What is consciousness? A new idea about what consciousness is and why we have it reveals how we could recreate it

---

What is consciousness? 
A new idea about what consciousness is and why we have it reveals 
how we could recreate it 
CONSCIOUSNESS is a slippery concept. It 
isn’t just the stuff in your head. 
It is the subjective experience of some of that stuff. When you stub your toe, 
your brain d
oesn’t merely process information and trigger a reaction: you have 
a feeling of pain. When you are happy, you experience joy. The ethereal nature 
of experience is the mystery at the heart of consciousness. How does the brain, 
a physical object, generate a non-physical essence? 
“Far from being some sort of magical property, consciousness is a tool 
of great, practical power” 
This experience-ness explains why pinning down consciousness has been 
described as “the hard problem”. Subjective experience doesn’t exist in any 
physical dimension. You 
can’t push it and measure a reaction force, scratch it 
and measure its hardness or put it on a scale and measure its weight. 
Philosophers have described it as the “ghost in the machine”. Even scientific 

---

ideas about consciousness often have an aura of the metaphysical. Many 
scientists describe it as an illusion, while others see it as so fundamental that 
it doesn’t have an explanation. Always at the centre of the riddle lies its non-
physicality. 
But what if consciousness isn’t so mystical after all? Perhaps we have just 
been asking the wrong question. Instead of trying to grapple with the hard 
problem, my colleagues and I at Princeton University take a more down-to-
earth approach. My background lies in the neuroscience of movement control, 
what you could call the robotics of the brain. Drawing on that, I suggest that 
consciousness can be understood best from an engineering perspective. Far 
from being some sort of magical property, it is a tool of extraordinary power. It 
is a tool that can be engineered into machines. Our new approach shows how.
Because the normal methods 
of observation and measurement don’t quite 
apply, the study of consciousness has always sat uneasily in mainstream 
science. A few decades ago, The International Dictionary of Psychology 
described consciousness as “a fascinating but elusive phenomenon; it is 
impossible to specify what it is, what it does or why it evolved. Nothing worth 
reading has ever been written about it.” Since then, consciousness has 
become an increasingly popular topic in science, generating numerous ideas 
and thousands of papers but very little agreement.
One approach searches for the neural correlates of consciousness: the 
minimal physical signature in the brain needed for subjective experience. 
There have been some interesting leads, but the hunt continues. Other 
researchers buil
d on the insight that consciousness isn’t just a stimulus 
processed in the brain. Their higher-order thought theory proposes that the 
brain contains a system that re-represents the stimulus at a higher level with 
added self-information, which is how we become conscious of it. Exactly what 
that higher-order information is, what cognitive purpose it serves and where in 
the brain it is constructed are all debated 
– although some people associate it 
with the prefrontal cortex. 
A particularly influential idea is known as global workspace theory. Here, 
information coming both from outside and within the brain competes for 
attention. Information that wins this competition becomes globally accessible 
by systems throughout the brain so that we become aware of it and are able 
to process it deeply. Also popular is the integrated information theory. It sees 
consciousness as an emergent property of complex systems and posits that 
the amount of consciousness in any system can be measured in units 
called phi. Phi is high in the human brain, but also present in everything from 

---

a hamburger to the universe, since everything contains at least some 
integrated information.
Then there is the idea that consciousness is an illusion. This is often 
misinterpreted. It doesn’t mean that consciousness doesn’t exist, or that we 
are fooled into thinking we have it. Instead, it likens consciousness to the 
illusion created for the user of a human-computer interface and argues that the 
metaphysical properties we attribute to ourselves are wrong. Researchers 
debate the exact source of these mistaken self-descriptions and the reason we 
seem to be mentally captive to them.
Engineering, and the science of robotics in particular, tells us that every good 
control device needs a model 
– a quick sketch – of the thing it is controlling. 
We already know from cognitive neuroscience that the brain constructs many 
internal models 
– bundles of information that represent items in the real world. 
These models are simplified descriptions, useful but not entirely accurate. For 
example, the brain has a model of the body 
– called the body schema – to help 
control movement of the limbs. When someone loses an arm, the model of the 
arm can linger on in the brain so that people report feeling a ghostly, phantom 
limb. But the truth is, all of us have phantom limbs, because we all have internal 
models of our real limbs that merely become more obvious if the real limb is 
gone.
By the same engineering logic, the brain needs to model many aspects of itself 
to be able to monitor and control itself. It needs a kind of phantom brain. One