Next: 2.5 Bliss Up: 2. Visual Languages Previous: 2.3 Decisions During Design   Contents
Subsections
2.4 Minspeak
Minspeak is a visual language designed for use in augmentative communication.
It was developed by Bruce Baker around 1980 and has been further developed
by linguists, clinicians, and people who rely on augmentative communication.
Handicapped people, who can not use speech or hand signs with unfamiliar
communication partners, rely on aids such as pointing to letters or symbols
on a board. However if composing a sentence requires many `keystrokes' the
communication is slowed down significantly. Thus simple pointing devices
can not be used as efficiently or naturally as speech or might be even
impractical for everyday use. Minspeak reduces the number of symbols by use
of semantic compaction.
This needs some more explanation. Natural language refers to spoken
languages such as English or German. Natural languages consist of hundred
of thousands words, each of which can be encoded in a sequence of characters.
The set of possible characters is called the alphabet, which usually consists
of 26 letters.
A selection set is a set of symbols. Conceptually based symbols
are symbols which encode a concept rather than representing phonetics. An
example for a phonetically based system is the alphabet.
Every image or sound evokes notions. The primary iconicity
describes the notion of to what degree a
particular graphic or sound resembles its
referent. Secondary iconicity describes the notions evoked by the
image or sound. An example is shown in Figure 2.4. The
primary iconicity of that figure is ``cup''. The secondary iconicity of that
figure would include ``drink'', ``thirsty'', ``dishes'', or ``white''.
Now that the terms are defined, let us review the definition of semantic
compaction. By systematic use of secondary iconicity of symbols, we can
reduce the number of symbols needed to represent natural language. As
stated above, natural language consists of many thousands words. Thus it is
impossible to find a unique symbol for every word. On the other hand, if the
26 letters of the alphabet are used, one needs many keystrokes or
selection operations to write a word. By exploiting the multiple notions an
image creates we are able to encode different meanings with one image. Which
meaning is the `proper' meaning of the image is based on the context this
image is used in.
As Baker puts it, Minspeak is just the commercial name of semantic compaction.
A typical Minspeak system such as ``Language, Learning, and Living'' by Tony
Jones has a set of approximately one hundred icons, which represent some
thousand words and morphologies by icon strings never exceeding four in
length.
Minspeak does not enforce one set of given symbols. In fact, everyone can
create a set of symbols (and their meanings) on their own. But as this is
a time consuming task, there are many predefined Minspeak systems available.
These systems may be further enhanced and customized though.
These examples are given in the Minspeak FAQ [CMV98].
Lindsay is five years old and uses the Liberator (a commercial Minspeak
speech device). It has 32 pictures on its overlay. In her system she uses
nothing but two icon sequences. These sequences produce either words or
complete sentences. Figures 2.5 and 2.6 show two
examples. In Lindsay's system the first picture always codes the topic, and
the second one codes the specific idea.
Thus in Figure 2.5 the question mark codes the idea of
questions, and the apple represents meal. In Figure 2.6
the rainbow codes the idea of color, and the apple represents
the color red.
Sara, who is 22 years old, has one, two, and three icon sequences. Her system
consists of approximately 60 icons, which represent about 800 words or phrases
according to part of speech (for example: noun, verb, adjective).
Figure 2.7 shows different meanings of the symbol apple.
Semantic compaction is the systematic use of secondary iconicity to reduce
the number of symbols in a conceptually-based selection set for
the representation of natural language.[BRB94]
2.4.1 Examples
+ NOUN = food
+ VERB = eat
+ ADJECTIVE = hungry
+
+ VERB = cook
Next: 2.5 Bliss
Up: 2. Visual Languages
Previous: 2.3 Decisions During Design
  Contents
