In this paper, we trace the biases in current natural language processing (NLP) models back to their origins in racism, sexism, and homophobia over the last 500 years. We review literature from critical race theory, gender studies, data ethics, and digital humanities studies, and summarize the origins of bias in NLP models from these social science perspective. We show how the causes of the biases in the NLP pipeline are rooted in social issues. Finally, we argue that the only way to fix the bias and unfairness in NLP is by addressing the social problems that caused them in the first place and by incorporating social sciences and social scientists in efforts to mitigate bias in NLP models. We provide actionable recommendations for the NLP research community to do so.

Randomized neural networks are immortalized in this AI Koan: In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6. Why is the net wired randomly?'' Sussman replied,I do not want it to have any preconceptions of how to play.'' Minsky then shut his eyes. Why do you close your eyes?''

In celebration of New York Magazine's 50th anniversary, this series, which will continue through October 2018, tells the stories behind key moments that shaped the city's culture. In the summer of 1988, the hedge-fund manager Donald Sussman took a call from a former Columbia University computer-science professor wanting advice on his new Wall Street career. "I'd like to come see you," David Shaw, then 37 years old, told Sussman. Shaw had grown up in California, receiving a Ph.D. at Stanford University, then moved to New York to teach at Columbia before joining investment bank Morgan Stanley, which had a new secretive trading group that was using computer modeling. A neophyte in the ways of Wall Street, Shaw wanted Sussman, who founded the investment firm Paloma Partners, to look at an offer he had received from Morgan Stanley's rival, Goldman Sachs.

Randomized neural networks are immortalized in this AI Koan: In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6. I am training a randomly wired neural net to play tic-tac-toe,'' Sussman replied. Why is the net wired randomly?'' Sussman replied, I do not want it to have any preconceptions of how to play.'' Minsky then shut his eyes.

In 1969, William H. "Holly" Whyte decided to analyze, and eventually decode, New York City's rambunctious street life. A famed author, Whyte, along with a handful of collaborators, was recruited by the city's planning commission to set up cameras and surreptitiously track human activity. Whyte and his team spent countless afternoons filming parks, plazas, and crosswalks, and even more time counting, crossing out, analyzing, and quantifying footage. Notations were made for how people met and shook hands. Pedestrian movement was mapped on pads of graph paper. To get accurate assessments of activity at a street corner, Whyte's researchers manually screened people caught waiting for lights to change. Imagine how much time it took to figure out that at the garden of St. Bartholomew's Church, the average density at lunch time is 12 to 14 people per 1,000 square feet.

Four Harvard Law students have their heads in the cloud--and they think the rest of the legal profession should join them. With their powerful new search engine called Evisort that harnesses cloud storage and artificial intelligence, they hope to revolutionize the costly and labor-intensive way that lawyers currently handle contracts and other transactional work, liberating them for more creative and interesting tasks. Developed by the students over the past two years, Evisort is "like Google for legal contracts," says Jerry Ting '18, co-founder and CEO, who came up with the idea as an undergraduate. While artificial intelligence is the cutting-edge of automating labor-intensive tasks such as document review, it hasn't yet been widely applied to contracts. Evisort jumps into that gap by enabling lawyers to quickly sort through thousands of contracts and other documents to unlock key insights for transactional work.

Structure and Interpretation of Computer Programs has been MIT's introductory pre-professional computer science subject since 1981. It emphasizes the role of computer languages as vehicles for expressing knowledge and it presents basic principles of abstraction and modularity, together with essential techniques for designing and implementing computer languages. This course has had a worldwide impact on computer science curricula over the past two decades. The accompanying textbook by Hal Abelson, Gerald Jay Sussman, and Julie Sussman is available for purchase from the MIT Press, which also provides a freely available on-line version of the complete textbook.

The strategy has been implemented in a simple QLISP program.

Plan modification *The research reported herein was sponsored by the National Science Foundation under Grant GJ-36146.