Siri, A Court Reporter? You Can’t Be “Sirious” !

by admin on January 29, 2014

siriBy:  Jan Ballman – RPR, CMRS

Siri.   Google Voice.  Hands-free technology.  These advances are both impressive and appreciated, even by Court Reporters and Captioners who make their living in the voice-to-text arena, for certain conveniences they provide.  I guess it’s no surprise, though, that since their emergence we are often asked:  “Why haven’t Court Reporters and Captioners been replaced by voice recognition?”  “Aren’t you worried about your future?”  “Is court reporting a dying profession, soon to be replaced by technology?”  Good questions, for sure.  But if we don’t look worried, it’s because we’re really not.   And that’s because we are confident that Siri is not, and likely never will be, good enough to be a Court Reporter or Captioner, because Siri lacks a critical component absolutely vital to doing our job.   Siri, like Scarecrow from the Wizard of Oz, needs a brain to be a Court Reporter.

The following answer, taken from the deposition of a neurologist, perhaps says it best:

 A.   “Our brains are a miracle.   Look at the Court Reporter here as an example, okay?  This is a miracle in progress, happening right before your eyes.  Let me explain what’s happening here.   I am speaking, so the information needs to come in through her ears, go through her temporal lobe, and then it has to go log itself into the language center.   She has to be able to comprehend what I’m saying, then it gets rerouted to the prefrontal cortex where she has to hold the information, because I’m talking faster than she is typing, okay?   Then she has to analyze it, integrate it, and synthesize it.   Then it has to go back to the cerebellum to be converted into those little symbols.   Have you ever seen them?   She is converting what I’m saying into a different language.  And the white matter tracts allow her to reroute all of this information accurately and simultaneous, seemingly without effort.  Seamlessly.  Okay?   That is why I believe Court Reporters will never be replaced.   Because no technical – no technology could ever replace the beauty and the miracle of that human brain.”

Today’s voice recognition, predictive text and word association technologies are based on statistical guesses.   According to James Kendrick, a Speech Recognition Specialist trained by IBM, who has been studying speech recognition technology for over ten years, we haven’t been able to advance much further than 90 percent accuracy,  and that’s leveraging  best-case scenario environments:    No background noise, only one person speaking, and utilizing high-quality microphones and noise-cancelling equipment.

Ninety percent will never cut it in our business.

Court Reporters execute at accuracy levels between 98 and 100 percent.  The difference in translating at 80-90 percent accuracy and 98 to 100 percent is huge when it comes to output, and it can be attributed to the human brain’s ability to comprehend versus simply hearing.   Comprehension facilitates determining what was said versus guessing, “Is this what you meant?”   You need only visit sites such as to appreciate how embarrassing guessing and getting it “almost right” can be.

Let’s delve a little deeper into how that human brain creates a superiority gap between Court Reporters and voice recognition technology.   When Court Reporters listen, be it to a mumbler or a babbler, to a foreign witness or a forensic pathologist, to a brain surgeon or a rocket scientist, we may have difficulty “hearing” whether the witness said “a” or “the”, “pan” or “can”, “Tim” or “Jim”, “Mary” or “Carrie,” let alone “marry,”  “carry,” “ankylosing spondylitis” or “apophysitis calcaneus,”  but our brains help us make the final determination, almost instantaneously, using comprehension derived from context, knowledge, and prior life experience.   Sometimes, when we simply did not hear nor comprehend what was said, we will interrupt and obtain immediate clarification, helping to ensure seamless accuracy.   As well, a lot of words sound alike but mean different things.   We need to know whether to write whether or weather, abhorrent or aberrant, epitaph or epithet, martial or marshal…

Accents, ambulance sirens, coughing and sneezing, papers shuffling,  glassware clanking, extemporaneous conversations, rapid-fire arguments, three people talking at once—these are everyday (more like every minute), real-life challenges Court Reporters face on every job.  But thanks to the amazing capacity of the human brain, we can filter noise, read lips, and even listen to one person talking while writing what another just finished saying at accuracy levels that still do and always will make Siri blush in a competence comparison.

Said another way, when it’s important and it has to be right, technology alone is still not good enough.   A text message?  No big deal.  A court transcript in a murder trial or a multimillion dollar lawsuit?  Way bigger deal.  And this holds true before we even begin to touch on other obvious problems with technology…like the fact it doesn’t know when (and therefore doesn’t tell us) it’s failing.  A Court Reporter will arrive early, have backup equipment at the ready, and will see and know immediately if his/her equipment is not working properly or begins to fail mid-job, and will take appropriate steps to immediately protect the record by proactively stopping the proceedings to making the necessary adjustments to ensure every word is captured and preserved.

So, Siri, we’ll concede your reign in certain arenas—places where typos and misfires are just not that big of a deal (raise your hand now if you’ve never found your car dialing Dale when you wanted to call Gail).  But when it comes to the important work Court Reporters do, you’re not even in the same arena, let alone league.

Say “Good Night,” Siri…



Well said, however, we argued this for years and years in Minnesota, now the courtroom is 100% electronic and we are in a different profession! Money is #1 everywhere and we are the only ones who care.
Don White
I was amazed when my son let me play with Siri on his phone, especially the quick and sometimes witty responses; but having been a verbatim court reporter for fifty years in many arenas, hearing reporter, freelance, grand jury, federal court, state court; and totally enjoyed the exposure, a true window on the world, the inside of courts, government agencies, ever changing language and law in both criminal and civil litigation, at times felt amazed that this field, although demanding difficult work, daily deadlines, and realtime reporting, which had to be at least 90% accurate, and the finished transcript 100 % accurate, was well-paid and provided excellent opportunities for actually physically seeing the world through work; hearing and reporting knowledge in every possible field of endeavor; after reading the opinion of the neurologist, agree that only humans can do what court reporters do with incredible accuracy under such a variety of circumstances, and will always be extremely proud to be a verbatim reporter.
I love this piece!!! Thank you for taking the time to put it together!
Tricia Weinberg
Great piece, Jan. I have shared it with the other reporters here.

Previous post:

Next post: