# Full Text: Predictive Processing Interpretation of the Mirror Test and Implications of a Reflection Prediction for Human Cognition

> Extracted from `2022_MirrorTest.pdf`

---

## Page 1

Predictive Processing Interpretation of the
Mirror Test and Implications of a
Reflection Prediction for Human Cognition
Sean O’Connor1, Daniel Ari Friedman2
1) Spoconno@gmail.com
https://orcid.org/0000-0002-0858-5635
Independent Researcher
2) DanielAriFriedman@gmail.com
https://orcid.org/0000-0001-6232-9096
Active Inference Institute
University of California Davis, Department of Entomology & Nematology
Version 1, uploaded November 29, 2022
Abstract
The "mirror test" has been used as a behavioral measure of mirror self-recognition for a variety
of species. In this article we apply a predictive processing interpretation to the results of the
mirror test in order to offer a novel perspective with which to understand mirror self-recognition
and self-directed behavior. Furthermore, we hypothesize that a “reflection prediction”, upon
which our predictive processing interpretation of the mirror test is built, may also offer a novel
perspective to understand how humans locate themselves relative to a mirror, imitate others,
and are self-aware from a social perspective. As we show that a reflection prediction may help
to explain how these traits may emerge in human cognition, we also point out that atypical
reflection predictions or atypical use of a reflection prediction may help to explain instances
where these traits are atypical in certain individuals.
1

## Page 2

Introduction
In “The free-energy self: A predictive coding account of self-recognition” Matthew Apps and
Manos Tsakiris explain how mirror self-recognition may be understood as the result of an
individual successfully learning to predict their own reflection in the mirror [1]. As they write:
“the viewing of an agent’s own actions in a mirror (including arm movements, facial
expressions etc.) will lead to optimised high level empirical priors about one’s body,
which will in turn modulate expectations in the visual system about the expected visual
consequences of one’s own actions.”
These optimized high level empirical priors are an individual’s expectations about their own
visual reflection and therefore what we refer to here as a “reflection prediction”. In this paper we
highlight that a successful reflection prediction explains mirror self-recognition in humans, as
pointed out by Apps and Tsakairis, and also offers a novel perspective on how to interpret the
results of the “mirror test” in other species.
Furthermore, we point out that the concept of a reflection prediction is crucial for the scientific
and medical community to consider because if a reflection prediction allows agency over a
reflection then reflection prediction errors may cause atypical agency over a reflection, and
reflection prediction error correction may help to reverse atypical agency over a reflection.
We conclude by pointing out that a reflection prediction should also be considered as a
mechanism the human brain may depend upon for the emergence and the successful
functioning of cognitive traits other than mirror self-recognition such as the ability to locate
oneself relative to a mirror, the ability to imitate others, and the ability to be self aware from a
social perspective. As with mirror self-recognition, we point out that if a successful reflection
prediction contributes to the proper functioning of these traits, then reflection prediction errors or
errors with the use of a reflection prediction may cause atypical functioning of these traits as
seen in certain individuals.
Predictive Processing Perspective of the Mirror Test
In 1970 Gordon Gallup Jr. devised a test to determine whether animals recognize themselves in
the mirror and therefore whether they were capable of mirror self-recognition [2]. To accomplish
this he anesthetized chimpanzees and monkeys, applied an odorless red dye to the animals eye
ridge and opposite ear, and after a period of time observed their behavior both with and without
a mirror. When the mirror was added to the experiment, chimpanzees with prior mirror
experience exhibited self-directed behavior towards the mark and spent more time staring into
the mirror, but monkeys with prior mirror experience did not exhibit self-directed behavior
towards the mark or spent more time staring into the mirror; therefore, Gallup determined that
chimpanzees were capable of mirror self-recognition and monkeys were not capable of mirror
self-recognition. This test has come to be known most popularly as the “mirror test” (also, “mark
2

## Page 3

test”) and it has been repeated many times across a range of species in the years since
Gallup’s original paper.
In this paper we will show that applying the predictive processing framework to the results of the
mirror test offers a novel perspective to interpret the results of the mirror test. Predictive
processing models suggest that individuals build internal predictive models of the external world
and seek to minimize the extent to which their predictions differ from reality. In predictive
processing models if an individual’s prediction is different from their experience they can engage
in prediction error minimization by either updating their prediction, a process called perceptual
inference, or through acting on their environment, a process called active inference [3]. If an
individual’s prediction is equal to their experience, the individual’s attention to the sensory
experience is dampened in a process called sensory attenuation and, for self-generated
movement, agency is achieved over the movement.
Applying this predictive processing framework to the results of the mirror test offers a
new perspective to understand mirror self-recognition1 and self-directed behavior, where
mirror self-recognition is due to sensory attenuation when an individual successfully
predicts their reflection and self-directed behavior is active inference to correct reflection
prediction errors. In the next two sections we will discuss the implications of each of these
claims.
Mirror Self-Recognition is a Product of Sensory Attenuation
As mentioned in the introduction of this paper, Apps and Tsakiris use a predictive coding model
to describe how mirror self-recognition in humans may be considered a product of sensory
attenuation when an individual successfully predicts their reflection [1]. In this section, we point
out that this concept is applicable not only for humans but also animal species, offering a new
way to interpret the results of the mirror test. When responding to the mirror test species fall into
one of three categories - those which treat their reflection as a conspecific indefinitely, those
which treat their reflection as a conspecific before ignoring their reflection entirely, those which
treat their reflection as a conspecific before displaying self-directed behavior. Traditionally only
those species which display self-directed behavior are considered capable of mirror
self-recognition, however, this predictive processing analysis of the mirror test offers a
new perspective where both species which learn to ignore their reflection in the mirror
and species which display self-directed behavior in the mirror are capable of mirror
self-recognition.
That individuals which ignore their reflection are capable of mirror self-recognition may seem
strange, however, it is supported by both the predictive processing model and the ecological
evidence. It is supported by the predictive processing model because sensory attenuation of
self-generated behavior leads to a dampening of attention towards external sensory stimuli and
has been claimed to lead to a sense of agency over that behavior [4,5], and thus help generate
self-other distinction [6,7]. It is supported by ecological evidence because studies have shown
1 And self-recognition in naturally reflective surfaces such as standing pools of water or mineral deposits
in caves or on the ocean floor
3

## Page 4

that even among species which are capable of self-directed behavior, individuals may not
display self-directed behavior during the mirror test. For example, one study on Asian elephants
found that only one of three elephants passed the mirror test [8]. The authors therefore suggest
the test may be flawed because the elephants simply may not care about the mark - as they
write, “A small paint mark may be trivial to them.” Another example of this is that studies have
found even for Chimpanzees, which were used in Gallup’s original experiment, fewer than half
of the individuals studied pass the mirror test [9] [10]. These studies highlight an inherent flaw in
the scientific interpretation of the mirror test; the mirror test measures self-directed behavior, not
mirror self-recognition, and many individuals or even species who fail to exhibit self-directed
behavior may nevertheless be capable of mirror self-recognition, a suggestion which is
consistent with the predictive processing analysis of the mirror test presented here.
Self-Directed Behavior is an Instance of Active Inference
If self-directed behavior is not an accurate measurement of mirror self-recognition, what exactly
is the self-directed behavior exhibited during the mirror test an accurate measurement of?
One possibility is what we call “mirror proprioception”, the understanding of the location of self in
space relative to a mirror. As Gallup writes: “self-directed and mark-directed behaviors would
seem to require the ability to project, as it were, proprioceptive information and kinesthetic
feedback onto the reflected visual image so as to coordinate the appropriate visually guided
movements via the mirror” [2]. We agree, however, it is important to note that self-directed
behavior is also not an accurate measurement of mirror proprioception in a species for the same
reason as mentioned above; the subjects being studied may simply not care about the mark.
Indeed, a number of species have passed “mirror use” tests that demonstrate their ability to use
a mirror to locate objects in space but have nevertheless failed to demonstrate self-directed
behavior during a mirror test [11], supporting the idea species which do not display self-directed
behavior may nevertheless have a visuospatial understanding of mirrors.
Therefore, self-directed behavior is an accurate measurement of neither mirror self-recognition
nor mirror proprioception. In this model, self-directed behavior is conceptualized as a specific
instance of active inference. Earlier we wrote that in predictive processing models, when
prediction of self-generated sensory input is different from the experience of self-generated
sensory input, prediction error correction has been proposed to occur either in the form of
perceptual inference (updating predictions) or active inference (acting on the environment).
Applying this predictive processing framework to the results of the mirror test offers a
new perspective to understand self-directed behavior as active inference to correct
reflection prediction errors. In the case of the mirror test, the mark causes a reflection
prediction error because the individual's experience of their reflection now differs from their
prediction of their reflection, and self-directed behavior is active inference to correct the
reflection prediction error caused by the mark. This view of self-directed behavior as active
inference also provides an explanation of why some individuals or even species seem to “not
care” about the mark during a mirror test; those individuals or species may be ignoring that
reflection prediction error because they are using perceptual inference to update their reflection
4

## Page 5

prediction in order to correct the reflection prediction error caused by the mark during the mirror
test. This prediction error correction from perceptual inference would theoretically re-allow
sensory attenuation because the updated prediction would allow reflection prediction to equal
reflection experience and make self-directed behavior unnecessary, and atypical perceptual
inference may thus explain certain instances of increased self-directed behavior.
While interpreting self-directed behavior during the mirror test as active inference is novel, the
idea is consistent with the concept of active inference in predictive processing models, if we
consider that the individual emerges to themselves as a part of the environment through a
mirror. For example, Constant et al. argue that the self can extend to objects in the environment
via niche construction, a process in active inference where an individual modifies the
environment based on their internal model. As they write:
“An agent not only learns a model of its environment, but also changes the environment
to fit its models. As a result, it tends to construct an environment that mirrors its
predictions, and, in a sense, makes the world its own.” [12]
Thus, if an individual has emerged to themselves as a part of the environment through a mirror,
the individual may exhibit self directed behavior as a way to “construct an environment that
mirrors [their] predictions”.
Therefore in this predictive processing interpretation of the mirror test self-directed behavior in a
species is a measurement of whether the species faced a strong enough selection pressure to
drive the natural selection of active inference in front of a reflective surface. This illuminates why
some species capable of mirror self-recognition display self-directed behavior while other
species capable of mirror self-recognition do not display self-directed behavior. Examples of
species that may have faced a strong enough selection pressure to drive the natural selection of
active inference in front of a reflective surface include those within which individuals are
threatened directly by diseases carried by blood or parasites which may be removed with
self-directed behavior, and/or species in which individuals are threatened indirectly by blood or
parasites because physical appearance has a reproductive fitness consequence. For example,
one of the species which has passed the mirror test by exhibiting self-directed behavior is the
cleaner wrasse [13]. According to this theory, the reason the cleaner wrasse exhibits
self-directed behavior is that they are threatened directly by blood or parasites and/or
threatened indirectly by blood or parasites causing a reproductive fitness consequence.
Considering the cleaner wrasse survives by eating parasites off of bigger fish in a mutualistic
relationship, it is indeed likely that they are threatened both directly and indirectly by parasites
which can be removed in front of a reflective surface; directly because they spend their lives
eating them off bigger fish, and indirectly because the big fish may be less likely to enter a
mutualistic relationship with cleaner wrasse that is already infected by parasites.
Reflection Prediction Errors and Error Correction
If a successful reflection prediction is responsible for agency over a reflection in the mirror as
first suggested by Apps and Tsakairis, it is important for the scientific and medical community to
5

## Page 6

consider that reflection prediction errors may cause atypical mirror self-recognition as seen in a
variety of individuals, and that reflection prediction error correction may help to treat atypical
mirror self-recognition. Moreover, certain behaviors may be considered attempts at reflection
prediction error correction analogous to the self-directed behavior observed during the mirror
test, as discussed earlier.
To elaborate on this we will point out that although a reflection prediction is theoretical, one
method to investigate this topic is to consider what errors a reflection prediction could have,
what the effects of this error would be, and how the brain could correct this error, then compare
these predictions with symptoms and behaviors actually observed in various individuals.
For example, one hypothetical reflection prediction error that we can consider is that an
individual predicts their reflection is missing a limb that the individual is not actually missing.
This circumstance would result in a reflection prediction error because the individual's reflection
prediction would be different from their actual reflection. The individual may use perceptual
inference to update their reflection prediction and correct the reflection prediction error, however,
if perceptual inference fails, active inference would be necessary to correct the reflection
prediction error. Active inference could take two forms - either amputate the limb so that
reflection prediction is equal to experience, or use a virtual reality mirror that allows the
individual to see their reflection without the limb.
The hypothetical reflection prediction error just described closely resembles body integrity
identity disorder (BIID) - a condition characterized by individuals desiring amputation. The model
presented in this paper offers an explanation for this experience if we consider that individuals
may suffer from a reflection prediction error where they no longer predict the reflection of a
certain limb. In this circumstance the desire of an individual to amputate may be considered
active inference for reflection prediction error correction, analogous to the self-directed behavior
mentioned earlier in this paper, because amputation will correct the reflection prediction error as
the individual will now no longer see the reflection of the limb; therefore, reflection prediction
and reflection experience will align and sensory attenuation may be achieved. Another way of
achieving this alignment would be to use a virtual reality mirror that allows the individual to see
their reflection without the limb that they no longer expect to see in the mirror. This is exactly
what occurred in one study [14] that used a virtual reality mirror to “virtually amputate” specific
limbs of two men with BIID; both men reacted to the virtual amputation positively, for example,
one said it was “emotionally overwhelming to finally see himself like this”, while the other
reported a reduction in his BIID symptoms in the days following the virtual amputation. Thus
BIID closely resembles a disorder caused by a hypothetical reflection prediction error where an
individual does not predict the reflection of a specific limb.
We therefore urge the scientific and medical community to consider other ways that reflection
prediction errors could explain atypical mirror self-recognition and consider how reflection
prediction error correction could help to treat these reflection prediction errors. If the brain is
willing to amputate part of the body in order to correct a reflection prediction error, as may be
the case with BIID, what other behaviors could be viewed as attempts at active inference to
6

## Page 7

correct reflection predictions errors, and what could be done in order to attempt to treat these
reflection prediction errors? A few suggestions are listed below in Table 1.
Table 1
Hypothetical
reflection prediction
error
Possible symptoms
due to reflection
prediction error
Possibly
related
disorder
Possible reflection
prediction error correction to
reduce symptoms
Individual predicts
reflection is different
than reality
Claiming person in
the mirror is someone
else, compulsive
plastic surgery
Mirrored-self
misidentification
VR mirror or plastic surgery
that allows reflection to more
closely resemble reflection
prediction
Individual predicts
reflection weighs less
than reality
Compulsively losing
weight
Anorexia,
bulimia
VR mirror that shows reflection
that weighs less than in reality
Individual predicts
reflection is bigger
than reality
Compulsively gaining
weight
Compulsive
overeating
VR mirror that shows reflection
that weighs more than in reality
Individual predicts
reflection looks dead
Emulating the dead,
suicidal behavior
Cotards
disorder,
depression
VR mirror that shows reflection
looks dead
Individual predicts
reflection is in pain
Pain
Psychogenic
pain
VR mirror that shows reflection
in pain
Individual predicts
reflection looks
paralyzed
Paralysis
Psychogenic
paralysis
VR mirror that shows reflection
paralyzed
Table 1. Hypothetical reflection prediction errors and their possible associated
symptoms, disorders, and methods for error correction.
Implications of Reflection Prediction for Human Cognition
This paper has focused on the implications of a predictive processing analysis on the mirror test
and the implications of a reflection prediction for mirror self-recognition and atypical mirror
self-recognition. Here we shift focus to consider that the brain may use a reflection prediction for
purposes other than mirror self-recognition; specifically but not exclusively, for locating oneself
in space relative to a mirror, for imitation, and for self-awareness. As with mirror self-recognition,
we also consider that if each of these traits depends upon a successful reflection prediction,
reflection prediction errors or errors with the use of a reflection prediction may help to
understand instances where these traits are atypical in certain individuals.
7

## Page 8

Reflection Predictions and Mirror Proprioception
In addition to mirror self-recognition, a reflection prediction may provide an individual with what
we refer to as mirror proprioception, the ability of an individual to locate themselves in space
relative to a mirror. A visuospatial relationship between self and reflection as a product of a
successful reflection prediction makes sense because in order for an individual to be able to
predict their reflection they have to understand how their reflection reacts to their motor
movements, for example, that as they move their hand further from themselves and closer to a
mirror, their reflection will move its hand further from itself and grow in size as it appears to
move towards the individual from inside the mirror. Mirror proprioception is necessary for
everyday tasks like shaving, putting on makeup, and driving. If the concept of mirror
proprioception does not seem intuitive, consider that species such as bats and dolphins use
echoes to locate themselves in space and that echos are a reflection of sound just as mirror
reflections are a reflection of light. Thus the suggestion that individuals use mirror reflections to
locate themselves in space is situationally analogous to bats and dolphins using echoes to
locate themselves in space.
One experiment that may demonstrate mirror proprioception is the “mirror illusion” [15]. During
the mirror illusion a mirror is placed on an individual’s midsagittal axis and the individual
synchronically moves their left and right hands, but cannot see one hand because of the mirror.
For example, if the mirror is placed such that it reflects the left side of the individual’s body, the
individual will not be able to see their right hand; instead, they see the left hand and the
reflection of their left hand. According to this study, as long as the actual hand and mirror illusion
hand are within 10 cm of one another, a phenomenon known as proprioceptive drift occurs
where the visual information from the illusory hand in the mirror overrides normal proprioception
and the individual’s perception of their body-part position changes.
For an individual to feel ownership and agency over the hand in the mirror may be expected
because our model claims an individual feels agency over their reflection when reflection
prediction is the same as reflection experience. In the mirror illusion, the reflection of an
individual’s left hand with a mirror on their midsagittal axis gives the individual the illusion that
their left hand is their right hand in the mirror. Moreover, in this experiment the individual was
synchronously moving their right hand with their left hand. Therefore, in our model the individual
would be predicting the movement of their right hand’s reflection and actually seeing a reflection
that looks what their right hand’s reflection would actually look like (the reflection of their left
hand) which would allow reflection prediction to equal reflection experience, granting the
individual ownership over the reflection where they feel like it is the reflection of their right hand
and agency over the reflection because they now predict the reflection will respond to the motor
commands of their right hand. As one subject in this experiment said: ”Whenever I moved my
right hand, I expected the hand in the mirror to move in the same way.”
Additionally, proprioceptive drift is expected because we claim a successful reflection prediction
may allow mirror proprioception, the ability of an individual to locate themselves in space
relative to a mirror. Therefore, during the mirror illusion when an individual feels agency over the
illusory hand in the mirror, we would expect the individual may feel confused about their location
8

## Page 9

in space relative to the mirror as proprioception and mirror proprioception collide. This is
consistent with the experience of proprioceptive drift as long as the mirror illusion hand and
actual hand are within a certain physical distance of one another - as the actual hand moves
further from the mirror illusion hand, experiences of proprioceptive drift and changes in
ownership/agency are less likely to occur; in this model, this may occur because distances
makes it tougher to convince the brain that the movements of the mirror illusion hand are due to
the right hand.
If a successful reflection prediction indeed facilitates mirror proprioception, reflection prediction
errors or errors with the use of a reflection prediction may cause atypical mirror proprioception.
For example, individuals may not be able to perform everyday visuospatial tasks in front of a
mirror, or may be mirror aversive and avoid environments with mirrors entirely. One example of
this may be seen in a mirror drawing task where individuals are asked to trace shapes while
visually focusing on their reflection in a mirror. One study has shown that some individuals with
autism spectrum disorder (ASD) perform better at this task than neurotypical individuals but
others perform much worse [16].The increased variance in the performance of this task in ASD
individuals may reflect the heterogeneity in the underlying mechanisms that lead to a common
diagnosis of ASD, however, according to our model, performing worse would be expected if
some individuals had atypical mirror proprioception and struggled to successfully locate
themselves in space relative to the mirror.
Additionally, atypical mirror proprioception due to reflection prediction errors or errors with the
use of a reflection prediction may cause an individual to feel located in space somewhere other
than where they are really located. For example, if an individual predicts their reflection is half
the size it actually is this model suggests the individual be confused about where they are
actually located in space relative to the mirror, and that such a reflection prediction error could
be treated with a VR mirror that shows the individual's reflection as half the normal size. If
reflection prediction errors or errors with the use of a reflection prediction do cause an individual
to predict that they are somewhere in space other than where they are actually located in this
way, this could help to explain out-of-body experiences as seen in a variety of individuals. As
one patient with depersonalization disorder explains:
“I look in the mirror and it doesn’t feel like myself I’m looking at. It’s like I’m floating, not
actually experiencing the world, and slowly fading away into nothing. It’s like I’m on
autopilot in somebody’s else body [17](see also [18][19]).” [6]
Therefore, we argue a reflection prediction may allow an individual to understand their location
in space relative to a mirror, and failures in a reflection prediction may cause an individual to not
know where they are located or feel located somewhere other than where they really are
located.
In this section we have discussed how a reflection prediction error may contribute to the
scientific understanding of mirror self-proprioception. We now turn to discuss how a reflection
prediction may contribute to the scientific understanding of imitation.
9

## Page 10

Reflection Predictions and Imitation
The ability of a species to imitate others is also a trait that may depend on the ability of
individuals within that species to successfully predict their own reflection. Unlike mirror
self-recognition and mirror self-proprioception, imitation is not proposed as an automatic product
of a successful reflection prediction, but rather the product of natural selection shaping the brain
to use the reflection prediction for imitation. The ability of an individual to predict their own
reflection before or without seeing it suggests that the reflection prediction is an association
between the motor neurons for an action and the sight of an individual’s reflection making the
contralateral movement. As an association between the motor neurons for an action and
the sight of another (the reflection) performing the contralateral action, an individual may
use the same association in reverse in an inversion of the predictive model [20] to
generate the motor neurons for an action given the sight of another performing that
contralateral action. In other words, a reflection prediction may facilitate the imitative function
of mirror neurons, defined as neurons that react both when an action is performed and when it
is observed.
An objection to this hypothesis that we anticipate is that children imitate others before they
recognize themselves in the mirror and therefore imitation cannot depend on a reflection
association. However the association between the motor neurons for an action and the sight of
another acting may allow for children to substitute others for their reflection before they are
capable of recognizing themselves in the mirror. Therefore while the evolution of imitation may
depend on the emergence of a reflection prediction, imitation itself does not necessarily depend
on the emergence of a reflection prediction developmentally, rather, individuals may also
develop this reflection prediction through observing and interacting with others. One example
supporting this claim is that “being imitated” can improve imitation skills [21,22] in toddlers. A
second example of this interpersonal association is that infants attempt to wipe something off
their faces when watching others wipe something off their faces [23,24]. A third example of this
interpersonal association is the enfacement illusion, where an individual is subjected to a tactile
sensation while witnessing someone else experience the sensation. This synchronic motor
experience simulates a mirror experience and causes individuals to respond to self-other
distinction tasks as though the other’s face was more like their own face [1,25–27], and causes
individuals to report that the others face was experienced as more like “me” than before they
were before the enfacement illusion [28]. According to our model this may be expected because
the synchronic motor experience may cause the individual to process the other as though the
other were their own reflection. This is also one possible alternate explanation for the rubber
hand illusion, where an individual watches a rubber hand undergo the same tactile experience
as their hidden hand and begins to feel agency over the rubber hand.
The idea that children may begin to establish their reflection association via others before mirror
self-recognition suggests additionally that children may build a prediction of themselves through
those they imitate. This can help to explain how individuals who have never had mirror exposure
may still have a prediction of themselves to use for imitation and other traits which may depend
upon a reflection prediction.
10

## Page 11

If imitation depends upon a successful reflection prediction, errors with a reflection prediction or
with the use of a reflection prediction may cause atypical imitation. For example, one
hypothetical reflection prediction error is a non-reversed reflection prediction. A non-reversed
reflection prediction would cause an individual to predict that as they waved their right arm their
reflection would wave its right arm rather than its left arm, as is actually the case in front of a
mirror.
If an individual has associated the movements of the left side of their body with the image of the
left side of the body of others, rather than the right side of the body of others as with a normal
reflection, then the individual may imitate the other incorrectly - if they attempt a contralateral
imitation, they will perform an ipsilateral imitation and vice versa. One disorder which exhibits
symptoms resembling this is autism spectrum disorder. Studies have indicated that autistic
individuals have deficiencies with mirror-image imitation [29], and according to one first person
account this sort of reversed imitation is consistent with the experience of some autistic
individuals. As Willey [30] writes,
“When someone facing me moved their left arm, I moved my right arm. When they
moved their right arm, I moved my left arm and so on and so forth. I knew all along that I
was making a mistake, but no matter what I did and no matter how many times I told
myself things like ‘her right arm equals my left arm,’ I could not transfer the knowledge to
the movement. After a few weeks of bilateral torture, I figured out I might find some
success if I practiced our dance steps from the back row; a vantage point that allowed
me to carbon copy the people who were facing the same direction I was.”
Therefore, while we are not suggesting a non-reversed reflection prediction is the sole cause of
ASD, these specific symptoms do closely match those that would be expected if an individual
did have a non-reversed reflection prediction that caused backwards imitation. One way to
attempt prediction error correction would be through the use of a backwards “anti-mirror”. An
“anti-mirror” is a full-length computer screen that can display one’s reflection as a mirror would.
A backwards “anti-mirror” will cause the individual’s reflection to perform the
ipsilateral/anatomical rather than contralateral/mirror movement. If our claim that some
individuals with autism have a backwards reflection prediction is correct then a backwards
“anti-mirror” should allow these individuals to feel more agency (evaluated using self-report or
other accepted neurobehavioral measures) than a normal mirror because the backwards
anti-mirror allows non-reversed reflection prediction to equal reflection experience, granting the
individual agency over their reflection via sensory attenuation. If true it is unclear whether this
prediction error correction would reverse backwards imitation in those who exhibit this trait,
however, there should be a clear positive correlation between those who exhibit backwards
imitation and those who feel more agency in front of a backwards anti-mirror.
In this section we have discussed how a reflection prediction may contribute to our
understanding of imitation. We now turn to consider how a reflection prediction may contribute
to our understanding of self-awareness.
11

## Page 12

Reflection Predictions and Self-Awareness
One final trait we will argue may depend upon a reflection prediction is self-awareness. Like
imitation but unlike mirror self-recognition and mirror self-proprioception, self-awareness is not
proposed as an automatic product of a successful reflection prediction, but rather the product of
natural selection shaping the brain to use the reflection prediction for self-awareness.
Many have implied a relationship between mirror self-recognition and self-awareness,
suggesting that If an individual is capable of self-directed behavior that they are self-aware. Yet
this claim does not stand up to logical scrutiny. For example, ants and cleaner wrasse fish
exhibit self-directed behavior in front of a mirror [31][13]. Are they self aware? Bats don’t exhibit
self-directed behavior but they do recognize their own echo [32], which is a reflection of sound
akin to how a mirror offers a reflection of light. Are bats self aware? Dog’s recognize their own
scent; does this mean they are self aware [33]? These rhetorical questions are of course
impossible to answer without defining self-awareness. Gallup et al. write:
“In its most rudimentary form self-awareness is the ability to become the object of your
own attention. When you see yourself in a mirror, you are literally the object of your own
attention, but most organisms respond to themselves in mirrors as if confronted by
another organism. The ability to correctly infer the identity of the image in the mirror
requires a pre-existing sense of self on the part of the organism making that inference.
Without a sense of self, how would you know who you were seeing when confronted with
your reflection in a mirror?” [34]
Yet as we showed above, in a predictive processing model an individual simply needs to be able
to predict their reflection in order to feel agency over their reflection. This implies
self-recognition, mirror and otherwise, does not require a sense of self as suggested by Gallup
et al., and indeed rather suggests self-recognition may be an evolutionary precursor for a sense
of self.
To explain this we suggest adding one word to Gallup’s definition of self-awareness:
self-awareness is the ability to become the object of your own social attention. This correlates
with the psychological concept of objective self-awareness, defined as an attention focused on
the self as a social object [35], and also in our opinion is more consistent with the colloquial use
of the term self-awareness than Gallup’s definition. A reflection prediction offers the potential for
an individual to become the object of their own social attention because their own physical
behavior has emerged to themselves in a third person perspective; therefore, they have a
mechanism with which they may react to their behavior the same way they would react to the
behavior of another. George Herbert Mead argued an individual becomes self aware only when
they “take the attitude of the other” on their own behavior. As he writes:
“When the response of the other becomes an essential part in the experience or conduct
of the individual; when taking the attitude of the other becomes an essential part in his
behavior- then the individual appears in his own experience as a self; and until this
happens he does not appear as a self.“ [36]
12

## Page 13

Mead and others have suggested that only modes of communication which are naturally
reflexive such as speech can allow an individual to take the attitude of the other because
non-reflexive communication such as facial gestures cannot be processed similarly by an actor
as it is by an observer [37], however, we propose that a reflection prediction can bridge this
theoretical gap because evolution may allow a reflection prediction to become reflexive, allowing
an individual's own gestural behavior to emerge to themselves as though they were watching
themselves perform the behavior from the perspective of another, offering a mechanism for the
individual to take the attitude of the other on their own behavior and therefore emerge as a
social other to themselves, allowing objective self-awareness.
Combining the concept of a reflexive reflection prediction with Mead’s logic suggests an
individual may use their reflection prediction to react to their own behavior the same way they
would react to another performing the behavior. This reaction then can be used by the individual
to predict the internal reaction of others to the individual's own behavior. G.H. Mead would say
the gesture has become a “significant gesture” because the gesture means the same thing to
both the actor and witness. We refer to this reaction as the “witness” reaction because the
actor is proposed to react due to becoming witness to their own behavior via a reflexive
reflection prediction, as an actual witness would witness their behavior. The witness
reaction implies the brain may have “witness neurons” which will appear similarly as “mirror
neurons” because they are both proposed to be simultaneously activated neurons which will be
shared between an actor and a witness; as mirror neurons are proposed to occur when a
witness activates the same neurons as an actor to simulate the actor, witness neurons are
proposed to occur when an actor predictively reacts to their behavior the same way a witness
may react to the individuals behavior to simulate the witness. This has at least three important
implications in regards to self-awareness.
1) Individuals may use the witness reaction for action monitoring, defined as the process for
evaluating the appropriateness of one’s actions [38], and response execution, the
process of performing those actions, and combine these two processes as a mechanism
for action simulation, simulating the reaction of others. As an individual can imitate
others, individuals may perform response execution during this process of action
simulation by imitating the reflection prediction.that arouses in them the reaction they
wish to arouse in others.
2) When an individual chooses which behavior to perform based on the internal reaction
they wish to arouse in others, this internal reaction has become the intention of the
behavior. Therefore we propose during action simulation the individual is associating the
intention of their behavior with the sight of another (their reflection) performing that
behavior; furthermore, that individuals may use this association between intention and
reflection prediction in reverse via an inversion of the predictive model [20] to infer the
intention others are attempting to arouse in them.
3) The individual has emerged as another to themselves, therefore, as individuals infer the
intentions of others, they may infer their own intentions during the witness reaction, as
they predict others thoughts, they may predict their own thoughts during the witness
13

## Page 14

reaction, as they empathize with others, they may empathize with themselves during the
witness reaction. As they narrate what others are doing, they may narrate what they
themselves are doing during the witness reaction. Etc. Therefore the intention of an
individual's behavior is not limited to one type of internal reaction; for example, an
individual may use the witness reaction to intend to arouse thoughts in another or may
intend to arouse emotions in another. Herbert Blumer, one of Mead’s students, would call
this a “process of indication”. As he writes:
“A very different view of human society is introduced when one recognizes that
the initiator and the recipient of a significant gesture indicate the meaning of the
gesture and then indicate the action to be taken on the basis of that meaning.
This process of indication, allowing the possibility of choice and diverse
response, breaks what would otherwise be the direct determination of action by
society. It changes the position of the initiator and the recipient of a significant
gesture from being mere agents of a transmitting society to the position of being
actors who are confronting, defining, and handling their situations.“[39]
Taken together, these implications of the witness reaction provide an explanatory mechanism to
better understand not only self-awareness but also; theory of mind, defined as the ability of an
individual to attribute mental states such as intentions, beliefs, thoughts, to others and
themselves; mentalizing, defined as “the capacity to reflect on and interpret one's own behavior
and that of others based on intentional internal mental states, such as beliefs, thoughts, and
emotions” [40,41]; and thinking through other minds (TTOM), defined as “the process of inferring
other agents’ expectations about the world and how to behave in social context” [42].
Furthermore, these implications show that it may not be accurate to think of objective
self-awareness and theory of mind as two separate concepts rather than two sides of the same
coin, because after an individual emerges to themselves as an other via a reflexive reflection
prediction’s witness reaction, self and other are inextricably linked; as an individual understands
and can define others better, they may understand and define themselves better, and as an
individual understands and defines themselves better, they may understand and define others
better. For example, as an individual learns to associate a name with the faces of others, they
may have learned to associate their own name with their own face. As they develop feelings
about another, they may develop feelings about themselves (resulting in positive or negative
self-esteem). As they judge the actions of others by a moral code, they may judge their own
actions by a moral code, and understand that others are judging their own actions by a moral
code. Etc. Therefore, self-awareness may be considered using other-awareness on oneself,
and theory of mind may be considered using self-awareness on others.
Finally, if a reflection prediction facilitates self-awareness then a reflection prediction error or the
error in the use of a reflection prediction may cause atypical self-awareness. For example, if an
individual has a reflection prediction that is not reflexive, that individual may not be able to
predict how others will react to their behavior and may therefore demonstrate behavior that is
inappropriate for a given social situation. Furthermore, since a reflexive reflection prediction is
proposed as a necessary mechanism for an individual to associate their own intentions with
14

## Page 15

their reflection, an individual without a reflexive reflection prediction may not be able to use this
association in reverse and therefore may struggle to understand the intentions of others.
This view of atypical self-awareness offers a new way of understanding atypical theory of mind
and intention understanding as seen in certain individuals. For example, there have been
multiple theories implicating mirror neuron dysfunction in intention understanding for individuals
with autism spectrum disorder [43]. However, witness neurons are proposed to appear similarly
in the brain as mirror neurons and witness neurons offer an easy explanation of intention
understanding, as presented in this paper. Therefore, researchers may be attributing atypical
intention understanding to atypical mirror neurons when they should instead be attributing
atypical intention understanding to atypical witness neurons. It is therefore important to clearly
differentiate between which brain processes involve mirror neurons and which involve witness
neurons.
Another example of how a reflection prediction error or the error in the use of a reflection
prediction may cause atypical self-awareness is if an individual consciously processes their
reflection prediction as another individual this may cause them to conceptualize their predicted
or actual behavior as the actual behavior of another. Two examples of individuals
conceptualizing their predicted behavior as that of another may include auditory verbal
hallucinations, when an individual does not recognize their thoughts and therefore claims they
are the thoughts of another, and syndrome of subjective doubles, where an individual claims
they have a doppelganger that is leading it’s own life.
One example of an individual conceptualizing their actual behavior as that of another may be
instances of multiple personalities as seen in dissociative identity disorder. An example of this
can be seen in the case of Donna Williams, who has been diagnosed with autism and
dissociative identity disorder and has two alternate personalities; Carol and Willie. The
experience of being Carol started with atypical mirror self-recognition, as indicated by this
autobiographical quote:
“Carol came in through the mirror. Carol looked just like me, but the look in her eyes
betrayed her identity. It was Carol all right. I began to talk to her, and she copied me. I
was angry. I didn't expect her to do that. My expression asked her why, and hers asked
me. I figured that the answer was a secret.” [44]
This experience supports the suggestion that a reflection prediction error or error with the use of
a reflection prediction may cause atypical self-awareness. Another example of an individual
conceptualizing their actual behavior as that of another due to a mirror is seen in how a mirror
can induce instances of asomatognosia in stroke recovery patients [45]. Asomatognosia occurs
when an individual loses recognition of one half of their body. One common symptom of
asomatognosia is somatoparaphrenia, where the individual loses agency over this part of their
body. In our model this may be expected if the mirror causes a reflection prediction error on half
of their reflection. In this instance the individual will not be able to successfully predict half of
their reflection, so they may not be able to recognize half of their body relative to mirrors and
thus may have a failure of self-awareness for that half of their body.
15

## Page 16

To summarize: a reflection prediction may facilitate self-awareness by letting an individual
emerge as another to themselves, and reflection prediction errors or error with the use of a
reflection prediction may cause atypical self-awareness as seen in a variety of individuals.
Conclusion
In this paper we have presented a predictive processing analysis of the mirror test, argued that
a reflection prediction may facilitate mirror self-recognition, mirror proprioception, imitation, and
self-awareness, and examined how reflection prediction errors or errors with the use of a
reflection prediction may contribute to atypical mirror self-recognition, atypical mirror
proprioception, atypical imitation, and atypical self-awareness.
Though grounded firmly in the predictive processing model built by others, this paper is largely
speculative and we have offered novel perspectives on a variety of topics that have been
discussed for decades or longer. These perspectives may wind up being correct or incorrect;
regardless, we believe an open discussion of these perspectives can greatly advance the
scientific understanding of the topics discussed in this paper such as mirror self-recognition and
self-awareness. This is particularly important due to the sheer number of symptoms of disorders
that this model may help to understand and possibly even treat, as we have attempted to point
out throughout this paper.
Many questions remain concerning this theory. Some examples include:
●
Where and how is the reflection prediction represented or implemented in the brain? Is it
separate entirely from the conscious body image or are they related in some way?
●
To what degree are individuals consciously aware of altered reflection predictions (i.e.
would someone with BIID tell you that when they closed their eyes and pictured their
reflection in the mirror, it is missing a limb)?
●
What exactly is a reflection prediction error? How or in what circumstances can a
reflection prediction be altered?
●
What causes perceptual inference to vary or fail?
●
Is there a way to treat reflection prediction errors other than reflection prediction error
correction? Is there a quantitative way to measure agency over a reflection in the brain?
As a first next step we urge the scientific community to gather more data and think more
critically about mirror self-recognition. The inability of an individual to recognize themselves in
the mirror is currently viewed by the scientific literature as merely a symptom of various
disorders and because of this there is a shortage of data on mirror self-recognition in the
scientific literature. If the scientific community instead considers that an individual not
recognizing themselves in the mirror may be tied to the atypical functioning of other traits, as
this model suggests, we believe significant progress can be made in understanding both
individual symptoms/disorders and how various symptoms/disorders relate to one another.
Finally we will conclude by pointing out that much of the model that we have presented in this
paper can be considered analogously for reflexive modes of communication as a path towards
16

## Page 17

self-recognition and self-awareness under the correct evolutionary circumstances. For example,
a species that communicates with odors may smell their own odor and treat their own odor as
the odor of another, learn to predict how that other will respond to them, gain agency over the
odor and ignore it via sensory attenuation, and use the odor prediction reflexively to generate a
witness reaction in order to predict how others will respond to their own odor. A species that
communicates with sound may treat their speech (or echo) as the speech of another, learn to
predict how that other will respond to them, gain agency over the speech (or echo) and ignore it
via sensory attenuation, and use the speech prediction (or echo prediction) reflexively to
generate a witness reaction in order to predict how others will respond to their own speech. Etc.
These modes of communication are different from bodily movement because they are naturally
reflexive and so this seems like a likely path for self-recognition and self-awareness to arise in
other species under the correct evolutionary circumstances.
17

## Page 18

References
1.
Apps MAJ, Tsakiris M. The free-energy self: a predictive coding account of self-recognition.
Neurosci Biobehav Rev. 2014;41: 85–97. doi:10.1016/j.neubiorev.2013.01.029
2.
Gallop GG Jr. Chimpanzees: self-recognition. Science. 1970;167: 86–87.
doi:10.1126/science.167.3914.86
3.
Perrykkad KS. Self in Autism: A Predictive Perspective. [cited 1 Mar 2022]. Available:
https://scholar.archive.org/work/yv36ywkqj5h45ieyyefzhqbvgi/access/wayback/https://au-ea
st.erc.monash.edu.au/fpfiles/28688580/Thesis_Perrykkad_FINAL.pdf?AWSAccessKeyId=e
00145a6f706457aab45051570081e49&Expires=1625774089&Signature=4OBqrrgdc%2B%
2BRi%2FTORAHLPb5vpQI%3D
4.
Leptourgos P, Corlett PR. Embodied Predictions, Agency, and Psychosis. Front Big Data.
2020;3: 27. doi:10.3389/fdata.2020.00027
5.
Gallagher I I. Philosophical conceptions of the self: implications for cognitive science.
Trends Cogn Sci. 2000;4: 14–21. doi:10.1016/s1364-6613(99)01417-5
6.
Ciaunica A, Seth A, Limanowski J, Hesp C, Friston K. I overthink—therefore I am not: An
Active Inference Account of Altered Sense of Self and Agency in Depersonalisation
Disorder. 2021. doi:10.31234/osf.io/k9d2n
7.
Haggard P. Sense of agency in the human brain. Nat Rev Neurosci. 2017;18: 196–207.
doi:10.1038/nrn.2017.14
8.
Plotnik JM, de Waal FBM, Reiss D. Self-recognition in an Asian elephant. Proc Natl Acad
Sci U S A. 2006;103: 17053–17057. doi:10.1073/pnas.0608062103
9.
Swartz KB, Evans S. Not all chimpanzees (Pan troglodytes) show self-recognition.
Primates. 1991;32: 483–496. doi:10.1007/BF02381939
10. Povinelli DJ, Rulf AB, Landau KR, Bierschwale DT. Self-recognition in chimpanzees (Pan
troglodytes): Distribution, ontogeny, and patterns of emergence. Journal of Comparative
Psychology. 1993. pp. 347–372. doi:10.1037/0735-7036.107.4.347
11. Heschl A, Burkart J. A new mark test for mirror self-recognition in non-human primates.
Primates. 2006;47: 187–198. doi:10.1007/s10329-005-0170-8
12. Constant A, Bervoets J, Hens K, Van de Cruys S. Precise Worlds for Certain Minds: An
Ecological Perspective on the Relational Self in Autism. Topoi. 2020;39: 611–622.
doi:10.1007/s11245-018-9546-4
13. Kohda M, Hotta T, Takeyama T, Awata S, Tanaka H, Asai J-Y, et al. Cleaner wrasse pass
the mark test. What are the implications for consciousness and self-awareness testing in
animals? bioRxiv. 2018. p. 397067. doi:10.1101/397067
14. Turbyne C, Koning P de, Zantvoord J, Denys D. Body integrity identity disorder using
augmented reality: a symptom reduction study. BMJ Case Rep. 2021;14.
doi:10.1136/bcr-2020-238554
18

## Page 19

15. Tajima D, Mizuno T, Kume Y, Yoshida T. The mirror illusion: does proprioceptive drift go
hand in hand with sense of agency? Front Psychol. 2015;6: 200.
doi:10.3389/fpsyg.2015.00200
16. Salowitz NMG, Eccarius P, Karst J, Carson A, Schohl K, Stevens S, et al. Brief report:
visuo-spatial guidance of movement during gesture imitation and mirror drawing in children
with autism spectrum disorders. J Autism Dev Disord. 2013;43: 985–995.
doi:10.1007/s10803-012-1631-8
17. Perkins J. Life on Autopilot: A Guide to Living with Depersonalization Disorder. Jessica
Kingsley Publishers; 2021. Available:
https://play.google.com/store/books/details?id=dYofEAAAQBAJ
18. Simeon D, Abugel J. Feeling Unreal: Depersonalization Disorder and the Loss of the Self.
Oxford University Press; 2006. Available:
https://play.google.com/store/books/details?id=M-Cgs_T7Cl8C
19. Sierra M. Depersonalization: A New Look at a Neglected Syndrome. Cambridge University
Press; 2009. Available: https://play.google.com/store/books/details?id=qOat1nDGZAkC
20. Friston KJ, Preller KH, Mathys C, Cagnan H, Heinzle J, Razi A, et al. Dynamic causal
modelling revisited. Neuroimage. 2019;199: 730–744.
doi:10.1016/j.neuroimage.2017.02.045
21. McDuffie A, Turner L, Stone W, Yoder P, Wolery M, Ulman T. Developmental correlates of
different types of motor imitation in young children with autism spectrum disorders. J Autism
Dev Disord. 2007;37: 401–412. doi:10.1007/s10803-006-0175-1
22. Ingersoll B. The Social Role of Imitation in Autism: Implications for the Treatment of
Imitation Deficits. Infants Young Child. 2008;21: 107.
doi:10.1097/01.IYC.0000314482.24087.14
23. Lewis M, Brooks-Gunn J. Toward a theory of social cognition: The development of self. New
Dir Child Adolesc Dev. 1979;1979: 1–20. doi:10.1002/cd.23219790403
24. Johnson DB. Self-recognition in infants. Infant Behav Dev. 1983;6: 211–222.
doi:10.1016/S0163-6383(83)80028-9
25. Tsakiris M. Looking for myself: current multisensory input alters self-face recognition. PLoS
One. 2008;3: e4040. doi:10.1371/journal.pone.0004040
26. Sforza A, Bufalari I, Haggard P, Aglioti SM. My face in yours: Visuo-tactile facial stimulation
influences sense of identity. Soc Neurosci. 2010;5: 148–162.
doi:10.1080/17470910903205503
27. Maister L, Cardini F, Zamariola G, Serino A, Tsakiris M. Your place or mine: shared sensory
experiences elicit a remapping of peripersonal space. Neuropsychologia. 2015;70:
455–461. doi:10.1016/j.neuropsychologia.2014.10.027
28. Tajadura-Jiménez A, Grehl S, Tsakiris M. The other in me: interpersonal multisensory
stimulation changes the mental representation of the self. PLoS One. 2012;7: e40682.
doi:10.1371/journal.pone.0040682
19

## Page 20

29. Avikainen S, Wohlschläger A, Liuhanen S, Hänninen R, Hari R. Impaired mirror-image
imitation in Asperger and high-functioning autistic subjects. Curr Biol. 2003;13: 339–341.
doi:10.1016/s0960-9822(03)00087-3
30. Willey LH. Pretending to be Normal: Living with Asperger’s Syndrome (Autism Spectrum
Disorder) Expanded Edition. Jessica Kingsley Publishers; 2014. Available:
https://play.google.com/store/books/details?id=Q7RNBAAAQBAJ
31. Cammaerts Tricot M-C, Cammaerts R. Are ants (Hymenoptera, Formicidae) capable of self
recognition ? J Sci. 2015;5: 521–532. Available:
https://difusion.ulb.ac.be/vufind/Record/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/219269/Details
32. Amichai E, Blumrosen G, Yovel Y. Calling louder and longer: how bats use biosonar under
severe acoustic interference from other bats. Proc Biol Sci. 2015;282: 20152064.
doi:10.1098/rspb.2015.2064
33. Horowitz A. Smelling themselves: Dogs investigate their own odours longer when modified
in an “olfactory mirror” test. Behav Processes. 2017;143: 17–24.
doi:10.1016/j.beproc.2017.08.001
34. Gallup GG Jr, Anderson JR, Shillito DJ. The mirror test. The cognitive animal: Empirical and
theoretical perspectives on animal cognition. 2002; 325–333. Available:
https://courses.washington.edu/ccab/Gallup%20on%20mirror%20test.pdf
35. Duval S, Wicklund RA. A theory of objective self awareness. 1972;238. Available:
https://psycnet.apa.org/fulltext/1973-26817-000.pdf
36. Mead GH, Huebner DR, Joas H. Mind, self, and society: The definitive edition. 1st ed.
Morris CW, editor. Chicago, IL: University of Chicago Press; 2015.
doi:10.7208/chicago/9780226112879.001.0001
37. Wiley N. Reflexivity in George Herbert Mead. Studies in Symbolic Interaction. 2021. pp.
61–72. doi:10.1108/s0163-239620210000052005
38. Suzuki K, Kita Y, Sakihara K, Hirata S, Sakuma R, Okuzumi H, et al. Uniqueness of action
monitoring in children with autism spectrum disorder: Response types and temporal
aspects. J Clin Exp Neuropsychol. 2017;39: 803–816.
doi:10.1080/13803395.2016.1266308
39. Blumer H. George Herbert Mead and Human Conduct. Rowman Altamira; 2004. Available:
https://play.google.com/store/books/details?id=u3BOxaRysbYC
40. Fonagy P, Target M, Steele H, Steele M. Reflective-functioning manual, version 5.0, for
application to adult attachment interviews. London: University College London. 1998;10.
Available:
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.466.3207&rep=rep1&type=pdf
41. Gergely G, Fonagy P, Jurist E. Affect regulation, mentalization, and the development of the
self. International Journal of. 2002. Available:
http://publications.ceu.hu/publications/gergely/2002/10159
42. Veissière SPL, Constant A, Ramstead MJD, Friston KJ, Kirmayer LJ. Thinking through
20

## Page 21

other minds: A variational approach to cognition and culture. Behav Brain Sci. 2019;43:
e90. doi:10.1017/S0140525X19001213
43. Yates L, Hobson H. Continuing to look in the mirror: A review of neuroscientific evidence for
the broken mirror hypothesis, EP-M model and STORM model of autism spectrum
conditions. Autism. 2020;24: 1945–1959. doi:10.1177/1362361320936945
44. Williams D. Nobody Nowhere: The Remarkable Autobiography of an Autistic Girl. Jessica
Kingsley Publishers; 2009. Available:
https://play.google.com/store/books/details?id=MD4pDgAAQBAJ
45. Paysant J, Beis JM, Le Chapelain L, André JM. Mirror asomatognosia in right lesions stroke
victims. Neuropsychologia. 2004;42: 920–925. doi:10.1016/j.neuropsychologia.2003.12.001
21


---
*Extraction method: pymupdf*