Your Words May Predit Your Future Mental Health

Mariano Sigman
G0:12
We have historical records that allow us to know how
the ancient Greeks dressed, how they lived, how they
fought ... but how did they think?
0:22
One natural idea is that the deepest aspects of human
thought -- our ability to imagine, to be conscious, to
dream -- have always been the same. Another
possibility is that the social transformations that have
shaped our culture may have also changed the
structural columns of human thought.
0:43
We may all have different opinions about this. Actually,
it's a long-standing philosophical debate. But is this
question even amenable to science?
0:53
Here I'd like to propose that in the same way we can
reconstruct how the ancient Greek cities looked just
based on a few bricks, that the writings of a culture are
the archaeological records, the fossils, of human
thought.
1:10
And in fact, doing some form of psychological analysis
of some of the most ancient books of human culture,
Julian Jaynes came up in the '70s with a very wild and
radical hypothesis: that only 3,000 years ago, humans
were what today we would call schizophrenics. And he
made this claim based on the fact that the first humans
described in these books behaved consistently, in
different traditions and in different places of the world,
as if they were hearing and obeying voices that they
perceived as coming from the Gods, or from the muses
... what today we would call hallucinations. And only
then, as time went on, they began to recognize that they
were the creators, the owners of these inner voices. And
with this, they gained introspection: the ability to think
about their own thoughts.
2:10
So Jaynes's theory is that consciousness, at least in the
way we perceive it today, where we feel that we are the
pilots of our own existence -- is a quite recent cultural
development. And this theory is quite spectacular, but it
has an obvious problem which is that it's built on just a
few and very specific examples. So the question is
whether the theory that introspection built up in human
history only about 3,000 years ago can be examined in a
quantitative and objective manner.
2:42
And the problem of how to go about this is quite
obvious. It's not like Plato woke up one day and then he
wrote, "Hello, I'm Plato, and as of today, I have a fully
introspective consciousness."
2:54
(Laughter)
2:56
And this tells us actually what is the essence of the
problem. We need to find the emergence of a concept
that's never said. The word introspection does not
appear a single time in the books we want to analyze.
3:12
So our way to solve this is to build the space of words.
This is a huge space that contains all words in such a
way that the distance between any two of them is
indicative of how closely related they are. So for
instance, you want the words "dog" and "cat" to be very
close together, but the words "grapefruit" and
"logarithm" to be very far away. And this has to be true
for any two words within the space.
3:40
And there are different ways that we can construct the
space of words. One is just asking the experts, a bit like
we do with dictionaries. Another possibility is following
the simple assumption that when two words are related,
they tend to appear in the same sentences, in the same
paragraphs, in the same documents, more often than
would be expected just by pure chance. And this simple
hypothesis, this simple method, with some
computational tricks that have to do with the fact that
this is a very complex and high-dimensional space,
turns out to be quite effective.
4:15
And just to give you a flavor of how well this works, this
is the result we get when we analyze this for some
familiar words. And you can see first that words
automatically organize into semantic neighborhoods. So
you get the fruits, the body parts, the computer parts,
the scientific terms and so on.
4:32
The algorithm also identifies that we organize concepts
in a hierarchy. So for instance, you can see that the
scientific terms break down into two subcategories of
the astronomic and the physics terms. And then there
are very fine things. For instance, the word astronomy,
which seems a bit bizarre where it is, is actually exactly
where it should be, between what it is, an actual
science, and between what it describes, the
astronomical terms.
4:59
And we could go on and on with this. Actually, if you
stare at this for a while, and you just build random
trajectories, you will see that it actually feels a bit like
doing poetry. And this is because, in a way, walking in
this space is like walking in the mind.
5:15
And the last thing is that this algorithm also identifies
what are our intuitions, of which words should lead in
the neighborhood of introspection. So for instance,
words such as "self," "guilt," "reason," "emotion," are
very close to "introspection," but other words, such as
"red," "football," "candle," "banana," are just very far
away.
5:37
And so once we've built the space, the question of the
history of introspection, or of the history of any concept
which before could seem abstract and somehow vague,
becomes concrete -- becomes amenable to quantitative
science.
5:55
All that we have to do is take the books, we digitize
them, and we take this stream of words as a trajectory
and project them into the space, and then we ask
whether this trajectory spends significant time circling
closely to the concept of introspection.
6:11
And with this, we could analyze the history of
introspection in the ancient Greek tradition, for which we
have the best available written record. So what we did
is we took all the books -- we just ordered them by time
-- for each book we take the words and we project them
to the space, and then we ask for each word how close
it is to introspection, and we just average that. And then
we ask whether, as time goes on and on, these books
get closer, and closer and closer to the concept of
introspection.
6:41
And this is exactly what happens in the ancient Greek
tradition. So you can see that for the oldest books in the
Homeric tradition, there is a small increase with books
getting closer to introspection. But about four centuries
before Christ, this starts ramping up very rapidly to an
almost five-fold increase of books getting closer, and
closer and closer to the concept of introspection. And
one of the nice things about this is that now we can ask
whether this is also true in a different, independent
tradition.
7:13
So we just ran this same analysis on the Judeo-
Christian tradition, and we got virtually the same
pattern. Again, you see a small increase for the oldest
books in the Old Testament, and then it increases much
more rapidly in the new books of the New Testament.
And then we get the peak of introspection in "The
Confessions of Saint Augustine," about four centuries
after Christ. And this was very important, because Saint
Augustine had been recognized by scholars, philologists,
historians, as one of the founders of introspection.
Actually, some believe him to be the father of modern
psychology.
7:50
So our algorithm, which has the virtue of being
quantitative, of being objective, and of course of being
extremely fast -- it just runs in a fraction of a second --
can capture some of the most important conclusions of
this long tradition of investigation. And this is in a way
one of the beauties of science, which is that now this
idea can be translated and generalized to a whole lot of
different domains.
8:17
So in the same way that we asked about the past of
human consciousness, maybe the most challenging
question we can pose to ourselves is whether this can
tell us something about the future of our own
consciousness. To put it more precisely, whether the
words we say today can tell us something of where our
minds will be in a few days, in a few months or a few
years from now.
8:42
And in the same way many of us are now wearing
sensors that detect our heart rate, our respiration, our
genes, on the hopes that this may help us prevent
diseases, we can ask whether monitoring and analyzing
the words we speak, we tweet, we email, we write, can
tell us ahead of time whether something may go wrong
with our minds. And with Guillermo Cecchi, who has
been my brother in this adventure, we took on this task.
And we did so by analyzing the recorded speech of 34
young people who were at a high risk of developing
schizophrenia.
9:22
And so what we did is, we measured speech at day one,
and then we asked whether the properties of the speech
could predict, within a window of almost three years, the
future development of psychosis. But despite our hopes,
we got failure after failure. There was just not enough
information in semantics to predict the future
organization of the mind. It was good enough to
distinguish between a group of schizophrenics and a
control group, a bit like we had done for the ancient
texts, but not to predict the future onset of psychosis.
10:00
But then we realized that maybe the most important
thing was not so much what they were saying, but how
they were saying it. More specifically, it was not in
which semantic neighborhoods the words were, but how
far and fast they jumped from one semantic
neighborhood to the other one. And so we came up with
this measure, which we termed semantic coherence,
which essentially measures the persistence of speech
within one semantic topic, within one semantic
category.
10:30
And it turned out to be that for this group of 34 people,
the algorithm based on semantic coherence could
predict, with 100 percent accuracy, who developed
psychosis and who will not. And this was something
that could not be achieved -- not even close -- with all
the other existing clinical measures.
10:53
And I remember vividly, while I was working on this, I
was sitting at my computer and I saw a bunch of tweets
by Polo -- Polo had been my first student back in Buenos
Aires, and at the time he was living in New York. And
there was something in this tweets -- I could not tell
exactly what because nothing was said explicitly -- but I
got this strong hunch, this strong intuition, that
something was going wrong. So I picked up the phone,
and I called Polo, and in fact he was not feeling well.
And this simple fact, that reading in between the lines, I
could sense, through words, his feelings, was a simple,
but very effective way to help.
11:36
What I tell you today is that we're getting close to
understanding how we can convert this intuition that we
all have, that we all share, into an algorithm. And in
doing so, we may be seeing in the future a very different
form of mental health, based on objective, quantitative
and automated analysis of the words we write, of the
words we say.
12:04
Gracias.
12:05
(Applause)

Credit:
Youtbe.com

Mariano Sigman@ted.com/talks