above-chance discrimination of a
visual stimulus is not accompanied by awareness implies that,
"awareness always requires a parallel and separate response to the
discrimination itself for it to be identified and acknowledged as such
by the subject" (p. 76, italics his). This general hypothesis can be
interpreted in two forms. The weak form of the argument suggests that
commentaries simply enable an acknowledgement of awareness to occur,
while the stronger states that awareness arises out of the very ability
to make a commentary (i.e., this ability endows a creature with
awareness).
Although it is impossible to adjudicate objectively between these
versions of the argument based on current evidence, Weiskrantz does
state his preference for the stronger (and more challenging) version.
Indeed, the weak form of the argument seems almost intuitively obvious.
It could perhaps even be said that this version leads us nowhere in
understanding the nature of awareness, given that ex hypothesi the
commentary is formed post hoc -- awareness must occur before one can
make a commentary to acknowledge it. Regardless of which position one
adopts, however, this general argument of commentaries as being a sine
qua non for awareness is certainly consistent with most of the
psychophysical and behavioural data on the neuropsychological deficits
described.
4. The Evolutionary Value of Consciousness
The disabilities that are incurred upon suffering such
neuropsychological deficits are reviewed in a chapter on the
evolutionary value of awareness. It is clear that sans awareness, one
is without the ability to manipulate explicitly thoughts or images for
the purposes of survival. While the residual processing that remains
following, for example, a striate cortical lesion or a temporal lobe
lesion is impressive, it cannot be used in thinking or activity by the
subject, who must be forced to provide their above- chance responses
(p. 66). Prompted by this observation, Weiskrantz asks the practical
question of whether it is possible to restore overt processing in these
syndromes -- a question that has obvious implications for patient
rehabilitation and retraining. As far as most of the neuropsychological
deficits are concerned, especially amnesia, Weiskrantz expresses his
pessimism about finding a means to do so.
In blindsight, however, he raises an intriguing question; one which some
have labelled the problem of "super-blindsight" (Tye, 1996). Imagine a
blindsight subject whose performance is optimised by providing him/her
with feedback after every trial. Over time, suppose that this person is
trained to will him/herself to respond without being cued; to "guess
when to guess." If the subject comes to believe these reports, and if
these are identical to those which normal healthy subjects would issue
when confronted with the same stimuli, the question is: would conscious
perception emerge as a result?
Dennett (1991), who initially proposed this gedanken experiment, is
confident that it will, but Weiskrantz does not share this optimism. On
the basis of subjective reports given by the extensively-tested GY,
Weiskrantz believes that the visual system of a super-blindsight subject
is unlikely to ever graduate to be conscious of the contralesional
visual world. In the many years that GY has been tested, he could only
use vague terms such as "feeling" that the stimulus had been presented,
even when he was confident of the stimulus having appeared. Clearly,
however, the answer is not obvious a priori and it would be a very
interesting experiment to attempt (not to mention one at the height of
heroism), for even if a super-blindsight subject is not fully
"conscious" of the stimulus presented, there is still the perplexing
question of how a such a subject's belief that a stimulus had occurred
would differ from a normally sighted individual's experience of its
presence.
5. The Method of Triangulation
Having emphasised the integral role of the commentary stage in sensory
awareness, the question arises as to its possible underlying neural
substrate. Weiskrantz does not shy from confronting this challenging
problem and proposes a general interdisciplinary approach towards it.
He could hardly go wrong here -- after all, "interdisciplinary" seems to
have been the catch phrase of the '90s (one need only look at the title
of this very journal for evidence of this). Flanagan (1992), for
example, has proposed a similar approach (the "natural method") which
encompasses each and almost every discipline of relevance (including
phenomenology and evolutionary science). Weiskrantz is more specific.
He believes that the three key disciplines of philosophy, neuroscience
and psychology should spearhead investigations into consciousness
("triangulation"), with psychology at the leading apex, and philosophy
and neuroscience at the trailing ends.
Weiskrantz claims that psychology is apt to be at the leading apex of
the triangle because consciousness cannot be understood by examining the
brain qua neurons. This is very much a personal view, of course, and
is far from universally accepted. Whilst being a scientist these days
almost invariably entails adopting a default materialist standpoint,
Weiskrantz is quick to separate himself from the reductionist and
eliminativist philosophies. This sets him apart from the
neuroscientists who might claim, for example, that, "It is premature to
try to describe how the brain really works using just a black-box
approach.... The language of the brain is based on neurons." (Crick,
1994, p. 256, italics his). However, Weiskrantz's point that "If a
Martian looked at the human brain and read every terrestrial textbook of
physiology describing its intrinsic activity, he would never discover
the organ allowed comprehension and generation of speech" (p. 194) is
well taken.
Consistent with the inclusion of philosophy in his method of
triangulation are his frequent references to the work of such
philosophers as Rosenthal, Searle, Dennett and Block. Unlike most other
scientific investigators of consciousness, therefore, Weiskrantz does
not ignore or simply dismiss the opinions of philosophers, which is
highly commendable (compare this, for example, to one such as Crick, who
does not hold philosophers in as high a regard, e.g., "Philosophers have
had such a poor record over the last two thousand years that they would
do better to show a certain modesty rather than the lofty superiority
that they usually display," Crick, 1994, p. 258). In fact, Weiskrantz's
very commentary argument has a philosophical grounding, one based on
David Rosenthal's Higher-Order Thought (HOT) theory (Rosenthal, 1986,
1993).
The HOT theory states inter alia that second-order thoughts about
first-order thoughts must occur for the first-order thoughts to be
rendered conscious. There does appear to be a strong relation between
Rosenthal and Weiskrantz's positions, insofar as Weiskrantz believes
that the commentary stage (analogous to Rosenthal's "second- order
thought") achieved in the processing of an input (or "first-order
thought") is what allows one to be aware of that input. It is a rare
occasion indeed when two theories arising from disparate fields and
separate observations (one philosophical and the other based on
empirical evidence) concur so elegantly, and I believe this does much
for the theories and positions of both Weiskrantz and Rosenthal.
6. "C" for Consciousness and Commentaries
After describing his general strategy on uncovering the neural substrate
of awareness, Weiskrantz attempts to actually apply it. He poses two
questions (p. 142) that he attempts to answer with his method of
triangulation:
1. What brain structure(s) are necessary and/or sufficient for visual
awareness?
2. What removes experience from the discrimination in blindsight and
what allows it to occur in normal perception?
Weiskrantz first briefly reviews a number of solutions that have been
proposed (such as Schacter's DICE model, Crick & Koch's gamma
oscillation hypothesis, quantum theory and Strong AI), and states his
position against the more esoteric of these (e.g., quantum theory). He
then continues by describing the relevance to visual awareness of areas
such as V1, extrastriate cortex, and all other closely-related
structures with a direct connection to V1.
In essence, the upshot of Weiskrantz's extensive and well-synthesised
review is that both V1 and extrastriate cortex are necessary for
consciousness, but are ultimately insufficient. In his view, V1 is a
major distribution point, but not the origin, of visual inputs. It is
postulated that the major role of "higher" visual cortices in normal
perception is to select a particular relevant stimulus attribute (say,
colour) and then feed this information back to V1. In addition, he
postulates the existence of a target area "C", quite separate from V1
and its closely associated areas, which may light up when the subject is
aware of a visual stimulus but not when she or he is unaware
(alternatively, it might light up in both the aware and unaware modes,
but more strongly when the subject is aware of the stimulus). This area
"C" can potentially receive visual information from both V1 and
associated areas, as well as from visual afferent fibres which
completely bypass these areas. Should such a region be shown to exist,
he surmises that it must ultimately feed back into the region of V1 and
its interconnections, as must the back-projections from "higher" visual
cortices, in order to gain access to the commentary stage. Since V1
effectively acts as a funnel in receiving visual information from these
different areas, its destruction prevents this information from being
fed forward to the commentary stage, resulting in a loss of awareness.
Motion, however, may still be able to gain access to the commentary
stage via subcortical pathways. This independence of motion processing
may be due to its evolutionary value (e.g., being aware of sudden
movements by an aggressive predator).
Weiskrantz does speculate on a number of possible candidates for area
"C", such as prefrontal cortex, while paying attention to empirical
findings from animal lesion studies. He expresses a hope that
neuroimaging studies will reveal the location of this area, and proposes
an experiment in which a blindsight subject is tested in both the aware
and unaware modes (with rapid- and slow-moving stimuli respectively)
while being imaged. Just prior to the book's publication, such an
experiment was actually conducted and is described in the Postscript.
The results indicated that there might not be an isolated "C" for visual
awareness; rather, the data demonstrated a shift of activity between the
aware and unaware modes, from dorsolateral to medial prefrontal cortex,
and from cortex to superior colliculus. In the absence of a
Post-Postscript, it appears that we must await further experimentation
in order for these data to be more completely interpreted and understood
in the context of visual consciousness.
7. General Considerations
This book is about commentaries and their relation to awareness. As
such, the title could be misleading, insofar as "consciousness" is taken
to be synonymous with "awareness." Admittedly, "consciousness" is an
irritatingly non-specific, polysemous term, and Weiskrantz does
acknowledge that it is "used in a large variety of different ways and
with all kinds of nuances, but here we restrict it to those
dissociations that have emerged from the neuropsychological syndromes
under review" (p. 164). This is certainly acceptable, although it was
somewhat disappointing that the "hard problem" of consciousness was not
more extensively addressed. In particular, the delicious mystery of
qualia is largely sidestepped: Weiskrantz does concede that, "If
[commentaries are endowing of awareness], I do not necessarily take this
to be an easy solution to the question of how qualia... arise" (p. 204).
But "what it is like" to undergo a particular phenomenal experience is
an issue at the very heart of consciousness -- it is what makes the
mystery of consciousness so intractable -- and I believe such issues
should have been addressed to a greater degree in this book.
In terms of style, this monograph was written with the non-specialist in
mind and, as such, there is some degree of repetition between chapters,
with the language at times being quite informal -- but always clear.
While the author has had to be selective in the material which he
presents, his choice of experiments is appropriate, well-organised and
undoubtedly succeeds in delivering an impression of the field as it
exists today. He goes to some lengths to keep the reader informed of
recent progress in the area, as reflected in his inclusion of a
postscript, written between the submission of the manuscript and its
publication. I imagine that a non-professional would find the book very
readable (perhaps with some assistance from its glossary and appendix),
despite its rather technical flavour. A basic knowledge of neuroanatomy
would be of benefit, especially in appreciating the subtleties of
experimental design which have the potential to be obscured by the
overlying complexities. Largely, however, the significance of the major
experiments he describes speak for themselves (e.g., those based on the
commentary-key paradigm).
To the experienced neuropsychologist, much of the monograph will be
familiar territory. After all, while the syndromes he describes may be
stunningly counterintuitive to the non-expert, they are the stock-and-
trade of neuropsychological research. However, even though the
extensive reviews may not enlighten, Weiskrantz's opinions on the role
of the commentary in awareness are well worth reading.
Consciousness Lost and Found does not claim to be a monograph that
will revolutionise neuropsychology. Instead, it is an excellent review
of the vast amount of research conducted primarily on two most
fascinating neuropsychological phenomena, and is a privileged insight
into the author's position on awareness and his impression of the status
of current research. This monograph makes a strong claim for the
potential that deficits such as blindsight and amnesia possess in
helping to expose the secrets of awareness. Given the strength of such
a potential, we would do well to continue the process of unravelling
this most tantalising of problems, the solution to which appears to lie
right beneath (or at least behind) our noses.
References
Campion, J., Latto, R., & Smith, Y. M. (1983). Is blindsight an effect
of scattered light, spared cortex, and near-threshold vision?
Behavioral and Brain Sciences, 6, 423-448.
Crick, F. (1994). The astonishing hypothesis. New York: Charles
Scribner's Sons.
Dennett, D. C. (1991). Consciousness explained. London: Penguin Press.
Driver, J., & Mattingley, J. B. (1998). Parietal neglect and visual
awareness. Nature Neuroscience, 1, 17-22.
Fendrich, R., Wessinger, C. M., & Gazzaniga, M. S. (1992). Residual
vision in a scotoma; implications of blindsight. Science, 258,
1489-1491.
Flanagan, O. (1992). Consciousness reconsidered. Cambridge, MA: MIT
Press.
Tye, M. (1996). Ten problems of consciousness. Cambridge, MA: MIT
Press.
Rosenthal, D. (1986). Two concepts of consciousness. Philosophical
Studies, 49, 329-359.
Rosenthal, D. (1993). Thinking that one thinks. In M. Davies & G. W.
Humphreys (Eds.), Consciousness: Psychology and philosophical
essays (pp. 198-223). Oxford: Blackwell.
Weiskrantz, L. (1986). Blindsight: A case study and implications.
Oxford: Clarendon Press.
Weiskrantz, L. (1997). Consciousness lost and found. Oxford: Oxford
University Press.
