header "The Captioning Process"

Why Isn't It Always Perfect?
""
Home | Next Page | Back | Site Map

This is the best question of all!  When you are viewing realtime captions, it’s hard to imagine that a person could possibly have the skill to write as fast as people are speaking, identify who is speaking, and use that little stenotype machine with 22 keys to represent ALL oral speech.
 Let me give you a quick lesson to show you exactly how this works.

The stenotype keyboard is composed in such a manner that the court reporter uses the left hand to begin the sound, the thumbs to insert vowels, and the right hand to end the sound – all simultaneously as the speech is heard.  

graphic of stenotype keyboard 

If you look at this sample stenotype keyboard, you can see keys on the left, the vowels at the bottom, and keys on the right.  

Now, for example, if you were to write the word cat, since court reporters write phonetically, they would place their left ring finger over the K, their left thumb over the A and their right little finger over the T, and all exactly at the same time strike those keys, and KAT would read out on their paper.  That is simple enough to understand, isn’t it?  If you were in class, the teacher would tell you to find the K, find the A, and find the T, and when you had your fingers positioned above all three, strike the keys.  (Eventually you would get faster – a heck of a lot faster!)

Well, now you might look at the keyboard and realize there are a lot of keys missing to create beginning, vowel and ending sounds.  

So what’s a court reporter to do?  They use combinations of keys.  Let me just explain one word that aptly demonstrates this concept. The word gleam.  You will notice there is no G on the left side.  And there is no L.  And on the right side, the side that ends the sounds, there is no M.  So gleam is written  TKPWHRAOEPL  and that’s exactly how it prints on the court reporter’s paper.  The court reporter can read that back, and the computer can translate that back into English because:  TKPW = G, and HR = L, and AOE = a long “e” sound, and PL = a final “m” sound.  Isn’t that easy to understand?  The court reporter strikes all those keys all at once.  It takes no more time than striking KAT.  

Now on to my next point.  Why are there sometimes mistakes?  
First, if the court reporter’s fingers were to miss one letter -- just one essential letter in that word gleam  -- then the translation would not work and you would either get a word that doesn’t translate or an incorrect word.  Consider that court reporters are often writing at 200 plus words per minute.  Occasional mistakes will occur.  Court reporters are generally certified at anywhere between 95 and 97.5% accuracy.  That’s pretty good considering most do a lot better than that and the volume of words that they capture at any given time.  
Secondly, the captioner doesn’t have the luxury of asking someone to repeat what they said or interrupt if they can’t quite hear something. Court reporters are human, and it's actually a darn good thing that they bring human reasoning to their skills.  

Consider the following:

 2 B oar naught two bee

Two bee ore knot 2 “B”

Too be or not two bee

 To be or not to be

Now, that's only six little English words!  Can you see just how difficult capturing spoken language accurately can be? 

 Understanding Live Captioning!  

Weer Not Bad Spelerz - by Patty White

 

""

web author Vykki Morgan, 
Associate Professor, Cerritos College
page last updated: 05/17/10

Home | Next Page | Back | Site Map