This performative reading will be mediated by a computer program, which is written to visually interpret certain terms in our texts as they occur in the duration of the event. The purpose of the project is to explore, in real time, the relationship between cognition, memory, and preexisting knowledge in the act of communication, playing with tools of artificial intelligence that strive to approximate the process by which we transmit and receive. From two original and relevant texts of our own devising, we have selected 25 terms and sent them to each other. We have not read the other’s source text, in order to work from the blind inference characteristic of encountering new language fr the first time. Placed side by side, these lists produce 25 pairs – we will each create 25 sentences, placing each each word pair in a context suggested by the combination of words and our personal associations with them. We are also giving the pair of words to a program called ConcpetNet, a way to access a kind of collective source of associations and inferred meaning, which will come up with its own common sense link between the words in each pair.
As our key terms occur in the reading, the program will trigger our individually composed sentences to appear on screen, along with the machine-generated sentences. In this way, the program is outputting a variety of contexts from our input. These contexts come from the texts themselves, our prior knowledge, and will tend to represent the social assumptions and political desires that we carry with us into the creation of meaning in real time. The new sentences will rise from the bottom of the screen and we will choose from amongst those possible sentences. The chosen sentences will remain and accumulate into a new, spontaneously generated, 25-sentence composition over the course of the reading. The project points to the notion that both humans and technology tend to infer meaning based on contexts from an array of preexisting options, which vary always in subjectivity.