Learning to Speak the Language of Life
Parlez-vous DNA? Habla Human Genome? Fluency in the Romance languages is great, but if you want to add some new life to your surroundings--literally--try uttering a few sentences in DNA.
Of course, the idea that the double helix represents not just a scaffolding for biochemicals but a new kind of language has been around ever since Watson and Crick first discovered it 40 years ago. Biologists regularly talk of the genetic “code,” gene “expression” and “editing” gene sequences. DNA-as-language is one of modern science’s most powerful metaphors.
“The metaphor does give you a richer set of approaches to organize how you think about a problem,” says Nobel laureate Joshua Lederberg, a pioneering molecular biology researcher and president emeritus of Rockefeller University.
But maybe language is more than just a metaphor. Maybe--just like Sanskrit, Chinese and English--DNA really is a language, with a grammar and syntax that determines how meaning is created.
“I imagine that a gene is a kind of biological sentence--not just a sequence of letters,” says David B. Searls, a University of Pennsylvania professor with advanced degrees in molecular biology and computer science. “I try to assemble them into a higher order--into phrases that can ultimately be combined into sentences.”
To the extent that’s possible, our ability to understand and manipulate life is radically transformed. Language becomes a practical tool, not just a helpful metaphor.
“If there is such a thing as a genetic grammar or dialect that tells us what sequences and phrases are necessary to make things work,” says biotechnology industry consultant Vivian Lee, “it would have tremendous ramifications. . . . Understanding the patterns of DNA is a hot button area.”
As with any language, the ability to parse a genetic sentence--along with a decent bio-vocabulary--ultimately enables us to “read” the genome of any plant, animal or living thing with the same facility that we now read a newspaper or a novel. Molecular biologists have just begun to build “parsers” that can analyze genetic sentences, much as Mrs. Grundy had you parse sentences in sixth-grade English.
More provocatively, a good command of genetic grammar means we can ultimately write and edit genetic “texts” with a confidence that is otherwise unimaginable. Understanding the syntax is essential to making sure our genetic scribblings make sense.
“That’s the single greatest strength of grammars,” says Searls. “They’re as adept at generating a sentence as recognizing one.”
In a DNA-as-language world, the biotech business turns into the bio-publishing business. Writers--novelists? poets? screenwriters?--and editors proof genetic copy before it’s sent off to the “printers” to be published as a new vaccine, pharmaceutical or life form. The bigger our vocabularies, the richer and more comprehensive our organic literature becomes.
Talk about the pen mightier than the sword. Who would write these texts? And who would publish them? As with other manuscripts, will desktop publishing emerge? If it does, could just anyone create life? What restrictions should apply? Would free speech be an issue?
At the root of efforts to bring good grammar to genetics has been the emergence of one of the hottest research areas in computer science: computational linguistics.
Human languages are natural; computer languages are not. They must be rigorously designed and constructed. As a result, computer scientists have developed a formidable body of mathematics describing how such languages work.
Linguists have been using these theorems and equations to better understand how we speak English and German. Scientists like Searls recognize that this mathematics--carefully applied--may reveal the underlying grammar of DNA and the gene. Indeed, as work in computational linguistics continues, the insights generated will no doubt influence molecular biology.
Insights from genetic analysis may, in turn, shape the way computer languages are designed. Consequently, a new fusion between computer science and genetic engineering is probably emerging.
Michael Schrage is a writer, consultant and research associate at the Massachusetts Institute of Technology. He writes this column independently for The Times.
