InterJournal Complex Systems, 365
Status: Submitted
Manuscript Number: [365]
Submission Date: 501
Neural network-like hierarchical sociogenesis as a common evolutionary logic for bio-machinogenesis and semeiogenesis
Author(s): Koji Ohnishi

Subject(s): CX.18, CX.14, CX.13, CX.67, CX.66, CX.30, CX.07

Category: Article


Well-made biomachines such as animal body, bee society (= bee super-organism), and genetic apparatus seem to have emerged by hierarchical sociogenesis of lower-level individuals (Hamilton, 1964 ; Ohnishi et al., 1999). Bee eusociety and animal body are altruistic society consisting of fertile queens ( queen bee, germ line unicell organisms ) and workers (worker bee, somatic line unicell organisms). In the emergence of protein-synthesizing/genetic machine, early RNA replicator ribo-organisms (ROs) would have had a life cycle consisting of tRNA-phase and tDNA-phase. Such early tRNA ROs would have associated together to make a society in which some tRNAs would co-operatively behave to other tRNAs, and have begun to generate woker-like (wl-) tRNAs which are earliest mRNA/mDNAs and rRNA/rDNAs (poly-tRNA theory, Ohnishi et al., 1993, 1999). The original-type tRNA remained as queen-like (ql-) tRNAs (= contemporary tRNAs). ||||| Why such hierarchical societies could have evolved to be well-made machines ? A possible answer is that a (two-layered) hierarchical (altruistic) behavioral and DNA-information- flow network could make a self-learning neural-network machine capable of learning the status of self and improve itself to better-made machine, in every generation. Two models are possible ; ||| [I] Queen-worker-type altruistic society model (bee-society, anmal body): (1) Queens are the uppermost elements of the hiearchical neural-network machine, and the final output information of the queens is DNA sequence information (of queen bee, germ line cells) to be transmitted to the next generation via gametes. (2) Informations from circumstancing environment amd/or from inside the self-system are inputted to workers, and make altruistic behavior to queens which finally output queen's DNA sequences which are kin to worker's DNA. (3) Worker's DNA sequence information is thus partially outputted from queens, since workers are kins of queen(s). (4) The quantity of the sum of DNA outputs from all workers to the next generation through the i-th queen, Q[i], is given by, O(Q[i]) = Sum[i] r[i,j]A(W[j]), where Sum[i] denotes summation over i, and A (W[j]) denotes the increase of (newly grown) Q[i]'s DNA-output to the next generation caused by the j-th worker( W[j] )'s altruistic behaviour to Q[i]. r[i,j] denotes the coefficient of genetic relatedness between Q[i] and W[j], and partially corresponds to the connection weight in neural-netwok theories such as the Back Propaganda method (BPM). (5) The final out put of queen's DNA(s) makes a feed-back informations for reproducing the worker-DNAs and queen-DNAs of the next generation. ||| [II] Ql-wl-type quesi-altruistic society model (intracellular tRNA society= genetic machine): Hamilton's kin selection rule would not work, since the relatedness between RO-individuals is not so strong as in bee eusociety. DNA information of wl-tRNAs considerably differs from that of ql-tRNAs, and is outputted to the next generation via "genome replication" (= simultaneous replication of ql- and wl-DNAs) evoked by the replication of ql-tDNAs. Thus, co-operative behaviors of wl-tRNAs to ql-tRNAs can generate a DNA-information-flow from wl-tRNA's to the final (ql-RO's) output (consisting of ql- and wl-tDNAs), which makes a feedback DNA-flow to the ql- and wl-tDNAs of the next generation. This is essentially similar to the neural-network-like DNA-information-flow in case [I]. ||||| Furthermore, (1) In every of these altruistic/co-operative societies, mature semeiotic systems are observed ; synaptic signs between sensory and motor neurons, dance-language synapsis between "sensory bee" and "motor bee" (Ohnishi et al.,1999), triplet codon rules between anticodon (= "image" or "signifian" in de Sausseur's terminology) and amino-acid specificity (= "concept" or "signifie"). (2) Language system in human society seems to have evolved in a very similar logic. (3) Arbitrary correspondences between signifian (signifier) and signifie (the signified) were found in wide-range of generalized cognitive systems emerged from hierarchical societies. Therefore almost identical common logic of semeiogenesis would underlie the evolution of different semeiotic systems such as human language, bee dance language, neuronal synaptic signs, and genetic codon systems. |||| Ref.: Ohnishi et al. (1999), Proc. of 4th Int.Conf. on Artificial Life & Robotics, 344-349.

Retrieve Manuscript
Submit referee report/comment

Public Comments: