Some may have read one of several articles from various outlets such as the New York Times, Philly, or Trib that, in summary, basically stated there’s a study that will be published in the Language journal. This study took a couple dozen volunteers and had them transcribe recorded statements that the news articles described as black dialect, but which I have also heard referred to described as urban English or street talk. The volunteers did not do well, and there was a high percentage of inaccuracy.
NCRA and PCRA responded to these articles openly. They basically called the title of the article(s) provocative, and pointed out that this involved volunteers taking down recorded statements as opposed to a live courtroom setting. They seem to believe as I do, that this study shows the desperate need for highly trained stenographers in courts.
Succinctly, there are two things the news often gets wrong: Law and science. We are talking about both at the same time, so I’m willing to go out on a limb and say that there are probably some inaccuracies or misconceptions about the actual study. Anecdotally, I look back to correspondence I’ve had with a researcher of a concept coined benevolent sexism. I actually wrote Jin Goh, a researcher in that study, and Jin Goh basically said we need more studies for the study to be conclusive, and the media misrepresented the study. The results were interesting and honest, but they did not mean that being nice to people was sexist as some news reported.
So of course the first step in understanding a study would be to read it and decide if reaching out to its authors is necessary. So I reached out to the group that publishes the Language journal only to learn that the study itself isn’t going to be published until June 2019.
So what can we say at this juncture? The study had a fairly small sample size, as best I can tell, of 27 volunteers. I am not entirely sure if those volunteers were stenographic reporters as of writing. This isn’t uncommon. We are in a world with finite resources and funding, so studies often don’t have the kind of money you’d need to reach conclusive results on any one topic over the course of one study. We can also note that this is, as best as I can tell, one of the first studies of its kind, so while it has interesting results, we need to remind ourselves these results are not, to our knowledge, representative of multiple studies over years. We need to remind ourselves that some researchers showed us months ago that studies and the news stories about them can be worded in a way that gets clicks, but not in a way that informs the reader. For example, a study was recently conducted that showed jumping out of a plane without a parachute does not increase your chance of injury. The catch? They jumped out of a parked plane. They did this intentionally to remind us all that news on studies can be slanted or cause misconceptions.
So what should we do? In my opinion, there’s only one thing to do. Open our eyes to the fact that a scientific study was conducted, and it’s apparent that humans mishear things a lot! Continue to adapt to different accents in our training and work. Continue to push to provide the best service possible to all lawyers, litigants, and caption consumers. I do think there’s a lot to be said for our performance in real courtroom settings. We ask for repeats all the time to make sure that what people say is honestly and accurately reflected, and that’s something they probably couldn’t get into a lab or study as easily. Perhaps with time we could even conduct our own study, and maybe it would find that the stenographer mishears less than the average person. Perhaps we’d find our hearing is average. We don’t know. That’s the point of studies.
Bottom line: Don’t let this thing ruffle your feathers. I saw a lot of reporters spew a lot of vitriol over the articles, and in the end, the theme of the articles were not “stenographers are bad,” but more “humans mishear things and we should be mindful of this in the administration of our courts because if the transcribers aren’t hearing it then it is likely the lawyers and judges aren’t either.” We’re good at what we do, and we’re better off proving that than attacking linguists on Twitter. We are better off making sure our service is the best service lawyers and litigants can find, period. Truthfully, researchers give us valuable insight into what we do, but it is we who perform every day who know what’s at stake for the lawyers, litigants, and judges we serve.
As an aside, I understand the verbiage of the headlines upset some readers and I agree that this all could’ve been written more artfully. I myself have used descriptors to try to explain the issue as it is and make it more clear for anyone that cares to read.
UPDATE MARCH 5, 2019:
I am very excited to say another article was released which published the linguist’s name, Taylor Jones. Taylor Jones’s site has a lot of very specific examples that I think are eye opening and important for everyone to read and understand, including examples like, “when you tryna go to the store?” I am delighted to have come across Jones’s website and work, and will be reaching out for comment and clarification on this study to understand exactly what it is about and how reporters might improve training. Previously, I believed I’d have to wait until June to see the study. At a glance, according to a January 2019 blog post by Jones, it does appear that they utilized and/or surveyed Philadelphia court reporters that were actually working in the courts. It is stated that evaluated sentence-by-sentence, accuracy was just under 60%, and when evaluated word-by-word, accuracy was about 82%. Without having yet received comment from Jones, I can say I am incredibly impressed by the blog, and anyone with interest in this study and developing better verbatim records should definitely swing by it and read some of the stuff there. At first glance, this really may be more of an issue than I had believed, and I’d encourage every reader to keep an open mind. Notably, Jones states he has worked with Culture Point to come up with a training suite to address this issue.
April 2, 2019:
In order to be subscriber-friendly I have attached all future updates on this to a new a blog post.