Last month, Ryan Williams, the president of the Claremont Institute, published an essay in The American Mind that argued multiculturalism is undermining traditional American values. I published a response here in The Federalist criticizing his thoughtful, well-written argument on a few significant points. My colleague, John Daniel Davidson, ran another article in The Federalist respectfully criticizing my objections. This is how discourse is supposed to work. But it seems Google is not aware of this.
Instead this week Google responded to the Williams article by denying the Claremont Institute the ability to purchase advertising for its 40th anniversary gala. This was the result of the American Mind bizarrely being labeled a racially or ethnically motivated publication, which the tech giant’s terms of service bar from running ads.
Yesterday Google’s acting director of political and stakeholder outreach told Claremont the label and refusal to sell advertising was a mistake, but only after several conservative outlets and half of conservative Twitter expressed outrage over the suppression of Claremont’s speech. That frankly just isn’t good enough. If this is happening to Claremont, then it is almost certainly happening to individuals and groups that cannot muster the support Claremont could to address the situation.
The real question here is how could such a mistake have been made in the first place? Was this the result of an algorithm? Or did some person or persons tag the American Mind for this treatment? Last year, the heads of both Twitter and Facebook testified before Congress about, among other things, political censorship on their platforms, Google was also invited to testify but declined the invitation.
Congress learned from the social media companies that they haven’t figured out how to police the content of millions of users without bias against conservatives. Twitter head Jack Dorsey was clearest in saying this bias against conservatives exists, but he offered no solution to the problem.
There are two basic ways that Internet companies flag content that supposedly violates their terms of service. First is reporting. It is possible that Google received complaints regarding Williams’ original essay, but in all likelihood Claremont was flagged using the second method, which is basically algorithms searching content for terms or phrases that make it more likely to be in violation.
In this case the term that most likely triggered the action was “multicultural,” which is interesting because the use of that word, or in my opinion the misuse of it, was at the core of my criticism of Williams’s essay. But whether one agrees with me or Williams on that issue, what all should be able to agree on is that there was nothing offensive or racialized about the way Williams uses that term. While historically “multicultural” has been used to describe societies becoming more racially diverse, Williams was clear that was not at the root of his usage. He was talking about something much closer to the idea of cultural relativism, in which traditional values are thrown overboard.
But why might a word like multicultural trigger a review in the first place? It might be a word that does appear in legitimately racist or offensive content, which is therefore captured by the algorithm. But the vast majority of the time the content Internet companies punish is right of center. So is the bias of typically progressive software employees driving the algorithm? Is Google, perhaps without even knowing it, importing or uploading a political worldview into its massive systems?
That would be bad, and seems likely to be the case. But it is not the worst part of this incident. The worst part is the American Mind was punished for engaging in exactly the kind of respectful discourse that our society needs to build greater understanding. If merely mentioning controversial words or subjects triggers a clampdown, how on earth is anyone supposed to even talk about those subjects? The result is not just a chilling of speech, it is to set barriers to the kind of discourse America so badly needs today.
I believe it was Ludwig Wittgenstein who said that language is not merely the vessel of our thought, it is also the driver. This is perhaps even more true for Internet companies, whose algorithms now make daily decisions about what materials we are and are not exposed to. Any kind of political bias in those systems can have dramatic effects, as Claremont found out.
Although Google’s reaction to Claremont is a particularly outrageous example of anti-conservative bias, it is far from an outlier. This is something conservatives have become accustomed to, but which we must not accept. Our outlets, our leadership, and our politicians must continue to keep tech companies’ feet to the fire on this issue. Conservatives have as much right to speak and make their case as the left does. As the Internet takes over more and more of our lives, we must work to ensure it is a place for lively and important debate, not a database of rightthink and wrongthink.