Content uploaded by Rich David Moccia
Author content
All content in this area was uploaded by Rich David Moccia on Nov 29, 2019
Content may be subject to copyright.
An evaluation of current perspectives on consciousness and
pain in fishes
Kristopher Paul Chandroo, Stephanie Yue & Richard David Moccia
Aquaculture Centre, Department of Animal and Poultry Sciences, University of Guelph, Guelph, ON, Canada N1G 2W1
Introduction 282
The neocortex and the neural correlates of consciousness 283
Convergence, homology and evolutionary psychology 286
Pain perception and brain structure 288
Cognition and behaviour in fishes 291
Conclusions 293
Acknowledgements 293
References 293
Abstract
There is growing societal and scientific interest in the welfare status of fish used for
commercial enterprise. As animal welfare is primarily concerned with the quality of
life of a conscious, sentient organism, the question of whether fishes are even capable
of consciousness must first be addressed in order to assess their welfare status.
Recently, there has been a resurgence of research investigating the biological basis for
human consciousness, and our current understanding of the cognitive mechanisms
underlying fish behaviour has likewise improved significantly. Combined, these
research perspectives create an opportunity to better comprehend the phylogeny of
traits associated with consciousness, as well as the emergence of consciousness itself
during vertebrate evolution. Despite the availability of this literature, contemporary
reviews or published studies investigating the probability of conscious states
occurring in fishes often do so without considering new perspectives or data. In
this paper, we review and critique recent publications that report equivocal
conclusions favouring the absence or presence of consciousness in various fishes.
By introducing other data into these analyses, we demonstrate that there are
alternative perspectives which support the existence of consciousness in fishes as a
plausible concept. An accurate assessment of the mental capacity of fishes will require
enhanced knowledge of their forebrain neuroanatomy, an understanding of how
such structures mediate behavioural responses, and an analysis of that information
within the context of contemporary theory related to the evolution of consciousness
in higher vertebrates.
Keywords animal welfare, consciousness, fish, pain, sentience
Correspondence:
Richard David Moc-
cia, Aquaculture Cen-
tre, Department of
Animal and Poultry
Sciences, University of
Guelph, Guelph, ON,
Canada N1G 2W1
Tel.: (519) 824-4120
(ext. 56216)
Fax: (519) 767-0573
E-mail: rmoccia@
uoguelph.ca
Received 26 Jan 2004
Accepted 31 Aug 2004
2004 Blackwell Publishing Ltd 281
F I S H and F I S H E R I E S , 2004, 5, 281–295
Introduction
There is growing societal and scientific interest in
the welfare status of fish used in commercial
enterprise. As animal welfare is concerned with
the quality of life of a conscious and sentient
organism, the question of whether or not fishes
are capable of conscious states must be addressed in
order to evaluate their welfare status. Surprisingly,
there is no universally accepted definition of con-
sciousness as it applies across the spectrum of
vertebrate phyla (Searle 2000). However, it is
generally agreed among researchers that conscious-
ness refers to a mental state of awareness of internal
and external stimuli. Depending upon the quanti-
tative or qualitative degree of awareness that is
present in an organism, consciousness can also be
described as existing in a primary or extended state
(Lindahl 1997). Primary consciousness may be
defined as the ability to generate a mental scene
in which diverse information is integrated for the
purpose of directing behaviour of self (Edelman and
Tononi 2000). Extended consciousness is thought to
involve ‘higher order’, advanced cognitive abilities
that involve, for example, a linguistic capability or
self-consciousness as self-knowledge (Zeman 2001).
Regardless of whether an animal is thought to have
primary or extended consciousness, both designa-
tions imply that they are sentient or self-aware
organisms.
The probability of consciousness occurring in
animals is typically assessed by comparing their
neuroanatomical, behavioural and physiological
characteristics with an array of human (or other
well-studied mammalian) biological features that
are closely associated with consciousness and emo-
tional states. Comparative investigations of this kind
have recently been published for fish species (Rose
2002; Sneddon 2003; Chandroo et al. 2004). The
conclusions attained by simplistic, comparative
assessment are often controversial, and much
debate exists with respect to the accuracy of this
analytical approach (Rolls 2000). This is especially
true when such evaluations are focused on ances-
tral or ‘primitive’ vertebrate species, such as fish.
Reasons for this controversy are wide ranging, and
include philosophical disagreement concerning
what comprises a legitimate form of scientific study
(Searle 1998), as well as an historical bias regarding
the interpretation of results derived from animal
behaviour research (Griffin 1998; Schilhab 2002).
Of equal controversy, is the disagreement with
regard to the phylogenetic relationships between
similar biological structures and their putative
functions among distantly related species (Striedter
2002).
Until recently, the scientific investigation into the
existence of conscious states in fishes has been
compromised due to a lack of primary research
investigating the aetiology of human consciousness,
as well as a limited supply of comparative studies
examining brain structure and cognition in fish
species. However, a simultaneous resurgence of
research investigating the neurophysiological basis
of human consciousness, including telencephalic
neuroanatomy and the underlying cognitive mech-
anisms of fish behaviour, has provided basic infor-
mation that permits the phylogeny of biological
traits associated with consciousness and conscious-
ness itself to be studied more objectively (Butler and
Hodos 1996; Baars 2002). Despite the availability of
this information, current reviews or published
studies investigating the probability of conscious
states occurring in fishes, often do so without
considering innovative, applicable data or alternat-
ive perspectives in its interpretation.
In this paper, we review and critique a number of
recent publications that have reported equivocal
conclusions on the existence of consciousness in
fishes. We primarily critique the work of Rose
(2002), and also address the work of Cabanac
(1999), Sneddon (2003), Sneddon et al. (2003) and
Cabanac and Cabanac (2000). Rose (2002) argues
that it is most likely impossible for fish to experience
pain or fear, while in contrast, Sneddon et al.
(2003) provide anatomical, physiological and beha-
vioural evidence that demonstrates nociception in
fish, concluding that fish can also perceive pain.
Based on empirical studies focused on physiological
and behavioural responses, Cabanac (1999) sug-
gests that fish do not have the ability for conscious-
ness and emotion. The introductory list of issues and
concepts found within (Rose 2002, p. 2) illustrate
Rose’s position that ‘anthropomorphic thinking
undermines our understanding of other species’,
and that ‘an evolutionary perspective is essential to
understanding the neurobehavioural differences
between fishes and humans.’ We agree with Rose
that an unjustified ascribing of mental abilities to
animals and the lack of an evolutionary perspective
will lead to inaccurate conclusions with regard to
the mental life of any animal. However, the evolu-
tionary perspective that Rose presents is heavily
dependent upon contrasts and anthropocentric
Consciousness and pain perception in fish K P Chandroo et al.
282 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
arguments. Primary literature on the neurobiolog-
ical features and learning behaviour of fishes seem
lacking in Rose’s review, despite the fact that there
is a plethora of suitable papers available (Moccia
and Chandroo 2003). Within his review, one
can find many discussions where neurobiological
literature pertaining to consciousness and pain
perception in humans are utilized as if they were
well-established facts, whereas the actual research
behind those subjects or ‘facts’ are often hypothet-
ical, preliminary or controversial. But it is our hope
that his foundation paper, in addition to this current
review, will help to open up the meaningful and
critical dialogue relevant to a better understanding
of consciousness and pain perception in fish.
The neocortex and the neural correlates of
consciousness
The neocortex is a heterogeneous, laminated brain
structure that comprises much of the cerebral cortex
in humans (Nieuwenhuys 1994). The neocortex
allows for sophisticated sensory processing, motor
functions and is associated with distinctive human
cognitive abilities. The role of cortical and non-
cortical brain structures in the generation of con-
scious states in humans has been the subject of
intense debate and study (Kanwisher 2001; Baars
2002). We interpret the central thesis in Rose
(2002) to be that conscious animals have a
neocortex and animals without a neocortex, such
as fishes, are by default, incapable of consciousness.
The information presented in Rose (2002) pertain-
ing to nociception, fear and learning processes in
fish are eventually tied to the central theme that the
neocortex is a prerequisite for any of these processes
to ‘reach’ consciousness. Edelman and Tononi
(2000), a reference source cited frequently within
Rose (2002), states that, ‘Many neuroscientists
have emphasized particular neural structures whose
activity correlates with conscious experiences. It is
not surprising that different neuroscientists end up
favouring different structures. As we shall see in a
number of cases, it is likely that the workings of
each structure may contribute to consciousness, but
it is a mistake to expect that pinpointing particular
locations in the brain or understanding intrinsic
properties of particular neurons themselves, will
explain why their activity does or does not contrib-
ute to conscious experience. Such an expectation is
a prime example of a category error, in the specific
sense of ascribing to things properties they cannot
have’ (Edelman and Tononi 2000, p. 19). Rose
clearly, and for good reason favours the human
neocortex as the structure of choice when it comes
to attributing consciousness to a particular brain
region, and suggests that specific areas of the
neocortex are crucial for consciousness to occur
(Rose 2002, p. 31). In order to support this claim,
Rose primarily presents ‘global workspace’ theory
(Baars 2002) or the ‘dynamic core hypothesis’ of
Edelman and Tononi (2000), as well as evidence
from clinical studies of humans afflicted with
chronic vegetative states due to catastrophic brain
injury (Laureys et al. 2000).
Although providing a brief, generalized descrip-
tion, Rose never sufficiently explains how, or why,
the neocortex is thought to be responsible for
consciousness, and careful examination of the
neural-based theories of consciousness yields a
viewpoint that does not necessarily support his
arguments. Edelman and Tononi (2000), Laureys
et al. (2000) and other work cited within Rose
(2002) commonly implicate the thalamocortical
system, and not the neocortex per se as the essential
neural substrate required for consciousness. This is
not a trivial detail, because as we explain later, it is
precisely this interpretational difference that permits
valid, alternative suggestions with regard to the
neural requirements and evolutionary history of
neural systems hypothesized to support conscious-
ness. Theories proposed by Edelman and Tononi
(2000) or tested by Laureys et al. (2000), account
for the fundamental properties of consciousness by
linking them to a particular type of neuronal
process found primarily within the thalamocortical
system (Tononi and Edelman 1998). The neuronal
process within the thalamocortical system that may
account for key properties of consciousness, is
essentially described as the widespread integration
of differentiated brain areas or functions (Baars
2002). A key tenet of the dynamic core hypothesis
or other theories describing related neuronal pro-
cesses, is that consciousness is ‘generated’ by a
neural process per se, and as such is not accurately
characterized as a specific thing or a location
(Tononi and Edelman 1998). Therefore, if a nervous
system has the appropriate characteristics that can
support this process in theory, then it is appropriate,
from a neurobiological perspective, to consider that
this nervous system has the potential to ‘generate’
consciousness. Depending upon which hypothesis
one ascribes to, the appropriate neuronal charac-
teristics could include a variable level of complexity
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 283
that reflects the interplay between functional segre-
gation and integration within a neural system
(Tononi et al. 1994), the level of degeneracy or
redundancy within a neural system (Tononi et al.
1999), the process of neural signal re-entry (Sporns
et al. 1991), particular thalamic functions (Llina
´s
et al. 1998), specific neural activity synchronized
at a particular gamma frequency (Sewards and
Sewards 2001) and other features exhibited by
vertebrate brains (Zeman 2001). It seems that Rose
(2002) offers no cyto-architectural or neurophysi-
ological data on the forebrain of fishes that can be
used to argue whether or not this brain region can
support any of the neuronal process mentioned
above. Without this analysis, Rose’s conclusion
that, ‘It is a neurophysiological impossibility for fish
to have consciousness’, is at best, unsubstantiated.
Instead of a data-based or theoretical analysis, Rose
relies on the belief that the fish brain is well
understood and thus it is highly implausible that it
could support consciousness (Rose 2002, p. 24).
The argument suggests that fish forebrains have
‘diminutive’ dimensions (Rose 2002, p. 10), or ‘poor
differentiation’ (Rose 2002, p. 28). Perplexingly, the
discussion presented in Rose (2002) that does
incorporate a selection of primary literature con-
cerning the central nervous system of fishes is
focused on the spinal cord and brain stem (Rose
2002, pp. 9–10, 22–23). Preliminary investigations
into fish neurobiology suggest that adequate infor-
mation currently exists to equally include or
exclude the fish forebrain from having the capacity
to support consciousness as defined by contempor-
ary theory (Chandroo et al. 2004). Thus, the
question of whether or not the nervous system of
fish permits consciousness, from a purely neurobio-
logical perspective, remains a very open question in
our opinion. Innovative application of brain ima-
ging techniques (Baars 2002) to fish species may
provide new insights into the function of the fish
telencephalon as it relates to neural processes
associated with conscious states.
Many arguments within Rose (2002) rely upon
controversial or untested interpretations of the
neural-based theories of consciousness. In explain-
ing why the neocortex exclusively is critical for
consciousness, Rose asserts that there is clear and
extensive evidence demonstrating that the human
neocortex satisfies several, essential ‘functional
criteria’, namely its unique structural features, that
permit the existence of ‘widely distributed brain
activity that is simultaneously diverse, temporally
coordinated and of high informational complexity’
(Rose 2002, p. 7). However, it seems this argument
is entirely circular: the human neocortex satisfies
the critical, ‘functional criteria’ for consciousness,
because the ‘functional criteria’ for consciousness
are directly derived from the anatomy of the human
neocortex. The majority of studies examining the
neural correlates of consciousness do not support
Rose’s claim that ‘the neurological basis of human
consciousness is becoming increasingly well under-
stood and is known to depend on functions of the
neocortex’ (Rose 2002, p. 31), or that ‘the funda-
mental neural requirements for pain and suffering
are now known’ (Rose 2002, p. 33) (Damasio
1998; Llina
´set al. 1998; Searle 1998; Tononi and
Edelman 1998; Searle 2000; Jack and Shallice
2001; Kanwisher 2001; Parvizi and Damasio 2001;
Zeman 2001; Baars 2002). Exactly how or why
certain brain areas are associated with conscious-
ness or pain perception is still largely controversial
(Block 2001; Dennett 2001), and explanations for
the associations are not by any means exclusive to a
single theory or particular brain region (Sewards
and Sewards 2000; John 2001). Rose reports that
consciousness ‘requires structurally differentiated
neocortical regions with great numbers of exactly
interconnected neurons’ (Rose 2002, p. 24), and
that ‘the type of neocortex most essential to
consciousness, i.e. the non-sensory association cor-
tex, comprises the vast majority of the human
cerebral cortex’ (Rose 2002, p. 7, 31). However, the
dynamic core hypothesis as proposed by Tononi and
Edelman (1998), and cited by Rose to defend his
argument, actually reports that the term dynamic
core deliberately does not refer to a unique, invari-
ant set of brain areas and that the core may change
in composition over time. The dynamic core is also
not necessarily restricted to the thalamocortical
system, which is an important concept. Tononi and
Edelman (1998) state that as neural participation in
the dynamic core depends upon shifting functional
connectivity among groups of neurones, rather
than on anatomical proximity, the composition of
the core can transcend traditional anatomical
boundaries. As Rose equates consciousness exclu-
sively with the neocortex, we suggest that his use of
theoretical neurobiology is misleading and may be
unsuitable for comparative assessment of fish brain
function.
In other attempts to single out the neocortex as
the exclusive structure enabling consciousness, Rose
uses the literature to dissect the thalamocortical
Consciousness and pain perception in fish K P Chandroo et al.
284 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
system (Rose 2002, p. 7, 13, 18). This is carried out
to distinguish the neocortex from subcortical brain
areas, implying that it is not significant that many
animals have strikingly similar subcortical brain
anatomy and function because they are not especi-
ally essential for the generation of consciousness. For
example, Rose states that ‘these [neocortical] struc-
tures and functional features are not present in
subcortical regions of the brain, which is probably
the main reason that activity confined to subcortical
brain systems can’t support consciousness’ (Rose
2002, p. 7). He also states that ‘…consciousness also
requires the operation of subcortical support systems
such as the brainstem reticular formation and the
thalamus, that enable a working condition of the
cortex. However, in the absence of cortical opera-
tions, activity limited to these subcortical systems
cannot generate consciousness’ (Rose 2002, pp. 6–
7). Although Rose cites data that reasonably support
his claims, we again find that other variations in
interpretation exist. Contemporary studies on the
neural correlates of consciousness does not seem to
support the suggestion that the thalamus behaves as
a ‘support system’, so that the neocortex is enabled
to generate consciousness. In fact, the role of
subcortical activity within the human thalamocor-
tical system is sometimes deemed as just as import-
ant for the ‘generation’ of consciousness per se,asis
neocortical function (Llina
´set al. 1998). In addition,
Rose (2002), cites other works (Tononi and Edelman
1998; Edelman and Tononi 2000; Laureys et al.
2000) and commentary that just as clearly suggests
that consciousness is more accurately described as a
global functioning state of the brain, rather than a
function of neocortical activity alone.
To further support his suggestion that the neo-
cortex is the exclusive domain of consciousness,
Rose extends a defence to include clinical studies
pertaining to human patients in chronic, vegetative
states. Rose describes a clinical condition that is
intended to demonstrate that damage to the human
neocortex renders a person vegetative and non-
conscious (Rose 2002; pp. 13–14, 21), thus the
neocortex must be responsible for any type of
conscious state. This is a reasonable interpretation
and hypothesis. However, there are three points of
information that illustrate a possible contradiction
in this logic. The first is that Rose fails to report that
all of his examples refer to a pathological condition
that radically affects the cerebral cortex in a way
that also compromises thalamocortical activity.
That is, thalamocortical activity and neocortical
function are confounded in Rose’s analysis. Sec-
ondly, Rose only reports cases of the persistent
vegetative condition in which damage has occurred
mostly to the neocortex, which permits an argu-
ment that the neocortex and not subcortical regions
are therefore responsible for consciousness. How-
ever, vegetative patients demonstrating near-nor-
mal cortical metabolic rates (i.e. preserved cerebral
function), but with damaged thalamic nuclei, have
been documented (Schiff et al. 2002). Our last point
refers to the effect of restricted neocortical lesions on
consciousness. Edelman and Tononi (2000, p. 54)
readily point out that ‘despite occasional claims to
the contrary, it has never been conclusively shown
that a lesion of a restricted portion of the cerebral
cortex leads to unconsciousness…no single area
seems to hold the key to consciousness’. In fact, the
only localized brain lesion that results in loss of
consciousness typically affects the reticular activa-
ting system, a non-cortical structure found in all
vertebrates. Again, it seems risky to equate con-
sciousness as the exclusive domain of the neocortex.
Rose constructs several concise arguments con-
trasting the neurobiology of humans with that of
other animals, and it therefore seems reasonable to
suggest that the human forebrain is both quantita-
tively and qualitatively different on most accounts
that matter to consciousness. These arguments
suggest a unique, causative link between the
physical size of the human neocortex, human
intelligence and the fundamental aspects of brain
organization that are supposedly specific to lamin-
ated mammalian brains. Rose emphasizes that it is
not just the presence of the neocortex that is critical
to consciousness, but a massive amount of neocor-
tical expansion is also required (Rose 2002, p. 7,
10). And it continues that this massive neocortical
expansion has allowed for the development of
certain anatomical and cognitive traits that are
distinctly human, including the lateralization of
functions between the cerebral hemispheres (Rose
2002, p. 13), or the ability to have a psychological
capacity (Rose 2002, p. 3). While there is no dispute
that humans possess greatly expanded mental
capacities that are associated with our brain struc-
ture, the application of this data to the analogous
question of mental capacity in fish and other
animals seems biased in our view. Consider the
assertion that a massive expansion of neocortex
must be present in order for consciousness to occur,
and that this neurological requirement is essential
for a psychological capacity (Rose 2002, p. 3, 29,
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 285
31). Edelman and Tononi (2000, p. 53) mention
several clinical reports of human patients, who have
lost (via surgery), or failed to develop, massive
neocortical expansion, and yet have reasonably
normal cognitive abilities and intelligence quotients.
These anomalous observations, considered together
with the reality that there is still little consensus
among neurobiologists as to how consciousness is
actually ‘generated’, point to the fact that modern
theories of the neural correlates of consciousness
are still just tentative explanations. But, we suppose
that this is fair enough at such an early stage of
debate of these complex paradigms. But Rose subtly
portrays some of these theories as biological facts,
without explaining the necessary caveats and
underlying assumptions inherent in these theories.
The observation is made that the majority of the
activity in our extensive brain matter is ‘unavaila-
ble’ to our conscious awareness, and therefore, for
an organism like a fish (i.e. having a smaller, less
complex brain), it is entirely logical that none of
their brain activity could be dedicated to conscious
experience (Rose 2002, p. 15). However, as pointed
out by Griffin (1998), the implied assumption
within Rose’s reasoning is that the proportion of
conscious to unconscious activity must be even
smaller in the non-human, animal brain. Rose fails
to provide any neurobiological data that could
justify that assumption (perhaps because such data
does not exist?), yet studies of human subjects
whose brain development or size had been limited
also do not support that premise (Edelman and
Tononi 2000). As Griffin (1998) remarks, ‘perhaps
only in ‘‘megabrains’’ is most of the information
processing unconscious…insofar as simple con-
scious thinking is effective and adaptive, it may be
one of the important functions a central nervous
system.’ Rose states that ‘expansion of the cerebral
hemispheres has also allowed lateralized functions
of the two cerebral hemispheres…’ and alludes to
the fact that certain lateralized cognitive functions
are manifestations of higher-order consciousness
(Rose 2002, p. 13). The idea that massive cortical
expansion is necessary for the lateralization of
cerebral functions seems simplistic. A brain is
considered to be cerebrally lateralized if one hemi-
sphere performs a different set of cognitive func-
tions, or is anatomically distinguishable from the
other (Bisazza et al. 1998). There are many studies
that show lateralization of cognitive functions
involved in social interactions, learning and per-
ceptual categorization occurs in many vertebrate
species, including fishes (Bisazza et al. 1998;
Vallortigara 2000), and that fundamental similar-
ities between the cerebral structures of all verte-
brates exist. Clearly, the association between
absolute brain size, brain organization, cognition
and consciousness is not as clear-cut as Rose
argues, and as such, his application of these
concepts to the question of consciousness in fish is
divisive.
Convergence, homology and evolutionary
psychology
Rose (2002) emphasizes the impressive differences
between the brain structure of fish and humans.
However, when the analysis of brain structure and
function is extended to all major vertebrate groups,
the vertebrate central nervous system appears to
have had a rather conservative evolution. The
structural or functional differences between species
can be accurately described as specialized adapta-
tions within a consistent overall organization
(Butler and Hodos 1996). We believe that Rose
would consider the neural substrate for ‘primary’ or
‘extended’ consciousness to be a specialized adapta-
tion exclusive to humans (Rose 2002, p. 6).
However, notably absent from Rose’s evolutionary
perspective is a consideration of the process of con-
vergent evolution. Convergent evolution appears at
all level of biological organization, and is a process
by which similarity between unrelated species
occurs because of adaptation to similar environ-
mental pressures (Wray 2002). Convergence can
occur on a functional level without the complete
convergence of underlying structural elements.
If Rose’s analysis is correct, and human neocor-
tical structure is the only neural substrate capable of
producing consciousness beyond a rudimentary
extent, then different neuroanatomical arrange-
ments of the forebrain should result in animals
with very dissimilar cognitive capabilities and little
or no manifestations of primary or extended con-
sciousness. Marino (2002) provides data and ana-
lysis that test this hypothesis and compares primate
and cetacean biology, and describes aspects of their
independent evolutionary history such as adapta-
tions to drastically different physical environments
(terrestrial vs. aquatic), as well as pronounced
differences in body shape and physiology. He also
demonstrates that cetacean forebrains are organized
in fundamentally different patterns from that
observed in the brains of primates, to the extent
Consciousness and pain perception in fish K P Chandroo et al.
286 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
that cetaceans can be considered as having a
completely different mode of cortical elaboration.
Interestingly, cetaceans demonstrate cognitive abil-
ities that are elsewhere only found in humans or the
great apes – abilities that were traditionally
assumed to be solely the result of human neocortical
structure and function. Marino (2002) clearly
shows that cognitive and behavioural convergence
can occur, even in the face of profound neuroan-
atomical divergence. The cognitive abilities shown
by cetaceans are also widely accepted as manifes-
tations of primary or extended consciousness
(Marino 2002). As cetaceans do not show the
‘extensive frontal and parietal lobe of neocortex’, or,
‘expansive, specialized six-layered type of cortex’
that Rose suggests is the most important brain level
requirement for conscious awareness and other
cognitive abilities (Rose 2002, p. 32), it would seem
that his argument can be challenged at the
conceptual level. Namely, the specific neurobiolog-
ical way that a species arrives at a functional
solution is not the only level by which to under-
stand it, especially when comparing disparate spe-
cies of animals. Thus, the neural substrate of
consciousness does not necessarily invoke the
involvement of a six-layered laminar structure, but
instead needs to fulfil some other aspect of a specific
neural process, such as those described within
Searle (2000) or Baars (2002). An examination of
brain structure and function between distantly
related species, which demonstrate cognitive adap-
tations characteristic of primary consciousness may
further reveal the nature of these neural processes.
We review the work of Marino (2002) as one
example of how different species have evolved
alternate mechanisms to increase their brain mass
and function. Enlargement and elaboration of the
forebrain has independently occurred multiple times
within different lineages of vertebrates, within fish
and mammalian species (Butler and Hodos 1996).
Contrary to the suggestions found in Rose (2002),
the forebrains of some fish do, in fact, represent
complex, elaborated structures within the verteb-
rate radiation. Anatomical features and functions of
the fish forebrain may be homologous or conver-
gently related to similar structures and functions in
mammals. Despite evidence to the contrary, Rose
strongly denies that any functional homologies
(especially limbic brain regions implicated in con-
scious states) exist in the fish brain (Rose 2002,
p. 28). His reasoning is that such homologies are
simply structural, and therefore it is misleading to
ascribe a comparable function with those struc-
tures. He also favours the thesis that functional
equivalency for any limbic structure found both in
fish and mammals is impossible, because fish do not
have a neocortex. We find the first reason perplex-
ing, because the majority of limbic structures in the
fish brain have not been defined simply because
their structures have been conserved during evolu-
tion, but specifically because they have similar
physiological and behavioural function as in other
vertebrates (Lo
´pez et al. 2000; Portavella et al.
2002). It is to be expected that limbic structures
will have modified interaction and function within
the fish and human cerebrum because those cereb-
ral structures must mediate different cognitive tasks.
However, the assertion that functional equivalency
of limbic structures is impossible because fish do not
have a neocortex implies that the influence of the
neocortex on other brain structures is to make those
brain structures functionally dissimilar. That con-
clusion seems unsupported by others (Butler and
Hodos 1996), and demonstrates an anatomical bias
in Rose’s reasoning. Many limbic brain structures
found in mammals have functionally similar coun-
terparts within the brains of fish (reviewed in
Chandroo et al. 2004).
Although Rose suggests that some mammals have
a primary, rudimentary form of consciousness
(because their cerebral hemispheres show limited
neocortical structure), he gives little indication as to
what primary consciousness actually implies in
terms of an animal’s psychological capacity. This is
significant, because delineating the exact nature of
the psychological differences between humans and
other species is necessary if we are to develop a valid
understanding of how and why a psychological
capacity arose during vertebrate evolution. In the
discussion that appears to address the question of
psychological capacities within vertebrate phyla, it
seems that Rose adopts an overly anthropocentric
view (Rose 2002, p. 3). That is, any psychological
capacity that can be observed in humans is assumed
to be uniquely human, so any suggestion that
animals have similar mental capacities can be
immediately dismissed as anthropomorphic. Rose
goes on to describe unique aspects of human
psychology, such as our capability for creativity
including art, science and the existence of religious
beliefs. He uses these specific human abilities as
examples that would make us seem so distinct, that it
would be highly inappropriate and misleading to
project any human-like psychological characteristic
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 287
whatsoever onto other species. And this may be a
reasonable premise. Yet, he then presents examples
of electroreception and signalling by electric fish, or
echolocation in bats and dolphins as unique capaci-
ties endowed to these animals that have no equal
counterpart in humans. This also sets the stage for
Rose’s opinion of consciousness (and therefore psy-
chological abilities as well), being likewise a unique
capacity, has no counterpart in most animals.
The opinions expressed by Rose reflect one of two
competing philosophical paradigms that character-
ize the approach to understanding the evolution of
the human mind. These approaches are termed the
Cartesian and Darwinian perspectives (Gibson
2002). Rose apparently ascribes to the Cartesian
perspective, which postulates that human and
animal minds are separated by major qualitative
differences in mental abilities. The Darwinian per-
spective postulates more continuity between animal
and human mental capacities; in other words, the
differences between the animal and human mind is
more a matter of degree rather than kind (Gibson
2002). As empirical data from comparative neuro-
biology and ethological studies have revealed lim-
itations in the explanatory value of the Cartesian
perspective for a number biological phenomenon,
including the learning behaviour of animals, we
tend to favour the later conceptual approach to
understanding the animal mind.
Animal behaviour research suggests that rudi-
ments of most human cognitive abilities also exist in
great apes (reviewed in Gibson 2002). These cog-
nitive abilities, many of which cannot be explained
by traditional associative theory, are thought to
occur through a process that is termed ‘mental
construction’. Mental construction refers to the
brain’s ability to generate a representation of
internal and external events. These mental repre-
sentations function as a predictive model of the
environment, allowing for the construction of new
knowledge. Utilization of this knowledge permits an
animal to express novel, adaptive behaviour (Topa
´l
and Csa
´nyi 1999; Edelman and Tononi 2000;
Gibson 2002). As cognitive abilities that are better
explained by means of mental construction have
been shown to exist in both animal and human
minds, it has been hypothesized that enhanced
human mental constructional capacities underlie
human creativity and mental flexibility (Gibson
2002). According to Gibson (2002), the improved
information processing abilities of the enlarged
human brain endows our species with greater
abilities to break concepts and actions into fine
component parts, and to combine these differenti-
ated components into higher order behavioural and
mental constructs. It is these mental construction
capacities that serve as a common foundation for
the wide-ranging behavioural domains in which
human intellectual abilities resemble and improve
on those of other primates (Gibson 2002). If we
assume that the capacity for mental construction is
associated with a psychological capacity, or is
characteristic of primary consciousness, then the
observation that mental construction occurs in a
variety of distantly related vertebrate taxa would
argue that basic psychological capacities, as well as
primary consciousness, is phylogenetically old. We
suggest that the ‘building blocks’ for psychological
capacity and primary consciousness may exist
within certain fundamental neural attributes and
processes shared by many vertebrate animals
(Chandroo et al. 2004). The hypothesis that the
ability for mental construction was associated with
the emergence of primary consciousness in animal
species, has also been proposed within tenets of
certain neural theories of consciousness (Edelman
and Tononi 2000). Rose (2002) contains no data
that enable us to determine whether or not mental
representation occurs in fish species, or that mental
constructs of a lesser complexity are not associated
with primary consciousness. The issue of mental
constructs in fish has been reviewed in Chandroo
et al. (2004), and we suggest that such a cognitive
ability is indeed feasible for fish species.
Pain perception and brain structure
Following the initial discussion on the neocortex
and consciousness, Rose (2002) gives a review of
pain perception in humans, with an eventual
application to the question of pain perception in
fish. Rose begins the pain perception review with an
analysis of nociception and pain in humans, appro-
priately making the distinction between transduc-
tion of tissue trauma into neural signals (i.e.
nociception), and central registration of nociception
(i.e. the processes involved in consciously experien-
cing pain). Rose states that ‘pain is a psychological
experience that is separate from behavioural reac-
tions to injurious stimuli’ (Rose 2002, p. 15) and we
accept this definition. However, for Rose, subcortical
events (e.g. behavioural reactions, and presumably
their accompanying autonomic nervous system
processes) involved with nociception have nothing
Consciousness and pain perception in fish K P Chandroo et al.
288 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
to do with the conscious awareness of pain,
because, ‘…the behavioural displays related to
noxious stimuli or emotion in humans, as in other
animals, are stereotyped, automatic behavioural
programs controlled by lower levels of the central
nervous system.’ (Rose 2002, p. 17) The supporting
evidence given for these suggestions is that one can
observe ‘pain-like’ responses in humans and ani-
mals with neocortical damage or impaired nocicep-
tive transmission preventing signals from reaching
the neocortex. Eventually, Rose’s argument reduces
to the premise that lack of a neocortex essentially
means that there can be no pain perception.
Conceptually, Rose’s view of the mechanisms of
pain perception is questionable. For example, con-
sider his reasoning behind the claim, ‘pain is a
psychological experience that is separate from
behavioural reactions to injurious stimuli.’ Contrary
to Rose’s assertions, the arousal of the subcortical
autonomic nervous system (and the associated
defensive behaviour response) plays a paramount
role in creating the psychological experience of pain,
and is not a functionally separate entity in normal
humans (Chapman and Nakamura 1999; Saper
2002; Craig 2003). Although it is perfectly accept-
able from an instructive point of view to subdivide
pain perception into anatomic categories (e.g. cor-
tical and subcortical), this dichotomous portrayal is
ultimately limited as far as gaining an accurate
understanding of human pain perception, as well as
in its application to the question of pain perception
in other species (Price et al. 2002). Although Rose’s
conception of the neural mechanisms of pain
perception can be demonstrated with humans or
animals with damaged, impaired central nervous
systems, one must ask – how does this relate to pain
perception in normal human subjects, and what, if
anything, does it say about the evolution of pain
perception in lower vertebrates?
Pain perception in humans is not simply a series of
reflexive behavioural responses accompanied by a
non-functional, distinctly separate, cortical-medi-
ated subjective feeling. Rose does not acknowledge
this concept, and as such, his arguments reflect a
dualistic perspective to pain perception. In addition to
his ‘consciousness as a computer monitor’ analogy in
which consciousness is portrayed as a passive win-
dow in which brain processes may come to aware-
ness (Rose 2002, p. 15), Rose cites a passage from
LeDoux (1996) that states, ‘the brain states and
bodily responses are the fundamental facts of an
emotion, and the conscious feelings are the frills that
have added icing to the emotional cake’ (Rose 2002,
p. 26). Rose seems to suggest that the conscious,
subjective aspects of pain perception have no tangible
or adaptive function, but are epiphenomenal. The
hypothesis that pain is a sensory end product of a
passive information transmission process has largely
been rejected among pain researchers (Chapman and
Nakamura 1999). It is curious how Rose can present
a dualistic perspective for a conscious state because
the work he cites in order to report that the neocortex
is the exclusive structure responsible for conscious-
ness explicitly requires the underlying assumption
that consciousness is not epiphenomenal, but func-
tional and adaptive (Edelman and Tononi 2000).
Although Rose argues that an evolutionary perspec-
tive is necessary for examining the question of pain
perception in fish species, his apparent dualistic view
of pain perception is largely incompatible with an
evolutionary account for its existence.
Although it is easy to gain the perception from
Rose (2002) that it is becoming very clear as to why
and how the neocortex is responsible for pain
perception, this portrayal does not fully reflect the
current scientific literature (Besson 1999; Chapman
and Nakamura 1999; Treede et al. 1999). In brain
imaging studies, only the anterior cingulate gyrus
(a ‘limbic’ system brain component) has demon-
strated a consistent response during the conscious
experience of pain in humans (Derbyshire et al.
1997). Interestingly, the anterior cingulate gyrus
does not show the classical neocortical, six-layered
structure, but its cellular conformation resembles a
five-layered structure, with other distinctions
(Nimchinsky et al. 1997). Although we could argue
that this neurophysiological fact casts doubt on
Rose’s claim that only specific, specialized neocortex
can be responsible for conscious perception of pain
in humans or animals, we instead suggest that this
line of reasoning in general is fruitless in any event.
Tononi and Edelman (1998) and others (Baars
2002) advocate that it is not any specific structure
(e.g. a cortex with five or six layers) or particular
location that is associated with the generation of
consciousness, but rather, it is the type of neuronal
activity per se that the structure participates in that
is critical for the generation of consciousness. While
Rose’s review brings forth some pertinent issues
needed to assess the question of pain perception in
fish, his interpretations of the neurobiological
underpinnings of human pain perception combined
with a lack of primary literature concerning fish
neurobiology and behaviour preclude him from
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 289
making any firm conclusions on the existence of
pain perception in those species.
We address the question of pain perception in fish
by first accepting the assumption that it is unlikely
that the conscious perception of pain evolved to
simply guide reactions to noxious events, or to
provide an experiential dimension to accompany
reflexes, but rather it allowed an organism to
discriminate their environment in ways that per-
mitted adaptive and flexible behaviour (Chandroo
et al. 2004). The neural systems involved in noci-
ception and pain perception, and the cognitive
processes resulting in flexible behaviour function,
probably evolved as an interactive dynamic system
within the central nervous system (Chapman and
Nakamura 1999). The creation of a subjective
feeling of pain is arguably a complex affair, inclu-
ding spinal, brainstem, thalamic and cerebral
structures, as well as essential autonomic nervous
system feedback (Willis and Westlund 1997; Chap-
man and Nakamura 1999; Saper 2002). Spinal,
thalamic and forebrain system interdevelopment
has been a primary mode of central nervous system
evolution in vertebrate species (Kevetter and Willis
1984; Butler 1994). Accordingly, it is reasonable to
suggest that the evolution of pain perception should
show concomitant developments in all those neural
structures, and these developments should be
reflected within an animal’s cognitive abilities.
Therefore, a pertinent question to ask when
attempting to determine the origins of pain percep-
tion during vertebrate evolution is when, and in
what form, does this unified, integrative system
appear? How do we recognize when nociceptive
signals arrive at the forebrain and participate in the
cognitive processes characteristic of conscious,
adaptive behaviour?
Work performed by Sneddon (2003) and
Sneddon et al. (2003) are a step in the right
direction to provide insights for the questions we
pose. Sneddon et al. (2003), Sneddon (2003) and
Ide and Hoffmann (2002) report experimental
studies that contribute significantly to the question
of nociception and pain perception in fish species.
Their studies contain data that describes the func-
tion of peripheral and central nervous system
structures involved in nociception, and the correla-
tion of nociceptor activation with autonomic
and behavioural responses. Sneddon (2003) and
Sneddon et al. (2003) interpret their work within an
ethological and physiological framework that allows
them to conclude that their results fulfil ‘the criteria’
for demonstrating pain in animals, and that accord-
ingly, fish can perceive pain. After defining pain in
humans as an ‘unpleasant sensory and emotional
experience associated with actual or potential tissue
damage,’ Sneddon et al. (2003) state that, ‘it is
impossible to truly know whether an animal has an
emotion because we cannot measure emotion
directly. Therefore, emotion does not feature in the
definition of pain in animals.’ They further describe
their criterion for pain perception by stating ‘…if a
noxious event has sufficient adverse effects on
behaviour and physiology in an animal, and this
experience is painful in humans, then it is likely to
be painful in the animal’ (Sneddon et al. 2003, p. 2).
The criterion that Sneddon et al. (2003) describe is
adequate for assessing clinical pain responses in
animals whose pain system is well understood
(e.g. domestic mammals). However, when this
criterion is used for the purpose of elucidating the
existence of pain perception in animals that are
significantly less understood (e.g. fish), subjective
leaps in the interpretation of the results are required
to come to firm conclusions on the issue.
Sneddon (2003) shows that injection of a known
noxious substance into peripheral tissue innervated
by nociceptors, causes several physiological and
behavioural reactions not found in control fish, such
as a significantly increased respiratory rate, a delay
in the time it takes for fish to resume feeding, and
rocking and rubbing behaviour. Sneddon (2003)
also shows that these induced responses greatly
diminish when morphine is administered intramus-
cularly. Clearly, the responses of fish to noxious
stimuli and morphine require the integration of
peripheral and central nervous system structures.
However, we must ask what relevance these results
have to the question of actual pain perception.
Sneddon (2003) first attempts to address that
question by suggesting that rocking behaviour of
fish may be similar to the rocking behaviour shown
by primates and zoo animals which are exhibiting
signs of poor welfare. Beyond the superficial simi-
larity, this suggestion is not further supported by
any other data presented. There is no effort made to
include or exclude alternative accounts for the
rocking behaviour that do not require the involve-
ment of consciousness for its explanation (e.g. the
effect of metabolic alterations caused by prolonged
hyperventilation, subsequent swim bladder or equi-
librium reactions, the effect of corticosteroid and
catecholamine responses, as well as the effect of
trigeminal nerve input on brainstem controlled
Consciousness and pain perception in fish K P Chandroo et al.
290 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
motor behaviour). In addition, Sneddon makes no
attempt to explain or compare the putative proxim-
ate and ultimate causes for rocking behaviour in
primates and fish. Such a comparison would need to
address the question of why rocking, primarily a
primate behaviour elicited by diverse aetiologies
(Nash et al. 1999), should have the same cognitive
underpinnings and social function in a rainbow
trout. This seems highly unlikely. Similarly,
Sneddon compares the rubbing behaviour of fish to
the act of rubbing an injured area to ameliorate the
intensity of pain as observed in humans and
mammals. She also states that rubbing behaviour
and rocking, may be potential pain-coping strategies
in fish (Sneddon 2003, p. 7). Again, the data or
analysis needed to confirm these suggestions are
absent, and as such, she cannot exclude non-
cognitive or non-conscious explanations for such
behaviour. Although Sneddon et al. (2003) have
described the composition and response character-
istics of fish nociceptors, they have not presented
any anatomical or physiological data to suggest that
afferent inhibition occurs in fish species (i.e. the
process behind pain amelioration via rubbing).
Sneddon suggests that the rocking and rubbing
behaviour of fish is complex in nature, and therefore
higher processing is involved (Sneddon 2003, p. 8).
This leads her to conclude that fish can perceive
pain. In light of the critique we have presented, we
suggest that this interpretation is premature. Based
on the information presented in Sneddon (2003),
labelling such behaviour as complex is a simple
value judgment. Even if nociception-induced rock-
ing and rubbing behaviour was complex, the
assumption that complex behaviour per se must
be indicative of ‘higher processing’ is incorrect
(Shettleworth 2001). Furthermore, the concept that
higher cognitive processing is synonymous with
conscious cognition is misleading (Bargh and
Ferguson 2000). We suggest that alternative cri-
teria should be used to determine if behavioural
responses are reflective of pain perception in fish.
The behaviour meeting this criterion should permit
the distinction between responses that involve the
integration of brain structures that are hypothesized
to be involved in the process of conscious cognition
and of those that do not. Such behaviour might
include those that are observed as the result of
interactive or declarative learning processes
(Chandroo et al. 2004).
Other studies that use autonomic and behavioural
responses of fish to assess whether they are
conscious have been performed by Cabanac
(1999). Based on empirical studies focused on
autonomic physiological responses, Cabanac
(1999), as well as Cabanac and Cabanac (2000)
suggest that consciousness and emotion evolved
with the appearance of amphibians or reptiles. They
reason that as fish do not show autonomic responses
to ‘emotional stress’ (i.e. tachycardia or behavioural
fever) in the same way terrestrial animals do, they
probably are not conscious. There are factual and
conceptual errors with this argument. Fish do in fact
exhibit tachycardia in response to circumstances
that one might expect a similar emotional response
to occur in mammals (Ho¨jesjo¨ et al. 1999). Caba-
nac’s hypothesis does not account for the observa-
tion that emotionally influenced autonomic
responses are associated with behaviour that is
relevant to surviving species-specific environmental
challenges. For example, an animal that evolved a
‘freeze’ strategy to a predation threat likely have
altered autonomic responses to an animal that
evolved a ‘flight response’. The autonomic response
to specific stimuli can even change with age within
single species (Ho¨jesjo¨ et al. 1999). The expectation
that certain autonomic or behavioural responses
should be similarly associated with specific emotions
among diverse vertebrate groups is unwarranted at
this time, although it is a useful paradigm to
examine. Future work should aim to account for
the mechanism by which behavioural or autonomic
responses are seated within cerebral processes that
may be associated with the existence of conscious
states.
Cognition and behaviour in fishes
Rose has reviewed a subset of the extensive
contemporary knowledge available on the neurobe-
havioural nature of fish. The primary literature
concerning the neurobiological features and learn-
ing behaviour of fish and other non-human verte-
brates is somewhat weakly presented in Rose
(2002), and it undoubtedly is the reason for many
of the arguments that support his conclusions. He
argues that ‘…most behaviour of fishes is not
dependent greatly on learning…’ (Rose 2002,
p. 8), ‘…instrumental and Pavlovian conditioning
are forms of associative, implicit learning that occur
in fishes…as cases of implicit learning, they operate
without awareness…’ (Rose 2002, p. 27), ‘…noth-
ing about the behaviour of a fish requires a capacity
for conscious awareness for its explanation’ (Rose
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 291
2002, p. 24), ‘…avoidance conditioning occurs
unconsciously and is not evidence of awareness of
pain or any other experience’ (Rose 2002, p. 29),
‘…in fishes, our brainstem-spinal systems are
adequate for generation of overt reactions’ (Rose
2002, p. 25) and ‘…behavioural specialization in
fishes is associated with expanded brainstem as
opposed to cerebral hemisphere development’ (Rose
2002, p. 10). All of those statements are misleading
in our opinion.
Rose claims that fish behaviour in general is not
greatly dependent on learning, except for the initial
development of species-typical behaviour (Rose
2002, p. 9). The idea that fish behaviour is
dominated by pre-programmed, invariant responses
to the environment, and that the main significance
of learning is to ‘prime’ the development of such
responses is perhaps more reflective of a historical
view of fish behaviour rather than of any current
data (Laland et al. 2003). The relevance of learning
to fish behaviour at various life-history stages has
been increasingly investigated over the past decade
(Overmier and Hollis 1990; Kieffer and Colgan
1992; Csa
´nyi and Do
´ka 1993; Bshary et al. 2002).
The learning processes shown by fish include
observational (McGregor et al. 2001), interactive
(Topa
´l and Csa
´nyi 1999), Pavlovian (Hollis 1984)
and avoidance (Zerbolio and Royalty 1983). Cur-
rent work shows that learning processes demon-
strated by fish are multifaceted phenomenon that
have clear fitness implications to fish species at
various developmental stages (Brown and Laland
2003; Griffiths 2003; Hoare and Krause 2003;
Kelley and Magurran 2003; Odling-Smee and
Braithwaite 2003).
Rose presents Pavlovian (i.e. classical condition-
ing) and avoidance learning exclusively to describe
learning processes in fish. He further defines Pavlo-
vian and avoidance learning as examples of implicit
learning. The argument then goes on to state that
implicit learning has no relationship to any con-
scious process, and must therefore occur without
conscious awareness. The suggestion that implicit
learning has no relationship to higher order cogni-
tion, or is inherently an unconscious process is a
concept that is not universally accepted because
there is empirical data that show otherwise (Maren
2001; Lovibond and Shanks 2002). Similarly, the
suggestion that avoidance learning in fish species is
purely a form of implicit learning with no other
significance to cognition is incorrect according to
Overmier and Hollis (1990). We suggest that
observational, avoidance and interactive learning
processes may require the formation of declarative
memories. The relationship between learning pro-
cesses demonstrated by fish, declarative memory
and conscious cognition has been reviewed in
Chandroo et al. (2004), and there is, in fact, an
objective basis for suggesting that some fish beha-
viour is better explained within a theoretical
framework that includes primary consciousness as
a function of their nervous system. Given the fact
that none of the learning processes we mention are
considered in Rose (2002), his statement that
‘nothing about a fish’s behaviour could be con-
scious’ seems unqualified and incomplete.
There are several arguments found within Rose
(2002) that suggest that fish are unique among
vertebrates in that their behavioural specialization
is dependent heavily upon brainstem development,
and that the fish cerebrum has little significance to
their behaviour beyond olfaction, or to ‘refine the
expression’ of brainstem functions (Rose 2002,
p. 9). He further states that, ‘the neurobehavioral
evolution of fishes has resulted in a highly diversi-
fied array of species in which the essentials of
neurobehavioral function are mediated mainly by a
neural system below the cerebral hemispheres’
(Rose 2002, p. 9). To support these claims, he
suggests that fish function, learn and behave
essentially normally (except for functions requiring
olfaction, which Rose claims is processed entirely
within their forebrain) after their cerebrum is
ablated. The logic here is misleading. The concept
that the fish cerebrum functions primarily as a
‘smell brain’ has been rejected by most comparative
neurologists for some time (Echteler and Saidel
1981). Although Rose brings forth the valid notion
that one function of the fish cerebrum is to
modulate behavioural expression, he fails to
acknowledge the implications of this function in
terms of fish cognition. Ablation of the fish cere-
brum does in fact impair learning and behaviour
that are hypothesized to involve expectancies,
complex spatial cognition, declarative memory or
mental construction processes (Overmier and Hollis
1990; Broglio et al. 2003; Chandroo et al. 2004).
Rose chooses to report only the learning and
behavioural processes that do not require the
involvement of such cognitive processes for their
explanation, and distorts the definition of avoidance
learning to suit his arguments. Rose’s theoretical
argument that evolution of the central nervous
system of fishes has resulted in mainly brainstem
Consciousness and pain perception in fish K P Chandroo et al.
292 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
development, and that their success as a species has
little to do with cerebral modification is refuted by
empirical evidence. All teleost fish have elaborate
forebrains (Butler and Hodos 1996), and the degree
of forebrain development is correlated with social
behaviour, communication abilities and other envi-
ronmental factors that may require integrative
cognitive capacities (Kotrschal et al. 1998). Fish
have evolved to exploit diverse environmental
niches, and show concomitant development within
all relevant brain areas. This brain development
may consist of increases or decreases in brain stem,
cerebellar, optic, olfactory, diencephalic and telen-
cephalic structural mass or complexity (Kotrschal
et al. 1998). The morphological changes that may
comprise the representative brain for any given
species are diverse, and contrary to Rose’s general-
ized assumption, some phylogenetic radiations
(e.g. actinoperygians) show a shift in brain mass
from primary sensory areas towards higher order
integration centres (Kotrschal et al. 1998). In addi-
tion, the integration of cerebellar, optic and telen-
cephalic functions to produce cognitive responses to
the environment may be similar in fish and mam-
mals (Broglio et al. 2003). Although Rose briefly
comments on the great morphological variation of
fish brains and the putative functions of the fish
cerebrum (Rose 2002, p. 9, 10), he fails to give a
balanced account of the implication that such
diversity and development have to cognitive func-
tions, and the question of consciousness in fish
species.
Conclusions
If the debate regarding the existence of sentience in
fish is to have valid conclusions, the basis of the
arguments must be made upon sound biological
principles, taking into account all sources of rele-
vant data. Our critique has demonstrated that the
input of recent behavioural, neurological or phy-
siological findings into the analysis can profoundly
change the possible conclusions reached about the
mental capacity of fishes. We have argued that if
one adopts a Darwinian perspective to the study of
animal minds, it is not simply a matter of more, less,
or no neocortex present that permits the existence of
a neural system that may support primary con-
sciousness. A current limitation of theories descri-
bing the neural basis of consciousness in humans
is that it essentially examines the intrinsic proper-
ties of complex neural systems, often without
considering the question of how those characteris-
tics arose during evolution (Tononi et al. 1994). If a
neural process, whether its substrate is laminated or
otherwise, allows some degree of mental construc-
tion to occur, and then it may be reasonable to
suggest that those animals may have evolved
primary consciousness. Autonomic and behavioural
responses that are used to prove or refute the
existence of conscious states in fish species need to
be assessed for their involvement within integrative
cognitive processes that are associated with mental
construction, declarative memory or other possible
indicators of primary consciousness. A sound
assessment of the probability that conscious
states occur in fish species will require expanded
knowledge of their forebrain neuroanatomy, an
understanding of how such structures mediate
behavioural responses to environmental challenges
and an analysis of that information within the
context of contemporary theory on the evolution of
consciousness.
Acknowledgements
We wish to thank the Ontario Ministry of Agricul-
ture and Food for partial, financial support of this
project. We also thank Dr Ian Duncan for his input
to many of the issues discussed.
References
Baars, B.J. (2002) The conscious access hypothesis: origins
and recent evidence. Trends in Cognitive Sciences 6, 47–52.
Bargh, J.A. and Ferguson, M.J. (2000) Beyond behavio-
rism: on the automaticity of higher mental processes.
Psychological Bulletin 126, 925–945.
Besson, J.M. (1999) The neurobiology of pain. The Lancet
353, 1610–1615.
Bisazza, A., Rogers, L.J. and Vallortigara, G. (1998) The
origins of cerebral asymmetry: a review of evidence of
behavioural and brain lateralization in fishes, reptiles
and amphibians. Neuroscience and Biobehavioral Reviews
22, 411–426.
Block, N. (2001) Paradox and cross purposes in recent
work on consciousness. Cognition 79, 197–219.
Broglio, C., Rodriguez, F. and Salas, C. (2003) Spatial
cognition and its neural basis in teleost fishes. Fish and
Fisheries 4, 247–255.
Brown, C. and Laland, K.N. (2003) Social learning in
fishes: a review. Fish and Fisheries 4, 280–288.
Bshary, R., Wickler, W. and Fricke, H. (2002) Fish
cognition: a primate’s eye view. Animal Cognition 5,
1–13.
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 293
Butler, A.B. (1994) The evolution of the dorsal pallium in
the telencephalon of amniotes: cladistic analysis and a
new hypothesis. Brain Research Reviews 19, 66–101.
Butler, A.B. and Hodos, W. (1996) Comparative Vertebrate
Neuroanatomy. Wiley-Liss, Inc., New York, USA.
Cabanac, M. (1999) Emotion and Phylogeny. Japanese
Journal of Physiology 49, 1–10.
Cabanac, A. and Cabanac, M. (2000) Heart rate response
to gentle handling of frog and lizard. Behavioural
Processes 52, 89–95.
Chandroo, K.P., Duncan, I.J.H. and Moccia, R.D. (2004)
Can Fish Suffer? Perspectives on sentience, pain, fear
and stress. Applied Animal Behaviour Science 86, 225–
250.
Chapman, C.R. and Nakamura, Y. (1999) A passion of the
soul: an introduction to pain for consciousness research-
ers. Consciousness and Cognition 8, 391–422.
Craig, A.D. (2003) Pain mechanisms: labeled lines versus
convergence in central processing. Annual Review of
Neuroscience 26, 1–30.
Csa
´nyi, V. and Do
´ka, A. (1993) Learning interactions
between prey and predator fish. Marine Behaviour and
Physiology 23, 63–78.
Damasio, A.R. (1998) Investigating the biology of con-
sciousness. Philosophical Transactions of the Royal Society
of London: Series B 353, 1879–1882.
Dennett, D. (2001) Are we explaining consciousness yet?
Cognition 79, 221–237.
Derbyshire, S.W.G., Jones, A.K.P., Gyulai, F., Clark, S.,
Townsend, D. and Firestone, L.L. (1997) Pain processing
during three levels of noxious stimulation produces
differential patterns of central activity. Pain 73, 431–
445.
Echteler, S.M. and Saidel, W.M. (1981) Forebrain connec-
tions in the goldfish support telencephalic homologies
with land vertebrates. Science 212, 683–684.
Edelman, G.M. and Tononi, G. (2000) A Universe of
Consciousness. Basic Books, New York, NY.
Gibson, K.R. (2002) Evolution of human intelligence: the
roles of brain size and mental construction. Brain,
Behavior and Evolution 59, 10–20.
Griffin, D.R. (1998) From cognition to consciousness.
Animal Cognition 1, 3–16.
Griffiths, S.W. (2003) Learned recognition of conspecifics
by fishes. Fish and Fisheries 4, 256–268.
Hoare, D.J. and Krause, J. (2003) Social organization,
shoal structure and information transfer. Fish and
Fisheries 4, 269–279.
Ho¨ jesjo¨ , J., Johnsson, J.I. and Axelsson, M. (1999) Beha-
vioural and heart rate responses to food limitation and
predation risk: an experimental study on rainbow trout.
Journal of Fish Biology 55, 1009–1019.
Hollis, K.L. (1984) The biological function of Pavlovian
conditioning: the best defense is a good offence. Journal of
Experimental Psychology, Animal Behavior Processes 10,
413–425.
Ide, L.M. and Hoffmann, A. (2002) Stressful and behavi-
oral conditions that affect reversible cardiac arrest in the
Nile tilapia, Oreochromis niloticus (Teleostei). Physiology
and Behavior 75, 119–126.
Jack, A.I. and Shallice, T. (2001) Introspective physicalism
as an approach to the science of consciousness. Cognition
79, 161–196.
John, E.R. (2001) A field theory of consciousness. Con-
sciousness and Cognition 10, 184–213.
Kanwisher, N. (2001) Neural events and perceptual
awareness. Cognition 79, 89–113.
Kelley, J.L. and Magurran, A.E. (2003) Learned predator
recognition and antipredator responses in fishes. Fish
and Fisheries 4, 216–226.
Kevetter, A.G. and Willis, W.D. (1984) Collateralization in
the spinothalamic tract: new methodology to support or
deny phylogenetic theories. Brain Research Reviews 7,
1–14.
Kieffer, J.D. and Colgan, P.W. (1992) The role of learning
in fish behaviour. Reviews in Fish Biology and Fisheries 2,
125–143.
Kotrschal, K., van Staaden, M.J. and Huber, R. (1998) Fish
brains: evolution and environmental relationships.
Reviews in Fish Biology and Fisheries 8, 373–408.
Laland, K.N., Brown, C. and Krause, J. (2003) Learning in
fishes: from three-second memory to culture. Fish and
Fisheries 4, 199–202.
Laureys, S., Faymonville, M.E., Luxen, A., Lamy, M.,
Franck, G. and Maquet, P. (2000) Restoration of
thalamocortical connectivity after recovery from persist-
ent vegetative state. The Lancet 355, 1790–1791.
LeDoux, J. (1996) The Emotional Brain. Simon and Schus-
ter, New York.
Lindahl, B.I.B. (1997) Consciousness and biological evo-
lution. Journal of Theoretical Biology 187, 613–629.
Llina
´s, R., Ribary, U., Contreras, D. and Pedroarena, C.
(1998) The neuronal basis for consciousness. Philosoph-
ical Transactions of the Royal Society of London: Series B
353, 1841–1849.
Lo
´pez, J.C., Broglio, C., Rodrı´guez, F., Thinus-Blanc, C.
and Salas, C. (2000) Reversal learning deficit in a
spatial task but not in a cued one after telencephalic
ablation in goldfish. Behavioural Brain Research 109,
91–98.
Lovibond, P.F. and Shanks, D.R. (2002) The role of
awareness in Pavlovian conditioning: empirical evi-
dence and theoretical implication. Journal of Experimental
Psychology 28, 3–26.
Maren, S. (2001) Neurobiology of Pavlovian fear condi-
tioning. Annual Review of Neuroscience 24, 897–931.
Marino, L. (2002) Convergence of complex cognitive
abilities in cetaceans and primates. Brain, Behavior and
Evolution 59, 21–32.
McGregor, P.K., Peake, T.M. and Lampe, H.M. (2001)
Fighting fish Betta splendens extract relative information
from apparent interactions: what happens when what
Consciousness and pain perception in fish K P Chandroo et al.
294 2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295
you see is not what you get. Animal Behaviour 62,
1059–1065.
Moccia, R.D.M. and Chandroo, K.P. (2003) Aspects of
animal welfare and aquaculture - a compendium of selec-
ted literature. Aquaculture Centre, University of
Guelph, Guelph, Canada (http://www.aps.uoguelph.ca/
aquacentre/aec/publications/; last accessed, 24 Sept
2004).
Nash, L.T., Fritz, J., Alford, P.A. and Brent, L. (1999)
Variables influencing the origins of diverse abnormal
behaviors in a large sample of captive chimpanzees (Pan
troglodytes). American Journal of Primatology 48, 15–29.
Nieuwenhuys, R. (1994) The neocortex. Anatomy and
Embryology 190, 307–337.
Nimchinsky, E.A., Vogt, B.A., Morrison, J.H. and Hof, P.R.
(1997) Neurofilament and calcium-binding proteins in
the human cingulate cortex. The Journal of Comparative
Neurology 384, 597–620.
Odling-Smee, L. and Braithwaite, V.A. (2003) The role of
learning in fish orientation. Fish and Fisheries 4, 235–246.
Overmier, J.B. and Hollis, K.L. (1990) Fish in the think
tank: learning, memory, and integrated behavior. In:
Neurobiology of Comparative Cognition (eds R.P. Kesner
and D.S. Olson), Lawrence Erlbaum, Hillsdale, NJ,
pp. 205–236.
Parvizi, J. and Damasio, A. (2001) Consciousness and the
brainstem. Cognition 79, 135–159.
Portavella, M., Vargas, J.P., Torres, B. and Salas, C. (2002)
The effects of telencephalic pallial lesions on spatial,
temporal, and emotional learning in goldfish. Brain
Research Bulletin 57, 397–399.
Price, D.D., Barrell, J.J. and Rainville, P. (2002) Integrating
experiential-phenomenological methods and neuro-
science to study neural mechanisms of pain and con-
sciousness. Consciousness and Cognition 11, 593–608.
Rolls, E.T. (2000) Pre
´cis of the brain and emotion.
Behavioral and Brain Sciences 23, 177–234.
Rose, J.D. (2002) The neurobehavioral nature of fishes and
the question of awareness and pain. Reviews in Fisheries
Science 10, 1–38.
Saper, C.B. (2002) The central autonomic nervous system:
conscious visceral perception and autonomic pattern
generation. Annual Review of Neuroscience 25, 433–469.
Schiff, N.D., Ribary, U., Moreno, D.R. et al. (2002)
Residual cerebral activity and behavioural fragments
can remain in the persistently vegetative brain. Brain
125, 1210–1234.
Schilhab, T.S.S. (2002) Anthropomorphism and mental
state attribution. Animal Behaviour 63, 1021–1026.
Searle, J.R. (1998) How to study consciousness scientific-
ally. Philosophical Transactions of the Royal Society of
London: Series B 353, 1935–1942.
Searle, J.R. (2000) Consciousness. Annual Review of Neu-
roscience 23, 557–578.
Sewards, T.V. and Sewards, M.A. (2000) Visual awareness
due to neuronal activities in subcortical structures: a
proposal. Consciousness and Cognition 9, 86–116.
Sewards, T.V. and Sewards, M.A. (2001) On the correla-
tion between synchronized oscillatory activities and
consciousness. Consciousness and Cognition 10, 485–
495.
Shettleworth, S.J. (2001) Animal cognition and animal
behaviour. Animal Behaviour 61, 277–286.
Sneddon, L.U. (2003) The evidence for pain in fish: the use
of morphine as an analgesic. Applied Animal Behaviour
Science 83, 153–162.
Sneddon, L.U., Braithwaite, V.A. and Gentle, M.J. (2003)
Do fishes have nociceptors? Evidence for the evolution of
a vertebrate sensory system. Proceedings of the Royal
Society of London, Series B 270, 1115–1121
Sporns, O., Tononi, G. and Edelman, G. (1991) Modeling
perceptual grouping and figures-ground segregation by
means of active reentrant connections. Proceedings of the
National Academy of Sciences of the USA 88, 129–133.
Striedter, G.F. (2002) Brain homology and function: an
uneasy alliance. Brain Research Bulletin 57, 239–242.
Tononi, G. and Edelman, G. (1998) Consciousness and
complexity. Science 282, 1846–1851.
Tononi, G., Sporns, O. and Edelman, G. (1994) A measure
for brain complexity: relating functional segregation and
integration in the nervous system. Proceedings of the
National Academy of Sciences of the USA 91, 5033–5037.
Tononi, G., Sporns, O. and Edelman, G.M. (1999) Meas-
ures of degeneracy and redundancy in biological
networks. Proceedings of the National Academy of Sciences
of the USA 96, 3257–3262.
Topa
´l, J. and Csa
´nyi, V. (1999) Interactive learning in the
paradise fish (Macropodus opercularis): an ethological
interpretation of the second-order conditioning para-
digm. Animal Cognition 2, 197–206.
Treede, R.-D., Kenshalo, D.R., Gracely, R.H. and Jones,
A.K.P. (1999) The cortical representation of pain. Pain
79, 105–111.
Vallortigara, G. (2000) Comparative neuropsychology of
the dual brain: a stroll through animals’ left and right
perceptual worlds. Brain and Language 73, 189–219.
Willis, W.D. and Westlund, K.N. (1997) Neuroanatomy of
the pain system and of the pathways that modulate
pain. Journal of Clinical Neurophysiology 14, 2–31.
Wray, G.A. (2002) Do convergent developmental mecha-
nisms underlie convergent phenotypes. Brain, Behavior
and Evolution 59, 327–336.
Zeman, A. (2001) Consciousness. Brain 124, 1263–1289.
Zerbolio, D.J. and Royalty, J.L. (1983) Matching and oddity
conditional discrimination in the goldfish as avoidance
responses: evidence for conceptual avoidance learning.
Animal Learning and Behavior 11, 341–348.
Consciousness and pain perception in fish K P Chandroo et al.
2004 Blackwell Publishing Ltd, F I S H and F I S H E R I E S , 5, 281–295 295
