Ghost channels are physical

Dr Bruce Long will be presenting this paper at the NZAP conference in New Zealand in December.

We can have detailed knowledge of complex systems like animal physiology, protein synthesis, and DNA translation/transcription, without having much, or any, information about the specific finitary evolutionary processes that originally produced them. Evolved representations embedded in the structures of complex natural systems either no longer carry, or else never did carry, represented information about those phenomena and processes. Otherwise, if organisms do carry causally induced representations of those processes, we may never be able to identify them, or what they indicate. The natural codes are no longer available and not derivable by deduction, induction, inference, or extrapolation.

However, we know from classical information theoretic principles that one token of a particular organism still carries non-accidental information about all of its conspecifics. The legumes in your garden carry information about all other legumes. Your pet goldfish carries information about all of its conspecifics, even though all tokens of their common distant evolutionary upstream causal source processes and phenomena have been lost. Similarly, one token of a particular type of celestial X-Ray source, like a specific kind of gas giant or a star of a particular type, carries information about all other entities of that type, but early conditions that produced original tokens of that type have dissipated.

Fred Dretske (1981) called the kinds of channels involved in these transmissions ‘ghost channels’. According to some probabilistic inferential conceptions of information transmission, causal pathways and token sources of origin are not necessary conditions for transmission. I’ll argue that ghost channels demonstrate that although information and transmission are importantly different from causation, ontic physical determination, and physical structure, these are still necessary conditions for the obtaining of both. Causal channel conditions, causation in causal signal pathways, and the interaction between channel conditions and causal signal pathways, are necessary conditions for all transmission.

Dretskian 'Ghost channel' B-C