This project takes a new approach to one of the most difficult questions in the study of human evolution: what underlies our unique communication system, language? Our project considers language as both a cognitive and a social phenomenon, by looking at how how language solves the problem of efficient communication. Underpinning the idea of “efficient” communication is Information Theory. Pioneered by Claude Shannon in 1948, Information Theory is one of the highest impact mathematical and conceptual frameworks of the last century. At its core is the insight that receiving information reduces uncertainty about a situation in a measureable way. Broadly, more unexpected events provide more surprising information than expected events: for example, if it was cold and raining in the morning, the information that is still cold and raining in the evening is unsurprising. But if it's warm and sunny, that's new - and probably more surprising - information!
Our project applies Information Theory to the social nature of human language and its cognitive scaffolding, and examines how people subconsciously structure the spread of information when they speak and write. In language, information is contained in small units (words), but also in larger units (e.g., phrases). Like the weather example, more unexpected words provide more information: if I say cat you will probably imagine a generic version of the common animal. But if I say ocelot, that provides you with a lot more information, further reducing your uncertainty about what I'm referring to.
Previous research has suggested that language users will spread information evenly when they speak or write to make successful communication (with hearers or readers) more likely. In other words, rather than putting all the most informative words at the beginning or end of a sentence, language users prefer to scatter them evenly across a sentence. Crucially, they do this for people they are communicating with, as part of a phenomenon known as audience design: they are curating their utterances for those who are listening. More evenly spread information makes for a higher likelihood that a hearer will understand an utterance - even if there is interference or noise in transmission.
We suggest that humans have evolved to subconsciously manipulate language in this way because language is fundamentally social: it strengthens the social relationship between a speaker and interlocutor, and must be designed for what an audience or interlocutor can most easily hear, understand, and remember. Language is also fundamentally cognitive, and evolved to take advantage of the communicative strategies that human brains use to transmit information within themselves. To show this, the project will focus on the distribution of information across sentences in different contexts.
Do the linguistic forms favoured during language change show evidence of selection for more even distributions of information?
How do speakers (or writers) adjust information content distributions for different audiences across diverse contexts?
Is the tendency to optimise information density a general feature of human social cognition?
Smooth Signals and Syntactic Change
An information theoretic account of word order shifts in English and Icelandic.
Noise resistance in communication: Quantifying Uniformity and Optimality
Details a new descriptive measure for utterance uniformity, and tests the noise resistance conferred by uniformity using a simulation with data from the PYCCLE corpus.
CAIL is funded by the ESRC Secondary Data Analysis Initiative (SDAI), grant #ES/T005955/1. The project is hosted by Newcastle University in the School of English Literature, Language and Linguistics, as part of the Language Evolution, Variation and Acquisition group and the Center for Behaviour and Evolution.