Klein, Colin and Barron, Andrew B. (2016) Insects have the capacity for subjective
experience.
Animal Sentience
9(1)
DOI: 10.51291/2377-7478.1113
This article has appeared in the journal
Animal Sentience
,
a peer-reviewed journal on animal cognition and feeling. It
has been made open access, free for all, by WellBeing
International and deposited in the WBI Studies
Repository. For more information, please contact
wbisr-info@wellbeingintl.org.
Insects have the capacity for subjective experience
Colin Klein
1*
& Andrew B. Barron
2*
1
Department of Philosophy, Macquarie University
2
Department of Biological Sciences, Macquarie University
*
equal authorship contribution
Abstract: To what degree are non-human animals conscious? We propose that the most
meaningful way to approach this question is from the perspective of functional
neurobiology. Here we focus on subjective experience, which is a basic awareness of the
world without further reflection on that awareness. This is considered the most basic form
of consciousness. Tellingly, this capacity is supported by the integrated midbrain and basal
ganglia structures, which are among the oldest and most highly conserved brain systems
in vertebrates. A reasonable inference is that the capacity for subjective experience is both
widespread and evolutionarily old within the vertebrate lineage. We argue that the insect
brain supports functions analogous to those of the vertebrate midbrain and hence that
insects may also have a capacity for subjective experience. We discuss the features of
neural systems which can and cannot be expected to support this capacity as well as the
relationship between our arguments based on neurobiological mechanism and our
approach to the “hard problem” of conscious experience.
Keywords: subjective experience, primary consciousness, vertebrate midbrain, superior
colliculus, invertebrate, insect
Colin Klein is Senior Lecturer in the Department of Philosophy
at Macquarie University. He works on philosophy of
neuroscience with a side interest in the perception of pain and
other homeostatically relevant states. In 2014 he received an
ARC Future Fellowship to look at interventionist approaches to
cognitive neuroscience.
http://www.colinklein.org
Andrew B. Barron is Associate Professor in the Department
of Biological Sciences at Macquarie University. With his team
at Macquarie, he is exploring the neurobiology of major
behavioural systems such as memory, goal-directed
behaviour and stress from a comparative and evolutionary
perspective. In 2015 he was awarded an ARC Future
Fellowship to develop a computational model of the honey
bee brain.
http://bio.mq.edu.au/research/groups/cognitive-
neuroethology/dr-andrew-barron/
Call for Commentary: Animal Sentience publishes Open Peer Commentary on all accepted target
articles. Target articles are peer-reviewed. Commentaries are editorially reviewed. There are
submitted commentaries as well as invited commentaries. Commentaries appear as soon as they
have been revised and accepted. Target article authors may respond to their commentaries
individually or in a joint response to multiple commentaries.
Instructions:
http://animalstudiesrepository.org/animsent/guidelines.html
Animal Sentience 2016.100: Klein & Barron on Insect Experience
2
1. Introduction
What follows is a synopsis of our argument in Barron & Klein (2016). Our intention here
is both to summarize our arguments from comparative functional neurobiology that
insects have subjective experience as well as to expand upon and clarify some points from
our previous article. Here we provide some further discussion of why we believe the
insect brain is capable of subjective experience and of the features of nervous system
organization which do and do not have this capacity. We conclude with reflections on the
relationship between our structural arguments and the so-called “hard problem of
consciousness” (Chalmers, 1996).
2. Consciousness and Subjective Experience
Consciousness is a complex, multifaceted phenomenon (Bayne, Hohwy, & Owen, 2016).
Terminology to describe consciousness has proliferated more quickly than our
understanding of the phenomenon. Most authors, however, mark off a very basic sense of
“conscious” that refers to the basic capacity to have subjective experience (Morin, 2006).
In Nagel’s (1974) familiar term of art, there is “something it is like” to be an organism with
subjective experience. Organisms capable of subjective experience do more than merely
react: they have a perspective on the world with a unique phenomenological feel.
We distinguish this minimal level of consciousness from more demanding conscious
relations. We think it is possible to have subjective experience without higher-order
thoughts (Edelman, 2003; Rosenthal, 2005), self-awareness of oneself as a subject
(Christoff, Cosmelli, Legrand, & Thompson, 2011; Morin, 2006), or reportable access to
one’s own phenomenal states (Block, 1995). We think, in short, that it is possible to simply
be aware, with no further reflection.
Such a distinction is, of course, philosophically contentious. Some believe that subjective
experience requires heavier capacities for self-reflection. We justify adopting this
distinction in three ways. First, we think that this is the modal position among
philosophers and consciousness scientists. Second, adopting such a distinction corrects
for potential anthropocentric bias. Third, the distinction alone does not secure our
conclusion. We argue that insects have the capacity for subjective experience. Even those
who think that sentience without self-reflection is possible are wary of including insects
on our side of the line. Hence work remains to be done.
In humans, the capacity for subjective experience is dissociable from the capacity for self-
reflexive consciousness. While the latter is dependent on cortical and midbrain structures
(Damasio, 1999), several authors have argued that the former is supported by the
midbrain and subcortical structures (Damasio & Carvalho, 2013; Mashour & Alkire, 2013;
Merker, 2005, 2007; B. Merker, 2013; Parvizi & Damasio, 2001; Penfield & Jasper, 1954).
We rely especially on the work of Bjorn Merker (2007), who draws on evidence from
anesthesia, vegetative state research, developmental disorders, brain damage and lesion
studies to create a compelling argument that the integrated structures of the vertebrate
midbrain are sufficient to support the capacity for subjective experience in humans.
Animal Sentience 2016.100: Klein & Barron on Insect Experience
3
Yet while cortical damage can profoundly affect the content of conscious experience, it
seems that there is no part of the cortex upon which the capacity for consciousness
reliably depends (Damasio, Damasio, & Tranel, 2012; Damasio & Van Hoesen, 1983;
Friedman-Hill, Robertson, & Treisman, 1995; Herbet et al., 2014; Kapur et al., 1994;
Merker, 2007; Penfield & Jasper, 1954; Philippi et al., 2012). By contrast, the basic capacity
for subjective experience is sensitive to damage to midbrain structures (Merker, 2007).
The primary locus of action of many global anesthetics is subcortical (Alkire, Hudetz, &
Tononi, 2008; Gili et al., 2013). Emergence from anesthesia (Långsjö et al., 2012; Mashour
& Alkire, 2013), and coma or vegetative state (Schiff, 2010) are similarly predicted by the
reengagement of subcortical structures.
Note here the important distinction between the capacity for subjective experience and
the particular contents of experience at a given time. The human cortex obviously makes
a considerable contribution to what we are aware of. Cortical damage may appear to
remove whole categories of conscious content, but determining the actual effects of such
damage requires careful investigation, given the complexity of inhibitory interactions
with sub-cortical regions (Sprague, 1966). Similarly, there ought to be considerable
variation in conscious content across phyla. Yet these are all variations which require the
capacity for subjective experience in the first place.
The evidence is thus that the basic capacity for subjective experience is supported by
subcortical structures. Why might this be the case? We adopt a proposal put forward by
Merker (2007), who offers a functional proposal for the midbrain and subcortical basal
ganglia structures that explains their role in subjective experience. These structures
combine processed sensory information on the state and structure of the environment
with processed information on the homeostatic needs of the organism. The outcome is a
unified multimodal neural model of the agent within its environment, which is weighted
by the current needs and state of the agent. Within the midbrain, different structures
perform different roles in this information economy (Figure 1). This modeling gives the
organism a unique, unified perspective on the world. This, argues Merker, is what makes
subjective experience possible.
Two features of this proposal are particularly relevant for our argument. First, the
integrated processing of spatial information in the midbrain enables a mobile animal with
spatial senses to solve the so-called re-afference problem (von Holst & Mittelstaedt, 1950).
A moving animal must disambiguate environmental movement from the sensory input
caused by its own motion relative to the environment. For active animals with well-
developed spatial senses, it is computationally more effective to resolve the re-afference
problem once in a unified sensory model than to resolve it in a dispersed and peripheral
way for each sense independently. In addition, different senses contribute different
information on how the body is moving; thus re-afference can be resolved with greater
accuracy and precision by integrating information from multiple senses (Merker, 2005).
In vertebrates, the layered tectum (or superior colliculus (SC) in mammals) of the roof of
the midbrain receives processed and topographically organized input from all spatially
structured senses, including vision, auditory, and somatosensory inputs (Damasio &
Carvalho, 2013; Harting, Updyke, & Van Lieshout, 1992; Klier, Wang, & Crawford, 2001;
McHaffie, Stanford, Stein, Coizet, & Redgrave, 2005; Merker, 2007). In mammals, inputs to
the SC include inputs from the vestibular system (Frens, Suzuki, Scherberger, Hepp, &
Animal Sentience 2016.100: Klein & Barron on Insect Experience
4
Henn, 1998), information on eye position (Groh & Sparks, 1996; Knox & Donaldson, 1995;
Van Opstal, Hepp, Suzuki, & Henn, 1995), and somatosensation (Merker, 2007). This
allows the influence of self-motion on the sensory fields to be factored out of the
constructed sensory model of the environment (Sparks, 1988). Hence the SC is vital for
organizing motion in space, for directed attention, and for reaching and grasping for
targets (Horowitz & Newsome, 1999; Krauzlis, Liston, & Carello, 2004; McPeek & Keller,
2004; Zenon & Krauzlis, 2012).
Figure 1: The vertebrate behavioral core control system. Following Merker
(2007), autonomous animal decision making can be considered to involve three
related domains: motivation, target selection, and action selection (A). These
domains can be resolved and decisions can be made by an integrated neural system
that contains information on the state of self, self-movement, environmental state,
and structure and memory of prior experience. These capacities are supported by
different midbrain structures (B – shown here not to scale). As a simplification,
regions are colored according to their primary function(s) described in A. The
superior colliculus (part of the tectum (TEC) forming the roof of the midbrain)
processes multisensory spatial information (Merker 2007). Hypothalamic
structures (Hyp) and associated nuclei, the pituitary (pt) and mammillary bodies
(M) collate information on the physiological status of the organism referenced with
prior experience, to identify needs to maintain a homeostatic optimum (Damasio
& Carvalho, 2013; Swanson, 2000). Integrative structures within the midbrain and
basal ganglia, including the periaqueductal grey (P), substantia nigra (N), thalamus
(Tha), striatum (St) and midbrain reticular formation (MR), integrate these
sources of information with forms of memory to update relevance to the organism
according to prior experience (McHaffie et al., 2005; Merker, 2007).
The mammalian SC thus acts as a point of convergence for spatially structured sensory
information, including information about the position, orientation, and movement of the
body (Masino, 1992; May, 2006; Merker, 2005; Sparks, 1988; Zenon & Krauzlis, 2012).
Processing within the SC creates a neural model of the mobile animal in space, which is
essential for resolving decisions about how to react to resources around the animal.
The second relevant feature of the midbrain is that information integration within it
allows for efficient action selection in complex environments. Merker (2007) has
described the functions of the vertebrate midbrain as a “behavioral core control system.”
The midbrain supports autonomous decision making, as well as serving as the “final
common pathway” for action planning. This is important, since adaptive behavior