Professional Documents
Culture Documents
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/240751960
CITATIONS READS
4 68
1 author:
Winfried Nth
Pontifcia Universidade Catlica de So Paulo (PUC-SP)
103 PUBLICATIONS 694 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
2016a Anlise de discurso com Peirce: interpretar, raciocinar e o discurso como argumento. Intexto
37: 34-46. e-ISSN 1807-8583. View project
All content following this page was uploaded by Winfried Nth on 25 May 2015.
Winfried Nth
Claus Emmeche (2000) raises the question Does a robot have an umwelt? His answer is
complex and differentiated. The concepts of umwelt and environment, says Emmeche, are
relevant to both biology and robotics. Organisms as well as robots interact with their
environment in processes of semiosis. However, environment must be distinguished from
umwelt, and umwelt in Uexklls sense can only be experienced by organisms not by
robots. Furthermore, even if robot had some experience of its umwelt, humans could hardly
ever know anything about such an experience.
Emmeche avoids one-sided positions, and since he always gives justice to the pros
and the cons of his arguments, it is difficult to disagree with him. However, the critique that
Ricardo Gudwin (1999) has developed in reply to Emmeche and indirectly also against
basic Uexkllian tenets indicates that some principles of Uexklls umwelt theory and their
semiotic implications might have been elaborated in a more pronounced way. My own
contribution to this discussion aims at clarifications in this direction, but also at some
semiotic extensions of the arguments concerning the processes of semiosis in the robots
interaction with its environment. (For further references see Nth 1997, 2000).
Let us begin with the common features of interaction of a robot and an organism with their
environment. There are at least four such common features, which support the argument
that not only organisms, but also robots have an umwelt in Uexklls sense.
The first is: both robots and organisms differentiate between the more encompassing
environment in Uexklls sense of Umgebung and a subjective environment, which Uexkll
calls Umwelt. The robot whose sensor devices can only detect green objects lives in an
umwelt which consists only of green things, but this green umwelt is only part of the more
encompassing environment of green, red, and black objects, which remains unaccessible to
the robots mind just as ultrasound, perceived by the bats, remains unperceived by humans.
The second reason why robots can be said to have an umwelt is the selective nature
of their interaction with their environment. The robot that reacts to the perceptual cues of
all green things while ignoring all red things perceives its environment selectively just like
humans who ignore ultra sound and perceive only sounds of a lower frequency. Since
robots equipped with different sensors thus have a different umwelt, there is also
subjectivity and relativity in their perception, and perceptual subjectivity and relativity are
Uexklls criteria of umwelt.
The third similarity is that organisms have, and robots may have, internal
representations of their umwelt. In Uexklls biosemiotics, the internal representation of the
organisms umwelt is called its Innenwelt robotics, is the model or symbolic
representation of the robots environment.
The fourth similarity is, so to speak, an anatomical one. It is apparent when we
compare Uexklls biological organism, as represented in his model of the functional circle,
with the design of a robot. Just as the biological organism is equipped with perceptual
organs and receptors as well as with effector organs and effectors, the robot has perceptual
modules and action modules to interact with its umwelt.
References