From: joedees@bellsouth.net
Date: Tue Sep 24 2002 - 16:13:51 MDT
Linguists Decipher Warning Message in
Genome
February 1, 2039
BOSTON--A group of researchers at MIT's Chomsky Institute
announced yesterday independent confirmation of their discovery
of a series of messages encoded in apparently dormant or unused
sections of the human genome. "We're able to report replication of
our results by at least three independent teams," explained the
team's project director Klara Tulip. "We hence feel quite
confident about the results and felt that they were significant
enough to warrant preliminary public release."
Exploiting evolved, mathematical models derived from iterative
analyses of network-available audio, video and text files in more
than 200 languages, the team scanned files in the Human Genome
Library for patterns consistent with the presence of a "semantic
system." "We were actually using the Genome Library as a control
data-set to be sure that our model wasn't producing false
positives," explains Tulip. "We'd developed a mathematical and
algorithmic formulation of a meta-language descriptive of all
known human linguistic systems and needed to test it against
some non-random data that we assumed had no semantic content.
We we're stunned to find that the genome contains sequences
consistent with an implied linguistic system."
Within days of discovering the presence of "semantic sequences"
the team had also isolated a "Rosetta Stone" enabling them to
partially decipher and translate a number of passages. "The
genome appears to contain a linguistic system of remarkable
economy," notes Tulip. "Like a coded message that includes
detailed instructions for how it is to be decoded."
Though declining to reveal the full results of their analysis,
noting that some 97% of the human genome consists of
biologically unused sequences with "a statistically significant
chance of containing decipherable semantic content," the team did
release translations of a "number of passages of public interest,"
including the warning "NOT TO BE REMOVED EXCEPT BY
END USER."
Among other messages, the team isolated at least 42 varied
repetitions of the instruction to "[not] fold, spindle, or mutilate"
and two apparently inconsistent warranties, one claiming "absence
of defect in material or workmanship for 180 days from
formulation" and one disavowing "all warranties of fitness for use
except as otherwise required." "Our initial analysis has uncovered
a number of repetitions, counter-factuals, and internal-
inconsistencies suggesting that these genomic messages are a
product of the same evolutionary forces driving reproduction of
the non-semantic portions of the genome," observes Tulip.
Responding to news of the team's discovery, critics, including a
number of prominent linguists and bioinformaticians, characterize
the research as a Rorschach Test revealing more about the
researchers' assumptions than about the meaning of human genes.
"You have to look closely at their model, at what their meta-
linguistic model assumes about the world," notes Harvard
Professor of Statistics Joseph Climb. "If you go into the world
with a sufficiently abstract model of 'language' you'll start finding
Shakespeare inside rocks and twigs."
Discounting such criticism as "mathematically unsophisticated,"
project leader Tulip points to the astronomical odds against "a
chance consistency that would permit our model to identify such a
vast pool of semantically significant sequences. Our genome has
something to say. The real question is why--what evolutionary
purpose could these messages serve?"
This archive was generated by hypermail 2.1.5 : Wed Sep 25 2002 - 13:28:59 MDT