Humanizing Political Language

On a recent episode of Your Undivided Attention, we explore how language shapes politics in subtle ways – either to make political concepts polarizing or less alienating and more unifying.

Our guests come from both sides of the aisle: Drew Westen is a political psychologist and messaging consultant who has advised the Democratic Party, and Frank Lutz is a communications consultant, polster, and pundit who has advised the Republican Party.

Drew Westen and Frank Luntz generally advocate for more straightforward, humanizing language. Westen and Luntz offer several examples of how and why we should make our language less polarizing:

‍

‍

Their suggestions highlight a few valuable takeaways:

• Language is a powerful tool. We can either choose language that divides and confuses, or language that unites and clarifies.
• Language is contextual and constantly evolving. The language that resonates with one group may not resonate with another. Similarly, the language we use today may not be the language we use tomorrow.
• Language that unifies will sound like the conversations we have with our extended families over the “kitchen table.”

Language plays an essential role in creating shared understanding. To learn more, listen to the full podcast episode with Drew Westen and Frank Luntz, outlined below.

‍

‍

WHAT WE’RE READING, LISTENING TO, AND WATCHING

• Last week, a Google engineer claimed that LaMDA, the company’s developing language model meant to power chatbots, was “sentient.” Quickly, AI researchers and ethicists pushed back on the claim. In particular, Timnit Gebru and Margaret Mitchell also note that assigning sentience to AI can result in accountability gaps, as responsibility for wrongdoing is transferred to the "independent being, rather than the company that created it.” (Note: Both Gebru and Mitchell were fired from Google in 2020 for raising the alarm about the safety of Google’s AI products, including that of perceived sentience.)
  
  Casey Newton outlines a brief history of Silicon Valley’s push for “human AI” and emphasizes the complications of people believing that AI is sentient whether or not it actually is.
  

• Representatives Frank Pallone (D-NJ), Cathy McMorris Rodgers (R-WA), and Roger Wicker (R-MS) introduced the bipartisan American Data Privacy and Protection Act, the most significant federal proposal on data privacy to date. While the bill shows progress in establishing federal privacy protections, children’s privacy advocates are concerned. In particular, advocates fear that the bill’s protections aren’t strong enough for kids and that, if passed, the measure would override state-level efforts, including the proposed California Age Appropriate Design Code Act, which is more robust in its protections for children under the age of 18.
  

• Jonathan Haidt joins the Lex Fridman Podcast to discuss The Case Against Social Media. In the 1.5 hour session, Haidt uses research to rebuke some of the claims Mark Zuckerberg made in his recent interview on Fridman. Main takeaway: social media is harmful, and changes aren’t going to come with platform tweaks. We need to fundamentally change the incentive structures of platform business models.