<s>
Speech-generating	B-Application
devices	I-Application
(	O
SGDs	O
)	O
,	O
also	O
known	O
as	O
voice	B-Application
output	I-Application
communication	I-Application
aids	I-Application
,	O
are	O
electronic	O
augmentative	O
and	O
alternative	O
communication	O
(	O
AAC	O
)	O
systems	O
used	O
to	O
supplement	O
or	O
replace	O
speech	O
or	O
writing	O
for	O
individuals	O
with	O
severe	O
speech	O
impairments	O
,	O
enabling	O
them	O
to	O
verbally	O
communicate	O
.	O
</s>
<s>
Speech-generating	B-Application
devices	I-Application
can	O
produce	O
electronic	O
voice	O
output	O
by	O
using	O
digitized	B-General_Concept
recordings	I-General_Concept
of	O
natural	O
speech	O
or	O
through	O
speech	B-Application
synthesis	I-Application
—	O
which	O
may	O
carry	O
less	O
emotional	O
information	O
but	O
can	O
permit	O
the	O
user	O
to	O
speak	O
novel	O
messages	O
.	O
</s>
<s>
The	O
first	O
known	O
SGD	O
was	O
prototyped	O
in	O
the	O
mid-1970s	O
,	O
and	O
rapid	O
progress	O
in	O
hardware	B-Architecture
and	O
software	B-Application
development	O
has	O
meant	O
that	O
SGD	O
capabilities	O
can	O
now	O
be	O
integrated	O
into	O
devices	O
like	O
smartphones	B-Application
.	O
</s>
<s>
Speech-generating	O
systems	O
may	O
be	O
dedicated	O
devices	O
developed	O
solely	O
for	O
AAC	O
,	O
or	O
non-dedicated	O
devices	O
such	O
as	O
computers	O
running	O
additional	O
software	B-Application
to	O
allow	O
them	O
to	O
function	O
as	O
AAC	O
devices	O
.	O
</s>
<s>
Speech	O
output	O
possibilities	O
included	O
both	O
digitized	O
and	O
synthesized	B-Application
speech	I-Application
.	O
</s>
<s>
Rapid	O
progress	O
in	O
hardware	B-Architecture
and	O
software	B-Application
development	O
continued	O
,	O
including	O
projects	O
funded	O
by	O
the	O
European	O
Community	O
.	O
</s>
<s>
The	O
first	O
commercially	O
available	O
dynamic	O
screen	O
speech-generating	B-Application
devices	I-Application
were	O
developed	O
in	O
the	O
1990s	O
.	O
</s>
<s>
Software	B-Application
programs	O
were	O
developed	O
that	O
allowed	O
the	O
computer-based	O
production	O
of	O
communication	O
boards	O
.	O
</s>
<s>
High-tech	O
devices	O
have	O
continued	O
to	O
become	O
smaller	O
and	O
lighter	O
,	O
while	O
increasing	O
accessibility	O
and	O
capability	O
;	O
communication	O
devices	O
can	O
be	O
accessed	O
using	O
eye-tracking	B-General_Concept
systems	I-General_Concept
,	O
perform	O
as	O
a	O
computer	O
for	O
word-processing	B-General_Concept
and	O
Internet	O
use	O
,	O
and	O
as	O
an	O
environmental	O
control	O
device	O
for	O
independent	O
access	O
to	O
other	O
equipment	O
such	O
as	O
TV	O
,	O
radio	O
and	O
telephones	O
.	O
</s>
<s>
Direct	O
access	O
methods	O
involve	O
physical	O
contact	O
with	O
the	O
system	O
,	O
by	O
using	O
a	O
keyboard	O
or	O
a	O
touch	B-Device
screen	I-Device
.	O
</s>
<s>
Users	O
accessing	O
SGDs	O
indirectly	O
and	O
through	O
specialized	O
devices	O
must	O
manipulate	O
an	O
object	O
in	O
order	O
to	O
access	O
the	O
system	O
,	O
such	O
as	O
maneuvering	O
a	O
joystick	B-Device
,	O
head	O
mouse	B-Device
,	O
optical	O
head	O
pointer	O
,	O
light	O
pointer	O
,	O
infrared	O
pointer	O
,	O
or	O
switch	O
access	O
scanner	O
.	O
</s>
<s>
With	O
direct	O
selection	O
a	O
body	O
part	O
,	O
pointer	O
,	O
adapted	O
mouse	B-Device
,	O
joystick	B-Device
,	O
or	O
eye	B-General_Concept
tracking	I-General_Concept
could	O
be	O
used	O
,	O
whereas	O
switch	O
access	O
scanning	O
is	O
often	O
used	O
for	O
indirect	O
selection	O
.	O
</s>
<s>
Speech-generating	B-Application
devices	I-Application
generally	O
display	O
a	O
set	O
of	O
selections	O
either	O
using	O
a	O
dynamically	O
changing	O
screen	O
,	O
or	O
a	O
fixed	O
display	O
.	O
</s>
<s>
Word	O
prediction	O
software	B-Application
may	O
determine	O
the	O
choices	O
to	O
be	O
offered	O
based	O
on	O
their	O
frequency	O
in	O
language	O
,	O
association	O
with	O
other	O
words	O
,	O
past	O
choices	O
of	O
the	O
user	O
,	O
or	O
grammatical	O
suitability	O
.	O
</s>
<s>
Another	O
approach	O
to	O
rate-enhancement	O
is	O
Dasher	B-Language
,	O
which	O
uses	O
language	O
models	O
and	O
arithmetic	B-Algorithm
coding	I-Algorithm
to	O
present	O
alternative	O
letter	O
targets	O
on	O
the	O
screen	O
with	O
size	O
relative	O
to	O
their	O
likelihood	O
given	O
the	O
history	O
.	O
</s>
<s>
Dynamic	O
displays	O
devices	O
are	O
usually	O
also	O
touchscreen	B-Device
devices	O
.	O
</s>
<s>
The	O
output	O
of	O
a	O
SGD	O
may	O
be	O
digitized	O
and/or	O
synthesized	O
:	O
digitized	O
systems	O
play	O
directly	O
recorded	O
words	O
or	O
phrases	O
while	O
synthesized	B-Application
speech	I-Application
uses	O
text-to-speech	B-Application
software	I-Application
that	O
can	O
carry	O
less	O
emotional	O
information	O
but	O
permits	O
the	O
user	O
to	O
speak	O
novel	O
messages	O
by	O
typing	O
new	O
words	O
.	O
</s>
<s>
Today	O
,	O
individuals	O
use	O
a	O
combination	O
of	O
recorded	O
messages	O
and	O
text-to-speech	B-Application
techniques	O
on	O
their	O
SGDs	O
.	O
</s>
<s>
SGDs	O
that	O
use	O
synthesized	B-Application
speech	I-Application
apply	O
the	O
phonetic	O
rules	O
of	O
the	O
language	O
to	O
translate	O
the	O
user	O
's	O
message	O
into	O
voice	O
output	O
(	O
speech	B-Application
synthesis	I-Application
)	O
.	O
</s>
<s>
Smartphones	B-Application
and	O
computers	O
have	O
increased	O
the	O
use	O
of	O
synthesized	B-Application
speech	I-Application
devices	O
through	O
the	O
creation	O
of	O
apps	O
that	O
allow	O
the	O
user	O
to	O
select	O
from	O
a	O
list	O
of	O
phrases	O
or	O
messages	O
to	O
be	O
spoken	O
in	O
the	O
voice	O
and	O
language	O
that	O
the	O
user	O
has	O
chosen	O
.	O
</s>
<s>
or	O
Assistive	O
Express	O
for	O
iPhone	O
provide	O
a	O
cheap	O
way	O
to	O
use	O
a	O
speech-generating	B-Application
device	I-Application
without	O
having	O
to	O
visit	O
a	O
doctor	O
's	O
office	O
or	O
learn	O
to	O
use	O
specialized	O
machinery	O
.	O
</s>
<s>
With	O
synthesized	B-Application
speech	I-Application
there	O
is	O
virtually	O
unlimited	O
storage	O
capacity	O
for	O
messages	O
with	O
few	O
demands	O
on	O
memory	O
space	O
.	O
</s>
<s>
Synthesized	B-Application
speech	I-Application
engines	O
are	O
available	O
in	O
many	O
languages	O
,	O
and	O
the	O
engine	O
's	O
parameters	O
,	O
such	O
as	O
speech	O
rate	O
,	O
pitch	O
range	O
,	O
gender	O
,	O
stress	O
patterns	O
,	O
pauses	O
,	O
and	O
pronunciation	O
exceptions	O
can	O
be	O
manipulated	O
by	O
the	O
user	O
.	O
</s>
<s>
Many	O
recently	O
developed	O
SGDs	O
include	O
performance	B-Application
measurement	I-Application
and	I-Application
analysis	I-Application
tools	I-Application
to	O
help	O
monitor	O
the	O
content	O
used	O
by	O
an	O
individual	O
.	O
</s>
<s>
As	O
AAC	O
devices	O
are	O
designed	O
to	O
be	O
used	O
in	O
all	O
areas	O
of	O
a	O
user	O
's	O
life	O
,	O
there	O
are	O
sensitive	O
legal	O
,	O
social	O
,	O
and	O
technical	O
issues	O
centred	O
on	O
a	O
wide	O
family	O
of	O
personal	B-Application
data	I-Application
management	I-Application
problems	O
that	O
can	O
be	O
found	O
in	O
contexts	O
of	O
AAC	O
use	O
.	O
</s>
<s>
Programming	O
of	O
Dynamic	O
Speech	B-Application
Generating	I-Application
devices	I-Application
is	O
usually	O
done	O
by	O
augmentative	O
communication	O
specialists	O
.	O
</s>
