Patrick van der Smagt (1999) There is no Motion Without Sensing. Psycoloquy: 10(005) Efference Knowledge (12)

Volume: 10 (next, prev) Issue: 005 (next, prev) Article: 12 (next prev first) Alternate versions: ASCII Summary
Topic:
Article:
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 10(005): There is no Motion Without Sensing

THERE IS NO MOTION WITHOUT SENSING
Commentary on Jarvilehto on Efference-Knowledge

Patrick van der Smagt
Institute of Robotics and System Dynamics
German Aerospace Center / DLR Oberpfaffenhofen
P.O.Box 1116, 82230 Wessling, Germany
http://www.robotic.dlr.de/Smagt/

smagt@dlr.de

Abstract

This commentary considers the thought experiment of Jarvilehto (1998) in relation to sensing and motion. Jarvilehto is right that efferent influences are part of sensing, but there can be other explanations for why this is so. In particular, his conclusion that existing theories of motion and perception are incomplete goes beyond the supporting arguments he provides.

Keywords

afference, artificial life, efference, epistemology, evolution, Gibson, knowledge, motor theory, movement, perception, receptors, robotics, sensation, sensorimotor systems, situatedness
1. That sensing and motion are closely related is a long-established view (e.g., Gibson, 1979). The influence of motion on perception is usually stressed in this context (Bain, 1855). But one can also readily conclude that motion can only exist in the presence of perception.

2. Jarvilehto (1998) attempts to demonstrate why efferent influences are important for receptors. Yet when perception and motion are fully united, this observation becomes self-evident.

I. SENSING

3. Jarvilehto (1998) does not provide a precise definition of sensing in his target article. In his thought experiment in section V, he attempts to describe an organism without receptors. However, the organism does obtain energy "from the energy field in the form of induction" when it is moving. The concept that the energy intake of the organism varies with its ego-motion implies that it does have a mechanism of sensing its ego-motion. Indeed, in paragraph 31 Jarvilehto notes: "Even the idea of inducing energy from the field assumes some sort of general receptor."

4. Second, Jarvilehto implicitly assumes that the organism has a method of sensing time (possibly by measuring its energy consumption), which is not evidently true for an organism which can live forever.

5. Even though Jarvilehto's first hypothetical organism's method of sensing is very inefficient, it is not in principle different from that of the further organisms which can sense "directly" through their receptor holes. Clearly, in considering both the law of energy conservation and the equivalence of energy and matter, the only conclusion can be that motion has an influence on (the cells of) an organism, and must therefore be sensed.

II. MOTION

6. Motion itself can only exist in the presence of perception, either the system's own or that of an observer. (Schroedinger's cat-in-a-box thought experiment demonstrates this nicely.)

7. Jarvilehto's assumption in paragraph 37 that "movement is [traditionally] only seen as a component modifying the sensations arriving through the senses" is incorrect. A first, obvious, counterexample is the biological eye: the retinal receptors effectively work as differentiators, thus detecting movement only. This movement can be self-induced (as in vertebrates) or only due to the environment (as in frogs). But visual systems usually go further than that; they often depend on optic flow or a similar motion-based quantity. In that case, perception is impossible without motion.

8. This principle is also key in many robotics application (e.g., Smagt, 1995; Smagt & Groen, 1997). Such systems follow the remarks that Jarvilehto makes in paragraph 38: by defining motion in terms of perception parameters, a circle (traditionally called a feedback system) is formed. The task to be performed by a system is defined in terms of sensory signals, and its success or failure in terms of their deviations from expected signals.

III. THE REASON FOR EFFERENT INFLUENCES

9. Jarvilehto's thought experiment seems to be derived from his rabbit experiment. Yet, as Schmid (1998) and Vogt (1998) point out, it is quite possible that the motor-vision correlation is responsible for the optic nerve activity.

10. In fact, it is well known that much of the preprocessing of sensory signals is done in the sensor itself. This principle has been clearly demonstrated by Mead (1989), who devised retina-like electronic components in which much of the visual preprocessing is naturally implemented. Hence the precise definition of what is part of the sensor and what constitutes the central nervous system should be a key component of Jarvilehto's framework.

11. So too should the systems discussed earlier, that define their goal in terms of sensory signals function through feedback of the motor state of the system to the sensors. Since deviations from the expected path are expressed in terms of sensory signals, these must be fed back to the sensors themselves, and can thus be transformed to quantitative error signals.

IV. CONCLUSION

12. In conclusion, I think that Jarvilehto is right in arguing that efferent information is a key factor in sensing. However, the reasons he puts forth for this argumentation are erroneous, and hardly novel.

REFERENCES

Bain, A. (1855) The Senses and the Intellect. London: Longmans, Green.

Gibson, J. J. (1979) The Ecological Approach to Visual Perception. Boston: Houghton Mifflin.

Jarvilehto, T. (1998). Efferent influences on receptors in knowledge formation. PSYCOLOQUY 9(41) http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?9.41 ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/ psyc.98.9.41.efference-knowledge.1.jarvilehto

Mead, C. (1989). Analog VLSI and Neural Systems. Reading, MA: Addison-Wesley.

Schmid, U. (1998) Bottom-up and top-down processes in learning, commentary on Jarvilehto on Efference-Knowledge. PSYCOLOQUY 9(76) http://www.cogsci.soton.ac.uk/psyc-bin/newpsy?9.76 ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/ psyc.98.9.76.efference-knowledge.4.schmid

Smagt, P. van der (1995). Visual Robot Arm Guidance using Neural Networks. Ph.D. Thesis, University of Amsterdam, 1995. http://www.robotic.dlr.de/Smagt/thesis/

Smagt, P. van der and Groen, F. (1997). Visual feedback in motion. Neural Systems for Robotics, ed. O. Omidvar and P. van der Smagt, pp. 37-73. Boston, MA: Academic Press. http://www.robotic.dlr.de/Smagt/papers/SmaGro97.ps.gz

Vogt, P. (1998) Afferent influences on effectors in knowledge formation, commentary on Jarvilehto on Efference-Knowledge. PSYCOLOQUY 9(82) http://www.cogsci.soton.ac.uk/psyc-bin/newpsy?9.82 ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/ psyc.98.9.82.efference-knowledge.8.vogt


Volume: 10 (next, prev) Issue: 005 (next, prev) Article: 12 (next prev first) Alternate versions: ASCII Summary
Topic:
Article: