Drew Margolin

Associate Professor

Drew Margolin, professor of communication at Cornell University, studies the way people communicate online and the role of accountability, credibility, and legitimacy within social networks. He says search engines are more complex than social network sites and that Google’s testimony is likely to create confusion among lawmakers.

Margolin says:

“After Congress struggled to understand how Facebook operated, and thus what its responsibilities ought to be, we should expect even more confusion in testimony from Google. Social networks sites are a recent invention, but the basic concept of what they do is intuitive to human beings, whereas search engines are much more complex, which makes their public responsibility much harder to gauge. They are not providing us with something that people are telling us, or answering a question we are asking, rather, they are providing us with a set of informed guesses that are most likely to satisfy our question.

“Google's particular technology relies on relationships among web pages, including the words they use and the ways they link to each other. Unlike social networking sites, Google does not control these underlying structures. It is just using the (online) world as it exists. This reinforces the possibility that search outcomes are ‘biased,’ in the sense that they do not show left-right balance or other kinds of balance, and yet at the same time makes it harder to require Google to correct the problem. American conservatives may be right that their viewpoints are not coming up in searches of particular issues, but this may be because these pages are poorly connected to the rest of the network.  

“I don't expect Google to be particularly forthcoming, but even if they are, it won't solve the conundrum, which is that the structure of information on which searches are built does not conform to traditional ideals of balance between opposing political viewpoints, and there isn't anything they can do about that. This doesn't mean Google doesn't have a responsibility to the public, or that its algorithm isn't exacerbating the problem, but that figuring out whether it is doing enough to fulfill its responsibility is very challenging.”