The origins of art
This is a guest post by Big Master Troll. To submit a post, see our Archives page.
“Let us look at a rose. Between it and the eye, in the dark room of the blank page, a number of authors are congregated, each with his own torch, his style. We have no perception of the rose without them. Each takes his turn: some torches cast a glowing white light which reveals the rose to be nothing more than a stalk, thorns, and a cluster of waxy petals. We are disappointed with the poverty of this impression. some others seem to be shining their torches straight into our own eyes. It is very colourful indeed; it is not unpleasurable; but after the seduction of dazzlement has been in progress for some minutes, we turn away, tired, knowing nothing new about the rose.
One man…his radiantly beautiful light seems to emerge from the heart of the rose, and intensifies as much as clarifies the perception; the rose suddenly seems the essence of all things to us.”
-Michael Ogilvie Imlah
1.0. Successful communication depends on an underlying similarity between the receiving agent and the transmitting agent. That is to say, if a signal is to transmit information, it must be presented in a form in which it can be understood by the receiving agent.
2.0. In any communication-game, there may arise situations where the message communicated is deceptive. That is to say, that the information communicated does not fully align with the underlying reality that is purportedly communicated by the information, often described as ‘signalling’. This makes sense, because agents often have different utility functions regarding the outcomes of any communication-game. Potentially ‘adversarial’ games of this type are common throughout nature. Clearly, such games are ‘won’ via obtaining a position of informational arbitrage - that agent has an advantage when they possess more information about another agent’s ‘real’ position, than they have about theirs.
3.0. Consequently, on the most basic level, there is an incentive to increase/decrease the valence/direction of any signal sent, as per an agent’s assessment of what will affect the receiving agent’s perceptions of the state of the game in an arbitrage increasing direction. This logic is of course equally valid in the reverse — it is advantageous to have a rival agent believe that one has ‘bought’ their signal to a greater extent that one has in reality.As a result of the above dynamics, signals sent between agents tend to move towards a greater ‘objective’ similarity with each other, as there is an incentive to signal as ‘strongly’ as possible in an advantageous direction, so as to create arbitrage situations, and in order to signal as strongly as possible, the signal must be presented in a form that is compatible with the ‘reading’ abilities of the receiving agent, which will in turn reflect some underlying nature of said agent.
3.1. Communication-games can take place across any timespan. For instance, the evolution by a carnivore of a weapon in the form of a claw, and the consequent evolution of an armoured hide by an herbivore, can be described as this process of similarity-creation. That is, a claw and the armour meant to protect against it, are in some sense more ‘similar’ to each other as a result of this arms-race, than two randomly selected features of animals that are not interacting in this way. This process of feedback-response is vastly more accelerated in more intelligent animals, where each agent is able to update their response to other-agent signals in a much faster way than can be done via genetic selection.
3.1.1. To clarify this further: while herbivores and carnivores are involved in a communication game, this communication, to put it colloquially, takes place on the level of ‘action’ rather than signalling. That is to say, the informational ‘level of play’ is lower. Their communication takes the form of non-deceptive actions by the participants, everything is left as it is. Some animals may develop higher level processes than this - for instance, an animal may begin to transmit information that is not ‘strictly true.’ For instance, many non toxic animals have developed colour schemas that mimic those of poisonous animals, although the updating/development of these schemas towards a similarity still takes place via the slow process of gene selection. At yet another level, you can have signals that update more rapidly - for instance, animals may shriek, posture, etc, to intimidate each other. Finally, for instance, you may have meta, human level signalling games, e.g. in a game of poker, one may not only signal falsely to deceive an opponent, but try to induce him to ‘bet deeper’ on a bad hand by attempting to signal to him that you have fallen for his bluff.
4.0. It seems likely that at some point, given some sufficient sensitivity of agents, the similarity between signals sent will become asymptotic. That the signals sent will become so similar as to achieve some kind of informational transparency to each agent i.e. both agents will be able to almost fully recognise the signal sent, and the signal received, as a result of their extreme similarity, i.e. a carrier of mutually agreed-upon information, which we call art.
4.1. To again clarify this statement, this is a definition of art that is somewhat compatible, and somewhat at odds with vernacular definitions. Typically, in everyday life, an attempt to communicate certain types of information will not be regarded as ‘art.’ For instance, if it is raining outside, and I ask a bystander if it is raining, and he answers that it is, this exchange would not be regarded as art. However, what I would like to stress, at the risk of stretching the definition of art too far, is that even this interaction takes place embedded ‘in art.’ The interaction takes place within a system of mutually understood symbols. Language is art — so much seems well known, but since it’s manipulation and use is so typical of humans, in the vernacular we reserve the designation of art for particularly human-rare informationally dense or creative forms of it.