© 2000, 2002. Shaun Gallagher. Paper presented at the Philosophy
Colloquium, Trinity College, Dublin (November 2000). An earlier version of this
chapter was published as "First Perception: A New Solution to the Molyneux
Problem," in Proceedings of the New York State Philosophical
Association (Syracuse, New York: 1996). This version is forthcoming in Shaun Gallagher. How the
Body Shapes the Mind. Oxford: Oxford University Press.
Neurons and Neonates:
Reflections on the Molyneux Problem
Shaun Gallagher
Department
of Philosophy and Cognitive Science
University
of Central Florida
gallaghr@mail.ucf.edu
A
central tenet of empiricist philosophy is that experience, in the sense of
relatively prolonged exercise or practice of the sense organs, educates perception. On this view, human perception is not
something that happens automatically in the first instance, or in the first
moments of life; it takes time and exposure to the natural environment to
develop properly. This idea has
informed many attempts to solve a long-standing problem first raised by William
Molyneux in a letter to John Locke over three hundred years ago. The question framed by Molyneux has
been called the central question of eighteenth-century epistemology and
psychology (Cassirer, 1951).
Although explicit discussions of the Molyneux problem decreased in
number in the nineteenth century, and fell off even more in the twentieth
century, the issues that it originally raised continue to be of interest, not
only to contemporary philosophers of mind, but to psychologists and cognitive
neuroscientists. Indeed, I want to
suggest that recent findings in developmental psychology and in neurophysiology
-- findings that pertain to the conceptual framework explicated in Part I --
have provided the relevant empirical data for answering once and for all the
Molyneux question.[1]
Molyneux,
in a famous letter to Locke, had entertained the question of what it might be
like for a blind man to gain sight; his question specifically asks what one
sense modality, touch, adds to another, sight -- whether previous tactile
sensations inform first perception in the visual modality.
Suppose a man born blind, and now adult, and
taught by his touch to distinguish between a cube and a sphere of the same
metal, and neighly of the same bigness, so as to tell, when he felt one and the
other, which is the cube, which the sphere. Suppose then the cube and sphere placed on a table, and the
blind man be made to see: quaere, whether by his sight, before he touched
them,
he could now distinguish and tell which is the globe, which the cube?[2]
Molyneux's
question at first glance looks deceptively simple, but it raises many
interrelated and complex issues.
Indeed, commentators have pointed out that Molyneux actually poses
several questions, and there have been some philosophically heated debates
about what Molyneux actually meant.
Some have suggested that he was asking about the perception of distance,
an issue that he clearly did raise in his first letter to Locke in 1688. Others have understood him to be asking
whether spatial perception is confined to touch, or whether vision also allows
for a genuine perception of space.
In this context, questions about the difference between tactile objects
and visual objects have exercised the minds of philosophers from Berkeley to
Merleau-Ponty and Gareth Evans.
For these reasons it is important that I identify at the outset the
philosophical issues I want to investigate.
My intention is to pursue what Evans called the
"background theories of perception" that inform answers to the
Molyneux problem (1985, p. 382).
Specifically, I want to identify the precise principles that best
account for perceptual experience.
To what extent is it correct to say that experience educates perception;
to what extent does perception operate intermodally, so that education of one
sense modality is also education of another? Answers to these questions will have a direct consequence
for how we think of spatial perception, the embodied nature of our access to
the world, and many of the other issues associated with the Molyneux question,
issues that still tend to divide nativists and empiricists.
On
the one hand, a negative answer to the question usually implies a theory of
perception in which access to a meaningful external world is not direct but
mediated in a process that necessitates an acquired capacity to synthesize
sensations belonging to different sense modalities. Such a capacity is acquired through experience, although, as
we will see, the concept of experience is itself open to various
interpretations. On the other
hand, positive answers to the Molyneux question may imply a more immediate
perceptual access to the world on the basis of an innate intermodal system in
which different sense modalities are already in communication.[3] In this case the concept of an innate
system is also open to various interpretations. Thus the Molyneux problem is directed in part at questions
about the respective contributions of nature and experience to perceptual experience.
First
Perception
What
is perception in its very first instance, without the contribution of previous
sense experience, without being informed by established conceptual schemas,
without the influences of habit, custom, language, and so forth? This question asks about first perception,
ontogenetically the earliest instance of perception, which, for obvious
reasons, is an extremely difficult state to describe or characterize. Thought experiments provide one avenue
for the philosophical investigation of this notion, and in the history of
empiricist philosophy there have been several famous ones that explore the
question of what first perception must be like. Condillac (1754), for instance, imagines what it would be
like for a statue, bereft of sensation, to awake in one modality of sensation
at a time. The empiricist hope
expressed in such a project is to return to first perception by stripping away
the multimodal complexity of adult perception -- reducing it to its barest
minimum within one sense modality.
I will try to make clear why the concept of Condillac's statue ignores
the complexities that are implicit in embodied perception, and cannot do
justice to the problems he wanted to address.
In the wake of Locke's Essay, discussions between
Condillac (1746) and Diderot (1749) make it clear that Condillac's statue is a
heuristic place-holder for both the newborn infant and the congenitally blind
subject of Molyneux's question.[4] Although Molyneux himself posed his
question as a thought experiment, philosophers and psychologists soon found
empirical instances of congenitally blind patients who through surgical
operations were given sight. I
will refer to these as Molyneux patients.[5]
In
Locke's view, nature contributes less than experience does to the production of
coherent perception. First
perception, which is no more than unorganized and meaningless sensation, can
hardly be called perception at all.
Perception needs to be educated by experience. For Locke, experience means the frequent and repeated
sensation which leads to the formation of habit or custom, and which shapes
judgment, which in turn, alters and improves perception. (For convenience, I will refer to
this concept of experience as 'Lockean experience'). If, for example, bare sensation presents the adult perceiver
with a pure phenomenal appearance of limited adumbrations, experience in the
form of "habitual custom" fills in to complete the perceptual
content. The empiricist thus
maintains that one must learn to perceive; perception is not delivered as an
innate power; uneducated or first perception cannot deliver a meaningful
world.
Furthermore,
on this empiricist model, first perception is always modality specific. Experience in one sense modality is not
sufficient to educate different sense modalities. As a result, Locke agrees with the negative answer of his
friend Molyneux: No, the newly
sighted man would not be able to discriminate, by vision alone, which was the
globe and which the cube. Locke
explains that the blind man, "at first sight," would not be able with
certainty to recognize by vision the globe or the cube, because of the
difference and the lack of natural connection between touch and vision. Because tactile space is
different from visual space, the two sense modalities do not naturally
communicate, and experience in one does not translate into experience in the
other. How the one relates to the
other can only be learned through experience of both. Locke accordingly suggests that we must consider how much
the perceiving subject "may be beholden to experience, improvement, and
acquired notions" (1694, p. 187).
The
principle that experience enters into and shapes perception is also clearly
expressed in the responses to the Molyneux problem made by Berkeley (1709),
Condillac (1746), Diderot (1749), and in the 20th century, by
neuropsychologists like Donald Hebb (1949) and philosophers like Merleau-Ponty
(1962). Their negative answers
appear to be reconfirmed by the empirical cases that show that Molyneux patients
in fact, at first sight, cannot visually recognize objects that previously were
experienced by touch alone.[6] In unexceptional cases of perception in
the normal adult the senses obviously communicate with each other to produce a
unified spatial experience. In the
initial visual perception for the Molyneux patient, however, "everything
is at first confused and apparently in motion. Discrimination between coloured surfaces and the correct
apprehension of movement do not come until later, when the subject has learned
'what it is to see' . . ." (Merleau-Ponty, 1962, p. 223). It thus appears that the empirical
cases of congenitally blind subjects who gain vision, exceptional though they
are, help to show the importance and necessity of Lockean experience for the
perceptual process. Such empirical
studies furthermore suggest that first perception is modality specific: for the
Molyneux patient what is confused in vision is not confused in touch. This indicates that, in terms of
development, unified perception is not something that blooms forth
automatically in an instantaneous production, but that we have to learn
perception and that our senses have to learn to communicate.
Let
me summarize in three propositions what I understand to be the important principles
of the empiricist reasoning on this issue.
(1) The development
of coherent perception depends on Lockean experience. At first, perception is confused; one
needs to learn to perceive. In Molyneux
patients the confusion of first perception means that colors are not contained
by definite shapes, and everything seems to be in motion.
(2) Sense modalities
do not naturally communicate. Sense modalities are
heterogeneous and have their own unique spatial and structural features.[7] They do not function intermodally; they learn to
communicate with one another (usually on an amodal level) only as the result of
Lockean experience. Part of the
learning of perception involves learning to integrate these different systems
on some perceptual or conceptual level.
(3) Lockean
experience in one sense modality does not educate other sense modalities. The Molyneux problem receives a negative
answer because experience in the modality of touch is not sufficient for
learning to perceive in the modality of vision.
These
three principles and the concept of Lockean experience provide a framework for
the traditional empiricist background theory of perception. In the following sections I will
argue that principles (2) and (3) have been overturned by recent discoveries in
developmental psychology, and that this requires that we rethink the first
principle, along with the concepts of experience and first perception that form
its basis.
The Developmental
Context
Historically, answers to the Molyneux
question have been related to the issue of first perception in the
newborn. Thus, for example,
Diderot contends that there is no doubt "that vision must be very
imperfect in an infant that opens its eyes for the first time, or in a blind
person just after his operation" (1749, p. 52).[8] Unsurprisingly, the same three
empiricist principles are often reflected in discussions of the developmental
issues. Recent discoveries in
developmental psychology, however, challenge the traditional empiricist theory
of perception in a way that can clarify the concept of experience at stake in
its formulation, and generate new principles. To see this we need to transpose the principles expressed in
(1)-(3) from the context of adult perception in tactile and visual modalities
to the context of developmental studies that involve tactile, visual, and
proprioceptive modalities.
In
the developmental context the question might be framed in this way: Where do we
find in the life of an infant, the emergence of coherent perception? To put the question in this way,
however, is already to grant too much to the traditional empiricist model. The notion of an emergence of
perception already implies that experience for the newborn infant is an
inchoate, incoherent and uncoordinated mass of sensations out of which coherent
perception develops. This is nicely attested to in William James' famous phrase
describing the infant's experience as
"a blooming, buzzing confusion" (1890).[9] This idea, which reflects principle
(1), is clearly expressed in the traditional view that exteroceptive perception
appears relatively late in the developing infant.
As
we know from our previous considerations (Chapter 3), developmental theorists
until recently believed that experience begins by being interoceptive, and that
the newborn is without external perceptual ability. On this view, James' "blooming, buzzing confusion"
persists until between the third and sixth month of postnatal life when the
interoceptive and exteroceptive domains begin to collaborate -- a collaboration
that is not found at the beginning of life. Recall that one reason cited for this lack of organized
exteroceptive perception involves the absence of a developed body schema. Thus psychologists like Wallon and
philosophers like Merleau-Ponty regard a developed postural schema as a
necessary requirement for coherent perception. For someone like Merleau-Ponty, this idea is completely
consistent with the notion that perception is embodied. Perception is disorganized at this
stage precisely because it is infected by the disorganization of the body.
On
the empiricist model, the neonate's experience is confused; the infant has not
yet learned to perceive. Moreover,
reflecting the second principle (sense
modalities do not naturally communicate), the reason for this confusion is that the
infant has not yet learned to integrate the proprioceptively defined spatiality
of its own body with the visual information it receives from the
environment. Proprioceptive space
involves an intra-bodily spatial framework; the organization of visual space
involves an egocentric spatial framework.
The development and integration of these two frameworks occur only with
prolonged motor and sense experience.
Intermodal communication or collaboration, important for the
establishment of a mature perception, has not yet been established between
proprioception and vision, or between proprioception and touch. Proprioceptive organization, as
Merleau-Ponty indicates, is established first, prior to, and as a necessary
condition for exteroceptive perception.
Reflecting the third empiricist principle (sense modalities do not
naturally communicate), however, proprioceptive organization alone is not a
sufficient condition for organized visual or tactile perception -- that is,
proprioceptive experience does not inform visual or tactile sense modalities.[10]
Still,
the neonate seems not to be in the same special circumstance of the Molyneux
patient. For example, the Molyneux
patient already has extended experience in the tactile domain and in other
non-visual sense modalities.
Nonetheless, as we have seen, the question of neonate imitation involves
the issue of intermodal translation and does so in a way similar to the case of
the Molyneux patient.[11] The recent developmental research
on neonate imitation clearly challenges the empiricist principles that underpin
the traditional view. Let us
consider again an unfortunate infant who is taken directly from the birth
process and, having undergone ten minutes of the usual medical procedures and
tests, is at this early age made to participate in a psychological
experiment. A person visually
presents to the newborn a facial gesture -- for example, a tongue protrusion or
a mouth opening. The question is:
Can the infant imitate the facial gesture? To do so the infant is required to translate its experience
from one sense modality, vision (indeed, something very close to its first
vision), into another, proprioception.
We
would expect those who answer Molyneux's question in the negative, for reasons
outlined by the three empiricist principles, to respond in a similar way to the
question about the possibility of neonate imitation. And we are not disappointed. Indeed, the traditional answers to these two questions, as
well as to the question of aplasic phantoms, are quite consistent. As we have seen before (in Chapter 3),
Piaget, consistent with a negative answer to the Molyneux question, maintains
that sense modalities initially represent heterogeneous spaces which the infant
must gradually learn to synthesize.
This would mean that imitation in the neonate is impossible since it
could occur only on the basis of a highly organized proprioceptive
experience. The infant would require
proprioceptive access to its own face in order to be able to find the proper
anatomical parts (for example, the mouth, the tongue, the lips) to use in the
imitation process, and it would have to know how to move them. Furthermore, the infant would require
an organized visual experience to be able to discern the gesture of the other's
face, and seemingly would have to make the connection between these two sense
modalities on some intellectual level.
Thus, according to Piaget (and most other classical theorists of
development), imitation is not possible prior to 8 to 12 months of age.[12] Accordingly, the newborn is caught up
in a complete confusion: it does not yet have an organized proprioceptive
experience of its own body (for example, it would not know where its mouth is
or what to do with its tongue); it cannot visually perceive the face of the
other person; and thus it certainly cannot imitate visually presented
gestures.
As
we have seen, however, newborn infants seemingly take issue with this traditional
answer. Previously cited studies
by Meltzoff
and Moore, and others, show that neonates less than an hour old can in fact
imitate facial gestures and even hand movements, that young infants imitate
facial gestures after a delay of as much as 24 hours, and that infants improve
their gestural performance over time (Meltzoff and Moore, 1977; 1994). For reasons outlined earlier, it is
also clear that neonatal imitation is not a matter of reflex behavior.
Both
neonatal imitation and the issues surrounding the Molyneux question involve the
concept of intermodal perception.
To move us closer to the terms of the Molyneux problem, consider the
following study.
Meltzoff and Borton (1979) evaluated
tactile-visual cross-modal perception in 32 infants (average age of 29
days). Two pacifiers with quite
different shapes were used -- one is shaped like a sphere, the other a sphere
with nubs (figure 7.1). Pacifiers
were hidden from the infant's sight prior to slipping them into the infant's
mouth. Half of the infants were
given the sphere, and the other half received the nubbed pacifier. They were allowed to suck on them for
90 seconds, during which time most infants engaged in tactile exploration with
the tongue, "furrowing their brows, as if the tactual exploration of the
novel object was of some interest" (Meltzoff, 1993, p, 223). The infants were then shown the two
objects side by side. In the
context
of such experiments object recognition is often measured by how long the infant
stares at a particular object within a fixed period. The results of this experiment showed that the infants did
look significantly longer at the shape they had felt. Of the 32 infants tested,
24 visually fixated for a mean percentage of 71.8% of the time on the shape
matching the one they had explored in the tactile modality -- figures greatly
above chance.
Meltzoff concluded that very young infants
"register the same information about the shape of the object even if it is
picked up through two different modalities, touch and vision" (1993, p.
224). Together with the intermodal
communication between vision and proprioception found in the case of neonate
imitation, and with the results of other experiments that demonstrate intermodal
communication between vision and hearing (Kuhl and Meltzoff, 1992; 1984), this
conclusion is consistent with the general thesis that first perception is
already shaped by an innate capacity, a built-in mechanism for intermodal
communication among sense modalities.
These experiments also make it clear that coherent perception that is not educated by experience
(at least postnatal Lockean experience) occurs from the very beginning of our
postnatal life.
What implications do
these studies have for Molyneux's question, or more generally for the background
theory of perception? We clearly need to revise the three
traditional empiricist principles in the light of the developmental context.
(1') Relatively organized perception is
possible from birth. It does not require
Lockean experience as a necessary condition.
First perception is not confused sensation. The body is already organized by an innate body schema, and
this organization is carried through to perception. Without Lockean experience, in a first instance, perception
is not
a blooming, buzzing confusion.
Having said that, there is still an important sense in which experience
in and beyond first perception does educate later perception in a
self-organizing fashion. Moreover,
there is also a sense in which experience, expanded to include prenatal
experience, and understood to have a physical effect on the experiencing body,
informs first perception. We
return to these points below.
(2') Sense modalities do communicate
naturally. Perception is intermodal
from the very start. First
perception already operates in an intermodal fashion. The perceiving subject does not have to
learn to integrate different systems, because they are already innately
integrated. Although separate
sense modalities can be said to have their own unique spatial and structural
features, they communicate intermodally in an egocentric spatial framework that
is fully integrated from the beginning with a proprioceptive intra-bodily
spatial framework (Chapter 6).
(3') Experience in one sense modality
does educate other sense modalities. In the case of neonate imitation vision directly educates
proprioception. In the experiment
conducted by Meltzoff and Borton (1979) the tactile modality directly informed
vision. Perception in one sense
modality is, in certain respects, sufficient experience for the recognition of
objects in another modality.
Answering
the Molyneux Question
The
reformulated principles seem clearly to contradict the traditional empiricist
theory of perception. The latter
is associated with and stands as the background theory of perception for the
negative answer to the Molyneux problem.
The empiricist theory extends further, of course. Most theorists who discuss the Molyneux
problem generalize their conclusions concerning the role of Lockean experience
in the education of perception, and the heterogeneity among sense modalities,
to apply across all instances of human perception, including first perception
in infancy. In other words,
Molyneux patients are not exceptions or counterexamples to the original
principles; they count as evidence in support of them. In contrast, neonates who imitate
facial gestures would count as counterexamples. The recent developmental
studies show clearly that the empiricist propositions (1)-(3) fail to describe
first perception in the case of the neonate. This suggests, at the very least, that theorists from Locke
to Hebb have been too quick to generalize.
On
the other hand, if we generalize the principles expressed in (1')-(3'), we are
led to affirm an innate connection between touch and vision. This would seem to imply, as Gareth
Evans suggests, a positive rather than a negative answer to the Molyneux
problem.[13] Could this be right?
Since
the time that Cheselden published, in 1728, his empirical account of a cataract
operation in a patient blind from birth or from infancy, and his observations
concerning the patient's visual perception, theorists as early as Berkeley
(1733) have cited cases of recovery from congenital blindness as support for
the negative answer to the Molyneux question.[14] After such operations patients are
unable to distinguish between objects.
Some patients, with the assistance or tutoring of touch, go on to learn
to visually distinguish objects; others give up the difficult education of the
visual sense, ignore their imperfect vision, and return to the life of a blind
person. In the face of this
empirical evidence, is it possible to maintain a positive answer?
The
problem is that the empirical studies of Molyneux patients are plagued by
ambiguities. These are well
rehearsed in the literature from La Mettrie (1745) to Oliver Sacks (1995) and
involve uncertainties about the timing of onset of blindness, the degree of
blindness in cases of cataract, and the confused experience of the patient
after the operation. The
uncertainties about the empirical data create uncertainties with regard to the
negative answer to the Molyneux question.
Gareth Evans, in an attempt to overcome such uncertainties, suggests an
experiment that would clarify the issue.
Evans suggests that by means of direct electrical stimulation of the
visual cortex, a pattern of experienced light flashes (phosphenes) in the shape
of a square or circle could be caused in a patient with congenital
blindness. If this were
possible, one could settle the matter once and for all. Evans argues that under such
circumstances the patient would indeed be able to tell the difference between a
circle and a square. His reasons
for this positive answer to the Molyneux question involve his concept of
embodied action in an egocentrically organized behavioral environment; that is,
his reasons are somewhat different but not necessarily inconsistent with the
evidence from neonatal imitation cited above.[15]
Now
despite the fact that I think the evidence from neonate imitation, and Evans'
conception of embodied action are in fact consistent with the reformulated
principles of perception (1') -(3'), I want to disagree with Evans' conclusion
and argue in favor of a negative answer to the Molyneux question. This negative answer depends, however, on a
particular interpretation of the question. If we take the question to mean: "Will the Molyneux
patient visually be able to distinguish the objects as such," then the
answer, I argue, is "no."
On a different interpretation of the question, which I will consider in
the next section, the answer may be different. Here, on the stated interpretation, I want to argue that
Locke, Berkeley, Condillac, Diderot, Merleau-Ponty, Hebb, and others are right
to say: No,
the newly sighted person would not be able to discriminate, by vision alone,
which was the globe and which the cube.
I think they base their answers on the wrong reasons, however, and more
generally they are led to the wrong principles of perception. To develop and maintain the correct
version of the negative answer, I need to provide some reasoned explanation of
the experimental evidence. After
all, what I take to be the most secure empirical data, the evidence from
neonate imitation and other experiments on intermodal perception, point to a
positive answer for the Molyneux question. The evidence from cataract operations, which supposedly
supports the negative answer, remains ambiguous. Evans' suggestion about cortical stimulation has not been
tried, but I think he is wrong about what it would show if it were actually
tried. That is, I think that even
under these conditions the subject would not be able to identify or tell the
difference between a square and a circle.
Locke
and the empiricists correctly give the negative answer, but for the wrong
reasons.[16] My response involves the formulation of
a completely different explanation for the negative answer to the Molyneux
problem. The solution is to be
found in the neurophysiology of vision.
Once understood it is easy to see that a negative answer to the Molyneux
question is not inconsistent with the reformulated propositions (1')-(3'), or
with the experience of the neonate, but is an explainable exception to the
principles that normally govern intermodal perception. In
effect, the reasons why the negative answer to Molyneux's question can be
consistent with the reformulated propositions rather than explained by the
empiricist propositions (1)-(3) are to be found in the way the body works, and
specifically on the sub-personal, neurophysiological level of explanation.[17]
As
we noted in discussing (1'), even if first perception in the newborn is already
somewhat organized, continued experience (in the Lockean sense) does educate
later perception in a self-organizing fashion. The difference between the first empiricist principle (1),
and the reformulated one (1'), however, depends on how we conceive of
experience more generally, its effects, and its timing. Although it would be
difficult to deny that postnatal Lockean experience educates and fine-tunes
perception, we can deny that the development over time of postnatal visual
experience is a necessary condition for relatively clear visual perception. The
neonate sees well enough to imitate facial gestures immediately upon birth
(given some moments for its eyes to adjust to light).
It
may be helpful to clarify just what the visual capacities of the newborn human
are. Atkinson and Braddick (1989)
and Slater (1989) indicate that newborn vision involves the following
characteristics:
sufficient acuity to resolve details at
close range;
contrast sensitivity great enough for the
perception of spatial patterns;
cortical activity and control (and not just
subcortical functionality);
overt control of attention and selection of
objects by means of eye movements;
preference for some objects over others
(e.g., moving objects rather than stationary objects; three-dimensional objects
rather than two-dimensional ones; high contrast rather than low-contrast
stimuli; patterns with curved rather than straight lines);
active searching for visual stimulation;
ability to habituate to one stimulus and
show preference for novel stimuli;
activation of visual memory -- the
possibility of habituation in neonates demonstrates a very short-term visual
memory (Slater 1989), but more durable memories have been demonstrated in
2-day-old babies who can retain a memory of the mother's face over a minimum
of 5 minutes (Bushnell and Sai,
1987).
sensitivity to motion parallax -- when the
infant moves its head, nearer parts of a viewed three-dimensional object appear
to move more than farther parts.
Neonates
are capable of discriminating between different geometrical shapes, such as
triangles, squares, and circles (Landau, Gleitman, and Spelke, 1981; Slater,
1989; Slater and Morison, 1985a; Slater, Morison, and Rose, 1983). Their vision is also characterized by
shape constancy, that is, across changes in orientation or slant neonates are
capable of recognizing the real shape of an object (Slater and Morison,
1985b). Neonate vision also shows
feature constancy, that is, the ability to recognize invariant features of an
object across certain varying features, such as moving versus stable objects
(Slater, 1989).
Qualifications
need to be made on all of these aspects.
Visual acuity in the newborn is approximately 30 times poorer than in
adults, but is still sufficient for some degree of pattern recognition (Slater,
1989). Further, the neonate's
vision is limited in terms of distance -- visual attention is not usually
directed to objects more than three feet away unless they are moving (Slater,
1989). Neonate vision is not fully
binocular (Atkinson and Braddick, 1989), and cortical control is far from fully
developed. Nonetheless, it is
clear that neonate vision is already educated to some degree.
To complete the explanation, however, one needs
to look on the neuronal level. To
ascertain whether there is an education process required for first perception
one needs to look at neuronal development, and it is here that we can find the
definitive answer to the Molyneux question. The established view on neuronal development from birth
onward begins with a statement about the initial condition: the patterns of
interconnection of human cortical cells are at first sparse (figure 7.2a)
compared to patterns in the adult (figure 7.2b; see Atkinson and Braddick,
1989; Shatz, 1992). If this is the
case, behavioral evidence tells us that there is nonetheless, in the visual
cortex, a sufficient neuronal structure in place that allows the newborn to see
as specified above. More recent
neurological evidence strongly supports this view. It suggests that the visual cortex of the newborn may be
more developed than first thought.
Indeed it may be much closer to the adult brain in regard to the
structure of the ocular dominance columns, groups of nerve cells in the visual
cortex that respond to input from one eye or the other (Crowley and Katz,
2000). Regardless of the initial condition, however, it is important
to consider what happens as the result of subsequent experience, or the lack of
subsequent experience.
