1 Introduction
The difference between a system with a first-person subjective
point-of-view and a system without is the explanatory gap between our
physical theories and what it is to have a mental life (Godfrey-Smith 2019,
1–2). There is undeniably something it is like to be a person
working on an essay for a class and seemingly nothing it is like to be
the chair in which the essay is written. Phenomenal consciousness,
henceforth referred to as p-consciousness, is what this experience of
something it is like to be something is called (Block 1995, 230).
Humans are confident in one another’s phenomenal consciousness through
the process of verbal reporting. We communicate using language and
convince each other of the existence of our p-consciousness through the
similarity of our described experience. We remain relatively confident
in the p-consciousness of other mammals due to the similarity of their
brains to ours; mammalian brains possess all the same key features as
our own, albeit smaller (Birch 2020, 288). We run into
methodological problems; however, in the case of non-mammals,
non-mammals may possess a homologous neural structure to mammals, but
they may not. To write off all other animal classes immediately from
possesing p-consciousness does not entertain the possibility of multiple
realizability, the ability of different underlying mechanisms to perform
the same functions.
In this essay, I will examine the work of Simona Ginsburg and Eva
Jablonka; they have developed an approach that links a certain forms of
enhanced learning with p-consciousness. They argue that organisms that
can perform these types of learning are p-conscious and thus have their
own subjective experience. These learning types are grouped and
collectively called unlimited associative learning (UAL). This means an
organism can learn about its environment and itself in an open-ended
manner (Birch et al. 2020,
56). The constraints on this learning still allow that it is
functionally impossible for all of the possible associative links to be
exhausted in the finite lifetime of a real organism (Birch et al.
2020, 56).
To identify UAL with consciousness, Ginsburg and Jablonka use a cluster
of enhanced associative learning forms that in humans, are believed to
be facilitated by p-consciousness (Birch 2020, 291–92). These are;
(1) compound stimuli, the stimulus can be made
of more than one perceptual feature in either a single or multiple sense
modalities (Birch et al. 2020,
56). (2) Novel stimuli, the conditioned
stimulus can be novel to the organism, (3)
second-order conditioning, the organism can build up open-ended
chains of association between stimuli and actions (Birch et al. 2020,
56). (4) Trace conditioning, there can
be a time gap between conditioned and unconditioned stimuli and,
(5) Flexible, easily rewriteable associations with
value, stimuli can be either positive or negative and the organism
can alter the value in response to environmental changes (Birch et al.
2020, 56).
These five points indicate how UAL can be useful as a research programme
once we have ascertained how UAL relates to a minimally sufficient
functional description of p-consciousness. We will examine the relation
of UAL to p-consciousness in the next section. First, is UAL
p-consciousness or just an indicator? Then we shall look at the
implications of UAL on p-consciousness. Finally, we shall examine the
some of the support and criticism of UAL.
2 Natural or nominal kind
In this section, we will further explore UAL, as its authors and
later collaborators envisage and understand it to be. Next, we shall
examine UAL in the context of another theory created by Nicholas Shea
and Tim Bayne, and we shall use their framework to assess the relation
between UAL and p-consciousness. Then finally, for this section, we
shall discuss the implications of whatever we find the status of the
relationship between UAL and p-consciousness to be.
Ginsburg and Jablonka’s theory is an evolutionary theory and proceeds in
stages marked by three transitions. First from non-life to life, from
automata to p-consciousness and, finally, to rational-symbolic (Ginsburg and
Jablonka 2020, 158). An evolutionary transition marker signals
each transition; once this diagnostic is present, the transition has
occurred, and the organism has gained a new mode of being not possessed
by its progenitors (Ginsburg and Jablonka 2020,
157–59). We concern ourselves with the second transition from
automata to p-consciousness; UAL is Ginsburg and Jablonka’s proposed
evolutionary transition marker for this transition. The authors maintain
that UAL is diagnostic to positively mark p-consciousness in an
evolutionary lineage (Ginsburg and Jablonka 2019,
226) and not necessarily anything more. However, if it is a good
minimal description of what is required functionally for
p-consciousness, then does it not constitute essentially being
p-consciousness?
To examine this question, we will use the theory mentioned above by Shea
and Bayne, their ‘natural kind methodology’. This framework is related
to our beliefs about p-consciousness; whether it is an underlying
natural property that entails our perceived ‘symptoms’ of
p-consciousness meaning, it is a natural kind (Shea and Bayne 2010, 470).
Alternatively, is it simply a family of cognitive abilities with no
underlying natural property, i.e., a nominal kind like the concept of
dirt (Shea and Bayne
2010, 470). If p-consciousness is a natural kind, to use Shea and
Bayne’s example, it is more like hepatitis C than dirt (Shea and Bayne 2010,
470). We can detect or diagnose hepatitis C without a
near-perfect detection method if we find enough symptoms of it and can
rule out other causes. This is the approach that the natural kind
methodology proposes; if p-consciousness is a natural kind we can detect
it with a diagnostic toolkit. Identifying a cluster of properties is not
a guarantee of the presence of a natural kind property to best explain
the cluster. However, suppose p-consciousness was a natural kind. In
that case, the methodology does allow more investigation than can be
made by pre-theoretic approaches that we already possess, such as verbal
reporting (Shea and
Bayne 2010, 470). Should p-consciousness be a natural kind, we
can collect a wide array of evidence and form nomological clusters. A
nomological cluster is a set of evidential properties that are
‘instantiated together better than chance’ and where ‘subsets of the
cluster supports induction to other elements of the cluster’ (Shea and Bayne 2010,
471).
Ginsburg and Jablonka’s UAL seems to fulfil these criteria for a
nomological cluster, each of the five features of UAL overlap and are
not individual and separate capacities. They have co-evolved with one
another (Birch et al. 2020,
56). For Jonathon Birch, Ginsburg and Jablonka to believe that
UAL is a good diagnostic, they must believe that it requires the
existence of the functionally coupled systems that result in
p-consciousness (Birch et al. 2020,
57). In other words, they have found a natural cluster of
evidential properties for which the best explanation is an underlying
natural kind property, i.e., p-consciousness. Hence, UAL is not
equivalent to p-consciousness; it is a nomological cluster that is best
explained by the existence of p-consciousness. Ginsburg and Jablonka are
p-consciousness realists and aim for their chosen nomological cluster to
provide evidence of natural kind p-consciousness.
A consequence of formulating their evolutionary transition marker in
Shea and Bayne’s natural kind methodology is explicitly stating the
pluralistic nature of the nomological cluster. Ginsburg and Jablonka’s
central thesis for UAL is that some forms of learning are facilitated by
p-consciousness (Ginsburg and Jablonka 2007,
219). Using the ‘natural kind methodology’, more evolutionary
transition markers (nomological clusters) could be found to add support
to Ginsburg and Jablonka’s theory. A synthesis of Ginsburg and
Jablonka’s and Shea and Bayne’s theories provides a comprehensive
research programme. We are licensed to search for UAL in non-human
animals, and we are licensed to search for more evidential properties
that are best explained by being facilitated by the natural kind
existence of p-consciousness. Discovering more interrelated evidential
properties allows us to form more nomological clusters to further aid
empirical research.
In the next section, we shall see support and criticism for Ginsburg and
Jablonka’s theory of their methodology and realist stance on
p-consciousness.
3 Support and criticism
Firstly, supporting evidence for UAL comes from the fossil record.
Ginsburg and Jablonka argue that in the fossil record for the Cambrian
period (542-485 million years ago (Ginsburg and Jablonka 2019,
407)), we can see features in some vertebrate organisms, i.e., a
nervous system (Ginsburg and Jablonka 2007,
222–23). Neural structure is strongly associated with UAL in
modern humans and other animals and allows for the five criteria
contained within the UAL nomological cluster. Ginsburg and Jablonka
argue that from an evolutionary perspective, the arrival in the Cambrian
period of open-ended associative learning is very sensible. There was a
sudden diversification of life and a massive increase in the number of
species during this period, known as the Cambrian explosion (Godfrey-Smith 2016,
11). Ginsburg and Jablonka believe that this ‘explosion’ was due
to the development of UAL (Ginsburg and Jablonka 2019,
419–20). UAL endowed organisms with the ability to adapt their
behaviour to their benefit within single organism lifetimes, which in
turn spurred on morphological diversity as an animal that discovered and
exploited a novel food source was likely to stay and reproduce in that
environment (Ginsburg and Jablonka 2019,
420–21). The organism’s offspring were more likely to exploit the
same food source, and from there, the pressures of natural selection
created more morphological niches to fill (Ginsburg and Jablonka 2019,
421). This is a compelling hypothesis and lends some strength to
the evolutionary approach that Ginsburg and Jablonka due to its
explanatory power. Additionally, it strongly uses the continuity
principle of the broader biogenic approach, that there is a continuous
unbroken lineage with our most complex cognition in the present day with
simpler forms in the past (Lyon 2006, 15). Ginsburg and Jablonka
determine the origins of p-consciousness to be far in the past, allowing
for a wide distribution of p-consciousness in the present day.
The first criticism we will examine is the application of UAL as an
indicator of p-consciousness to artificially intelligent systems. It
looks pretty conceivable that an AI system that could currently or in
the near future be built able to demonstrate UAL fully. Would this AI
then be considered p-conscious? The position of Birch, Ginsburg and
Jablonka is simply no (Birch et al. 2020,
56). They argue that UAL is an evolutionary transition marker for
biological systems, not computer systems (Birch et al. 2020,
56). I find this a somewhat weak argument; they imply that what
matters are biology and the fact that some evolution has occurred; we
shall examine these two premises. The premise of biological necessity
seems contradictory to the weight they place on functionalism. Ginsburg
and Jablonka state that a form of learning, UAL, is a marker for
p-consciousness and that it is multiply realisable so that non-mammalian
animals can become candidates for p-consciousness. Suppose the crucial
aspect of p-consciousness is UAL functionalism. In that case, I do not
see why AI is logically ruled out, as Birch, Ginsburg and Jablonka
believe, as a candidate for functionally realising UAL in a different
material substrate. The second point regarding evolution is perhaps even
easier to critique. Many AIs are evolutionary algorithms; staggeringly
large numbers of iterations of the AI are produced in discrete
generations, and a form of Darwinian evolution is applied (Mirjalili et
al. 2020, 3). The ‘fittest’ AI is the end product. Hence the
evolutionary aspect need not be a barrier to AI being considered a
candidate for p-consciousness using the UAL evolutionary transition
marker. From the realist position on p-consciousness we would need to
know more about what sort of natural kind p-consciousness actually is.
This is something that theory-light approaches, frameworks that do not
operate with a fully specified theory of consciousness (Birch 2022, 140–41),
such as Ginsburg and Jablonka’s and Shea and Bayne’s, purposefully avoid
preventing overt anthropomorphism. So it seems inconsistent to reject
artificial cases without careful examination when they undergo evolution
to achieve UAL functionally.
Now we shall examine a criticism of the markers of UAL. UAL, as an
indicator of p-consciousness, cannot give verdicts on organisms that do
not display the five criteria listed in the introduction. There is a
risk of false negatives. Animals that do possess p-consciousness run the
risk of being labelled as not p-conscious. The problem here is that
criteria such as (4) trace conditioning might
provide evidence of p-consciousness in humans. However, it is unclear
why its absence in a non-human animal should mean that the animal does
not possess p-consciousness (Shevlin 2021, 308). An organism that
relies heavily on one sense would have very little reason to have
developed or retained a capacity for multisensory learning (Shevlin 2021, 308).
The connection between learning and p-consciousness can be more broadly
questioned. Henry Shevlin suggests that many examples of p-conscious
examples in humans, such as perceptions and bodily sensations, are
seemingly unconnected to the human capacity for learning (Shevlin 2021, 308).
Although Shevlin’s point is not without merit, Ginsburg and Jablonka’s
UAL does provide the ability for assessment and diagnosis; knowledge of
perception and bodily qualia in humans requires a verbal report. This
cannot be done with other animals. We should perhaps not use UAL to rule
out an organism from p-consciousness but only to positively ascribe
p-consciousness to animals capable of UAL. Perhaps with the
identification and addition of more nomological clusters, we may be able
to strengthen the negative case to a tolerable uncertainty.
Our final point is to question whether p-consciousness is a natural kind
at all. The ‘natural kind methodology’ that Ginsburg and Jablonka have
contributed to relies on the premise that p-consciousness is a natural
kind. If it is not and is, in fact, a nominal kind, their methodology
will not identify p-consciousness, only unlimited associative learning.
Daniel Dennett proposes a qualia irrealist stance that equates
p-consciousness with access consciousness (Dennett 2015, 2). Access consciousness
henceforth referred to as a-consciousness, is the availability of a
perception for reasoning and rationally guiding action (Block 1995, 227).
Dennett argues that whilst it seems intuitive that the qualia of
subjective experience exist, this is, in fact, only an intuition (Dennett 2004,
60–61). Furthermore, Dennett argues that we do not know the
limits of sufficiently detailed physical theories. Therefore, it is at
least possible that some piece of supposedly non-scientific knowledge,
such as the ‘redness’ of red, could be determined scientifically (Dennett 2004,
60–61). Hence, there is a possibility that p-consciousness is not
a natural kind. Suppose this is the case and it is a nominal kind;
Ginsburg and Jablonka’s research programme would be doomed. UAL testing
would only be able to tell us about varying capacities of
a-consciousness in animals and not provide a definitive and helpful
diagnostic of p-consciousness as intended.
4 Conclusions
To conclude, we have seen that Ginsburg and Jablonka’s evolutionary
transition marker theory fulfils the description of the natural kind
methodology proposed by Shea and Bayne. This leads us to believe that
Ginsburg and Jablonka must be p-consciousness realists. Shea and Bayne’s
framework allows us to answer the question as to whether UAL is
equatable to p-consciousness. We have found it is not and is a
nomological cluster of evidential properties, best explained by the
natural kind existence of p-consciousness in the realist position. The
connection between Ginsburg and Jablonka’s and Shea and Bayne’s
respective frameworks allows us to note the possibility of a more
comprehensive research programme following the natural kind
methodology.
The evolutionary perspective of Ginsburg and Jablonka’s theory provides
a compelling hypothesis as to the early origins of p-consciousness and
its role in the Cambrian explosion of new species. This contains that
the distribution of p-consciousness in the animal phyla may be extensive
across many different neural structures, all realising p-consciousness
in different ways.
I also dispute that UAL can be dismissed out of hand when applied to AI
due to the functionalism approach employed by Ginsburg and Jablonka.
They appeal to multiple realisability to allow for animals with wildly
different neural structures to mammals to be included as candidates for
p-consciousness. Additionally, AIs are already produced with
evolutionary algorithms, so their dismissal from the evolutionary
transition marker framework could be categorised as biased and
biological chauvinism.
We have also concluded that UAL is susceptible to false negatives and
should be supplemented with other nomological clusters to reduce this
uncertainty. We should also exercise caution when applying the UAL
diagnostic to exotic organisms that differ significantly from ourselves.
Some UAL criteria do not seem reasonable when applied to some creatures
in some environments where the precise form of learning being tested
would have little use.
Whilst we admit the possibility due to Dennett’s critique that
p-consciousness might not be a natural kind, the possibility for further
empirical research using the natural kind methodology is too alluring to
give up at present. Not only can researchers into non-human
p-consciousness perform empirical research with animals using the UAL
diagnostic. Additionally, they can search for more nomological clusters
to reduce uncertainty in their assessments, making UAL and its
underlying natural kind methodology a profitable investment of time and
resources going forward.