In order to be more competitive, today’s successful lawyer must provide the client with timely and accurate information. In the past an attorney might dictate a letter or report to the client which would then be placed in line for transcription by the secretary and might take up to a week before it was ultimately out the door and on it’s way. This is no longer acceptable to most clients who demand a report letter within 24-48 hours of a significant event. How can this challenge be met? One solution is implementation of voice recognition technology.

Great strides have been made in the past several years in the software products we have come to know as voice or speech recognition. It is no longer a simple dictation program but a much more complex and advanced process which takes the spoken word and translates it into a finished document through a series of formatting, editing, and spell checking functions with accuracy rates of 95% and higher. Today’s software is capable of sending the finished document, via e-mail or through the firm’s network, to the attorney’s secretary or to the client as well as to a printer. These systems can be programmed to understand the user by direct voice input or downloading the user’s sample documents into the program from diskette. The program will automatically create a vocabulary for the way you write. Most programs come packed with a legal vocabulary which includes classic and specialized legal terms, court names, case history, weight of authority, Latin and French law phrases, litigation documents, hyphenates, reporters, abbreviations and constructions. Supplements may be added as needed.

Voice recognition systems become "smarter" the more they are used. The software learns to recognize tone and inflection and increases it’s dictionary with each new word or phrase used. Dictated paragraphs or sections can be reviewed and revised using simple mouse or voice commands. Marcos, speech commands that allow you to enter boilerplate or execute complicated commands, are also available. The system also helps by suggesting alternatives when it recognizes a word incorrectly.

It can take several weeks to train any voice recognition program, however it is not unusual to be dictating effectively at more than 90 words per minute after one week.

Minimum system requirements for these programs are; 133 MHz Pentium Processor, IBM-compatible PC Windows 95 (32 MB memory) or NT 4.0 (48 MB memory), hard disk, 200 MB, and CD ROM for installation, 16-bit sound card or built-in audio system.

Although these systems are not yet 100% accurate, it is my belief that with the strides made over the past 5 years, we should see a 99%+ accuracy rate soon. Firms in the local Knoxville area using or experimenting with voice recognition systems are still in the minority, however most administrators and those attorneys with more than a passing interest in technology are watching and waiting for that product which will make our firms more productive.

© Copyright 2004 ALA-Knoxville Chapter. All Rights Reserved.
Questions or comments about this site should be directed to

 The Web Master

DISCLAIMER: The Knoxville Chapter of the Association of Legal Administrators is a separate legal entity from the Association of Legal Administrators (ALA). ALA licenses the use of its name, mark, logos and other protected properties to chapters which are in good standing. ALA disclaims all liability or responsibility whatsoever for the actions, representations and liabilities of the Knoxville Chapter, specifically including those of any nature whatsoever arising from or out of the content of the other features related to the Knoxville Chapter Web site.  In no event shall the ALA be deemed the guarantor of the Knoxville Chapter.