The experts do not agree on whether Denyse DuBrucq's tactile speech simulator will enable deaf persons to understand speech without reading lips. But they do agree that, if it does work, it will represent what one Gallaudet College researcher called "an outstanding accomplishment." uBrucq has formed five entities to perform research and to produce, sell and export the simulator -- called the TactilEar -- and its associated instructional devices. The business and legal end of the operations seems as complex as the scientific and technical end.

She says that the tactile speech simulator planned by her Ear Three Systems Manufacturing Co. of Vienna will translate speech into recognizable vibrations on the wrist of anyone wearing the cuff-like device.

The keys to the TactilEar are a speech analyzer, metal strips that bend when exposed to low voltage and low current and that have pegs on one end to make contact with the user's skin, and a code developed by DuBrucq.

The analyzer will break speech into its component sounds and tranform these sounds into DuBrucq's code, activating different combinations of metal strips so that the user feels "touches" in varying patterns.

Four contact points in one row on the top of the wrist are for consonants, and four in a parallel row are for vowels. Three parallel rows of three points each on the underside of the wrist reflect what the inventor says are the nine mouth forms used to make sounds.

So, for example, the sound of the letter M would activate all four of the consonant points in a vertical row as well as three mouth-form points in a diagonal from lower left to upper right.

The speech analyzer was invented by Motorola Inc. engineer John Marley. Motorola sells it to domestic and foreign automakers, who use it as part of a computerized mechanism that makes engine adjustments every 5 milliseconds in many 1982 models.

Marley retains the rights for speech applications and will produce an analyzer for Ear Three Systems. DuBrucq expects his first prototype early in 1983 and production models in March.

At this point, Marley's analyzer has not been connected to DuBrucq's prototype TactilEar, which consists of the metal strips housed in two shallow, open-face plastic boxes with what looks like a nylon watch strap.

DuBrucq won't be able to demonstrate that the device creates tactile patterns that people can be taught to recognize until the analyzer is programmed properly, and attached to the strips, microphone and 9-volt battery that supplies the power.

But there are skeptics, partly because this inventor has been attempting for less than a year to accomplish what others have been working on unsuccessfully for as long as 30 years. They also have voiced the fear that announcement of DuBrucq's plans could raise false hopes.

DuBrucq grew up in Milwaukee, specialized in biology and genetics in graduate school, and earned a doctorate in science education. When she was teaching at Gallaudet's Model Secondary School for the Deaf, two of her students had Ushers Syndrome -- whose victims not only are are born deaf, but also lose their vision during adolescence. Their needs generated the idea of tactile stimulation. She began development of the TactilEar after leaving Gallaudet.

Dr. James Pickett, who is engaged in tactile research at Gallaudet, cautioned that his experiences and knowledge of others' research leads him to believe that DuBrucq will "run into trouble once she tries to train people" to recognize the tactile patterns comprised by her code. He said that researchers have produced "pretty good results" only when the number of speakers and their vocabularies are limited.

But he added, "I think the particular approach" that DuBrucq is taking for the coding "is a good one."

"Eleven years ago, I set out to use vibratactile tactics to do the same thing" that DuBrucq is attempting, but did not succeed and switched to the visual approach, recalled Dr. R. Orin Cornett, chairman of the Center for Studies in Language and Communication at Gallaudet. He is perfecting a visual aid for lip readers called the Autocuer, which DuBrucq says gave her a starting point for her own work.

In expressing skepticism about DuBrucq's eventual success, Cornett noted that he was using a more sensitive part of the body--the finger tips--in his unsuccessful tactile research. e said the difficulty in perceiving vibrations as the number of contact points increases has stumped him and two Princeton University researchers who have been working on the problem for 30 years.

One of them, Dr. Carl Sherrick, cited three problems in tactile stimulation, including obtaining a device such as a speech analyzer that can break speech into "the proper segments," and getting a sufficiently light transducer "to excite the skin mechanically or electrically.

"The third is the really big problem," Sherrick said: Can a person clearly distinguish tactile patterns such as those that constitute DuBrucq's code? He believes that, because "intensive-enough training yields surprising results . . . nobody can precisely foretell" what kinds of patterns might work. Sherrick has agreed to sit on the board of DuBrucq's nonprofit research foundation.