In a series of papers studying the history of American innovation, Packalen and Bhattacharya indexed every one-word, two-word, and three-word phrase that appeared in more than 4 million patent texts in the last 175 years. To focus their search on truly new concepts, they recorded the year those phrases first appeared in a patent. Finally, they ranked each concept's popularity based on how many times it reappeared in later patents. Essentially, they trawled the billion-word literature of patents to document the birth-year and the lifespan of American concepts, from "plastic" to "world wide web" and "instant messaging."
The overall story, Bhattacharya told me, follows the shift from "atoms to bits"—from the loud world of trains and cars in the 19th century to the invisible life of software. But within that meta-narrative (and this is where the colors come in handy), you can see moments where one industry dominated the patent literature—like chemistry (black) in the 1930s, medicine (red) in the 1980s, and computers (green) in the last few decades.
Since the 1970s, medicine and computers have reigned over patents like no two categories have dominated any previous period of invention in U.S. history. In his Nobel address, Mullis described his midnight eureka as a solitary moment of invention on an lonely California road. But rather than seeing him as a lonely inventor, it might make more sense to view him as a product of his times, a medical scientist working in the 1980s, at the apex of medicine's potency in the patent literature. His PCR patent was a part of, and a catalyst for, its own chain reaction of innovation in genetics, from "genomic DNA" to "DNA sequencing" to "monoclonal antibodies." Patents that introduce entirely new fields of study (like PCR) spur much more new research than subtle tweaks to old ideas, Packalen and Bhattacharya found. Still, past research has shown that organizations like the National Institutes of Health and the National Science Foundation are more likely to subsidize projects in highly familiar areas. Indeed, one of the major implications of Packalen and Bhattacharya's research is that, by awarding established scientists in well-understood fields, the government is implicitly discouraging the most radical innovation.
Another theme of Packalen and Bhattacharya's research is that innovation has become more collaborative. Indeed, computers have not only taken over the world of inventions, but also they have changed the geography of innovation, Bhattacharya said. Larger cities have historically held an innovative advantage, because (the theory goes) their density of smarties speeds up debate on the merits of new ideas, which are often born raw and poorly understood. But the researchers found that in the last few decades, larger cities are no more likely to produce new ideas in patents than smaller cities that can just as easily connect online with their co-authors. "Perhaps due to the Internet, the advantage of larger cities appears to be eroding,” Packalen wrote in an email.
U.S. Invention: From Follower to Leader