Page 17 - The 'X' Chronicles Newspaper - September 2023

Brains May Predict The Future
To Make Sense of the
Present, Brains May
Predict The Future
Continued from Page 16
The researchers observed a greater brain
response when the study’s subjects came
across the unexpected word “dog,”
characterized by a specific pattern of
electrical activity, known as the “N400
effect,” that peaked approximately 400
milliseconds after the word was revealed.
But how to interpret it remained unclear.
Was the brain reacting because the
word’s meaning was nonsensical in the
context of the sentence? Or might it have
been reacting because the word was
simply unanticipated, violating whatever
predictions the brain had made about
what to expect?
In 2005, Kutas and her team conducted
another study that pointed to the latter
possibility. People were again asked to
read a sentence one word at a time on a
screen: “The day was breezy so the boy
went outside to fly ____.” Because “a
kite” seemed the most likely way to
finish the sentence, the subjects expected
to see “a” next, a word that had no
intrinsic meaning but did predict the
word to follow. When the participants
saw “an” instead (as in “an airplane”),
they experienced an N400 effect,
seemingly because the brain had to
process a mismatch between its
expectation and reality. The effect was
apparently unrelated to the meaning of
the word or any difficulty in processing
the presented stimulus itself.
The 2005 finding seemed like a great fit
for the predictive coding framework. But
in April 2018, a paper published in eLife
reported that several labs have been
unable to replicate the result. Now, other
researchers have started to respond, some
claiming that subtleties in the replication
methods still favor the prediction-based
interpretation.
This back-and-forth reflects much of the
debate that’s surrounded predictive
coding. Experiments like Kutas’ are
subject to many interpretations. They can
be explained by models other than
predictive coding, and they fall short of
definitive proof of the hypothesis
because they don’t delve into the actual
17
mechanisms at play. While the idea that proposes that the brain makes
the brain is constantly making inferences probabilistic inferences about the world
(and comparing them to reality) is fairly based on an internal model, essentially
well-established at this point, proponents calculating a “best guess” about how to
of predictive coding have been seeking interpret what it’s perceiving (in line with
ways to prove that their particular the rules of Bayesian statistics, which
version of the story is the right one — quantifies the probability of an event
and that it extends to all of cognition. based on relevant information gleaned
from prior experiences). Rather than
Bayesian Brains and Efficient waiting for sensory information to drive
cognition, the brain is always actively
Computing
constructing hypotheses about how the
world works and using them to explain
The foundational insight that the brain experiences and fill in missing data.
perpetually makes and evaluates its own That’s why, according to some experts,
predictions about ongoing experiences we might think of perception as
wasn’t always taken for granted. The “controlled hallucination.”
view of neuroscience that dominated the
20th century characterized the brain’s In that vein, the Bayesian brain also
function as that of a feature detector: It explains why visual illusions work: Two
registers the presence of a stimulus, dots blinking in rapid alternation on a
processes it, and then sends signals to screen, for example, look like a single dot
produce a behavioral response. Activity moving back and forth, so our brains
in specific cells reflects the presence or unconsciously start to treat them like a
absence of stimuli in the physical world. single object. Understanding how objects
Some neurons in the visual cortex, for
instance, respond to the edges of objects move is a higher-level type of
knowledge, but it fundamentally
in view; others fire to indicate the influences how we perceive. The brain is
objects’ orientation, coloring or shading.
simply filling in gaps in information — in
this case, about motion — to paint a
But the process turned out to be far less picture that’s not entirely accurate.
straightforward than it seemed. Further
tests found that as the brain perceives, But in spite of the clear role that
say, a longer and longer line, the detector generative models and expectations play
neurons for lines stop firing even though in brain function, scientists have yet to
the line hasn’t disappeared. And the fact
that so much information seemed to be pinpoint exactly how that’s implemented
at the level of neural circuits. “The
communicated through mysterious top- Bayesian brain story is relatively agnostic
down feedback connections suggested about what the underlying mechanisms
that something else was going on.
are,” said Mark Sprevak, a professor of
philosophy of mind at the University of
That’s where the “Bayesian brain” comes Edinburgh in Scotland.
into play, a general framework with roots
dating back to the 1860s that flips the (Continued on Page 18)
traditional model on its head. The theory