Vogt (1998) takes the traditional beliefs on knowledge formation so much for granted that it is difficult for him to see that it is precisely those beliefs I am questioning in the target article (Jarvilehto 1998a). Vogt maintains that the thought experiment shows the significance of afferent influences on effectors rather that of than efferent influences on receptors. However, within the framework of the organism-environment theory, the process of knowledge formation may be described in either way, because knowledge formation always involves both afferent and efferent elements of the organism-environment system.
2. Vogt claims that my thought experiment is wrong, because an organism without sensors is impossible, in principle. I agree that such an organism is impossible to realize in practise (Jarvilehto 1998a, para. 24), but this is a THOUGHT experiment. The impossibility of implementing such an organism does not invalidate the argument, because the main idea in the experiment is not to model real organisms, but to explore alternatives to the traditional role of receptors and of efferent influences on them. In the first part of the thought experiment I try to demonstrate that it is possible to imagine an organism which could achieve knowledge even in the absence of receptors.
3. Vogt's confusion could be partly due to some obscurity in the description of the experiment. Vogt has separated four parts, while I have in mind only three: (1) An organism without receptors (external) in a homogeneous environment with fixed walls, (2) an organism with receptors in a static heterogeneous environment without walls, and (3) an organism with receptors and efferent influences in a dynamically changing heterogeneous environment. The second and third parts are demonstrations of the necessity of receptors and efferent influences on them in certain types of environment; afferent and efferent processes not only offer better possibilities for perception, but they are essential under these conditions.
4. I agree that if we artificially stabilize the environment then knowledge formation may be based on afferent activity only (Vogt, para. 11), although this is probably never the case with living organisms. However, I argued earlier (Jarvilehto 1998b, para. 7) that sensing and moving are not simply related to receptors and muscles, respectively, but are in fact related to both. Thus, from the point of view of perception, it does not matter if we speak of efferent influences on receptors or afferent influences on effectors. Moreover, Vogt's claim that an organism without effectors can be considered to be an "agent" is contradictory, because an agent without the possibility of acting is no agent at all. An agent is something that acts and acting means efference, at least in some form.
5. Vogt claims that I make a mistake when speculating about possible computer simulations of the organism without receptors and cite Tani and Nolfi (1997) in support. He is partly right, because that paper does not provide direct support to my interpretation. I cited it because it was the only paper I knew indicating an attempt in that direction. However, since publication of the target article I have become acquainted with some of the work of Peter Todd, who has in fact made successful software simulations of organisms ("minimats") with no receptors at all (Todd and Yanco 1996). I was quite surprised to discover the alignment of some recent developments in robotics and AI with the principles of the organism-environment theory.
6. Vogt thinks the concept of knowledge should not be used for simple organisms but reserved for a "higher cognitive level," i.e., for the case when the organism knows what it is doing. However, even with humans, conscious knowledge (e.g., the ability to report what one is doing) is only a fraction of what one knows. Vogt's examples (para. 7) do demonstrate knowledge: a machine on two wheels can hardly be called an "organism," but a lot of what humans know is certainly stored in its structure. I do not know what "reactive behavior" means, exactly, but detecting a pheromone certainly implies that the ant can use some parts of its environment in a different way than some other parts. If so, then it must know its environment at least in the sense that it can achieve results, which are important in its life process. To call this merely "reactive" behavior is contentious and reflects the endeavor of the investigator to treat other organisms as mechanical devices and to preserve for humans alone the privilege of knowing.
I would like to thank Ms. Suzy McAnsh and Mr. Jukka Heikkila, M.Ed., for help in the preparation of my reply.
Jarvilehto, T. (1998a) 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
Jarvilehto T (1998b) What is a machine? Reply to Schmid on efference-knowledge. PSYCOLOQUY 9(79) http://www.cogsci.soton.ac.uk/psyc-bin/newpsy?9.79 ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/ psyc.98.9.79.efference-knowledge.5.jarvilehto
Tani J and Nolfi S (1997) Self-organization of modules and their hierarchy in robot learning problems: A dynamical systems approach. Sony Computer Science Laboratory Inc., Technical Report SCSL-TR-97-008.
Todd PM and Yanco HA (1996) Environmental effects on minimal behaviors in the minimat world. Adaptive Behavior 4: 365-413.
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