904 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1946 
Anyone who considers carefully the relation between aural and 
visual perception is likely to question the need for using a three- 
dimensional form of pattern for visible speech. The ear hears two- 
dimensionally in frequency and intensity. The time dimension is sup- 
plied by the memory. Why not then show only a frequency-intensity 
speech pattern and make similar use of the visual memory? The ques- 
tion is a perfectly valid one and has received considerable thought 
during the investigation under discussion. In fact, an experimental 
study of two-dimensional patterns is being carried on simultaneously 
with the three-dimensional studies. At the present time it is confined 
to visual discrimination tests of various two-dimensional displays. 
Earlier, one girl studied a particular form of pattern for a few months, 
but when the discrimination test methods were developed this train- 
ing was discontinued in order to concentrate the available effort upon 
the more fundamental aspects of two-dimensional display. 
Both the limited training and the visual discrimination tests made 
so far seem to indicate that it will be more difficult to read the two- 
dimensional patterns than it is to read the three-dimensional type. 
Three-dimensional speech patterns are analogous to print moving from 
right to left on a telegraph tape, while two-dimensional patterns are 
analogous to seeing this moving print through a narrow slit, only as 
wide as the lines that form the letters. No doubt we could learn to read 
print moving past such a narrow slit, but it would rather obviously be 
more difficult than unrestricted reading because we normally perceive 
whole words rather than bits of letters. It may well be that this man- 
ner of reading is a result of the requirement that the eye be focused 
upon the print in order to obtain a satisfactory memory impression. 
Although the focusing requires visual effort, large pattern areas may 
be recorded in a single “exposure” so that the effort is not excessive. 
But to record patterns a bit at a time, or two-dimensionally, by the 
visual process would require almost continual concentration and should 
therefore exact more effort for the same accomplishment. 
It is possible then even though the eye should in theory be able to 
understand two-dimensional] patterns of speech it may be inefficient for 
this purpose. 
In addition to the problem of understanding others, deaf and 
severely deafened persons are faced with the problem of controlling 
their own speech. Making sounds without ears to hear them is some- 
what like drawing pictures with no eyes to see the results. When 
the hearing is largely absent in a child, it is necessary to teach speech 
without normal control exercised by the hearing. Even in cases where 
a vocabulary is developed before severe loss of hearing sets in, there 
is a gradual deterioration of the speech. In both cases visible speech 
in the forms described here should be of considerable help in speech 
