This article appeared in the August 2, 1993 issue of Forbes Magazine, shortly before the provisions of the Television Decoder Circuitry Act took effect. It was written by Joseph R. Garber.
I’m not a fan of the way he described the ADA (Americans with Disabilities Act) and TDCA (Television Decoder Circuitry Act), as I’m a big supporter of both bills. Despite that, I thought the article was a good one.
It takes four to five years of rigorous schooling to become a court reporter, which is one reason graduates can command $50,000 to $100,000 annually.
Nowadays personal computer software can understand the spoken word and translate speech into digital commands. Does that mean PCs will put court reporters out of business? Not likely. In these litigious times, legal transcription is a growth industry that consumes all the talent and all the technology available. Far from replacing court reporters, therefore, PCs will make them more productive.
What can stenographers do that computers can’t? Even the brainiest computer has–and will have, for decades to come–problems determining whether a witness’ mumbled grunt translates as “uh-huh” or “uh-uh.” Nor, for a long time to come, will computers be able to sort out a dialog when, as is common in courts and in depositions, several people speak at the same time. Then there are the witnesses with heavy accents, speech impediments or nearly inaudible voices. Sure, someday computers may be able to understand them-but will they be able to sense what a nodding head means, or remember that the witness pointed to the accused as being the man seen at the scene of the crime?
Probably not. But what the PC is doing is increasing the back-office productivity of the 40,000 or so U.S. court reporters, in many cases replacing the reporter’s assistant. It isn’t replacing the professional who sits near the witness box.
Stenographers were among the earliest users of office automation equipment. First patented in the 1800s and largely unchanged in either keyboard or platen design since the 1930s, the standard stenotype machine costs $3,000. The court reporter depresses its 22 keys (typically 4 to 6 of them simultaneously) to generate, at about 250 words per minute, a mostly phonetic record of what was spoken. Extra letters are added to distinguish homophones, like “they’re” and “their.” A starter steno’s dictionary has about 20,000 words; reporters will expand it with abbreviations suited to the proceedings at hand-a shorthand for “cardiotoxicity” or “Milken,” for example.
Back in the old days, before stenotype machines came with a floppy disk, a megabyte of memory and PC interface, a court reporter’s real work began after the typing was over. That work consisted of translating a day’s worth of phonetic gibberish into something a lawyer could read. The chore took hours, and leaned heavily on stenographers’ assistants.
Nowadays the technology turns 200 pages of stenotype phonetic code into an English transcript in 14 seconds. Needless to say, such magic comes at a price. Cheetah Systems, a privately held firm in Fremont, Calif., charges $10,000 for its turnkey transcription system.
Cheetah’s chairman, Gary Robson, made his first killing in 1986 at age 29, by selling a microelectronics design shop he and his brother owned to Imperial Chemical Industries. Looking around for an encore, he listened to his wife, a court reporter, describe what she needed in a stenography machine. In 1987, Cheetah was born. It was, in Robson’s words, “the eleventh company in a ten-company industry.” Now Cheetah is number three, and gaining fast on numbers one and two, Quixote Corp. and Xscribe.
Why the quick success? Robson says his software was the first in 1987 to provide speedy translation on a PC platform, a 24-hour-a-day hotline for its customers (they do a lot of transcribing at night) and such useful features as automatic punctuation. With about a million lines of “C” and assembly language code, Cheetah’s package purrs on a standard 386 or 486 PC.
The Robson clan didn’t stop there. The threesome were quick to spot the implications of the costly and bureaucratic Americans with Disabilities Act and Television Decoder Circuitry Act. Starting in July of this year, every new 13-inch or larger television sold in America must have caption decoding chips in it. Most broadcasters will start captioning all their programs. Most corporations, or at least those wanting to avoid Disabilities Act suits by their hard-of-hearing employees, will start captioning training videos and video conferences.
Thus Cheetah, building on its powerful court reporting software, developed a $15,000 package for television captioning, bringing it to market well ahead of its competitors. The stenocaptioner listens to a show, phonetically transcribing what he or she hears into a Cheetah system. The phonetic symbols are then translated in real time by the computer into English and automatically fed into the broadcast signal. Cheetah, still number three in the court reporter market, is number one, with a 60% share, in the faster-growing video captioning business.
Cheetah has been a boon to court reporters, opening to them a whole new line of business; reporters, after all, are the only people capable of producing on-screen transcriptions of the rapid-fire dialog during live television events such as sports and news, or during business meetings and video conferences. A former court reporter turned freelance business captioner is delighted with her new career. “No lawyers,” she explained.
Gary D. Robson 