The rules for using language are complicated enough that nobody fully understands how it works, and yet young children have no problem using and exploiting it with ease.
“How can a child learn something so complicated, something we cannot yet emulate with computers despite the advances of the last few years?” asks Emeritus Professor Laurie Bauer from Victoria University of Wellington’s School of Linguistics and Applied Language.
Compound words, such as Green Beret and Cat’s Eyes, are of particular interest to Bauer.
Although a Green Beret is a soldier not a hat, and a Cat’s Eye is a road marking not a feline eyeball, making these apparently unruly distinctions is second-nature to people.
How we are able to make sense of these words and construct new ones is a major focus of Bauer’s work and he is one of the world’s leading authorities on compound words.
In 2017, he was awarded the Humanities Medal by Royal Society Te Apārangi and also published his 20th book, Compounds and Compounding, which explores this complex area of word-formation.
“Consider a ‘word’ like oestrogen pill. It denotes a pill that contains or provides you with oestrogen. But a sea-sickness pill does not contain or give you sea-sickness,” says Bauer.
“So how do you interpret these things when you hear them for the first time? How do you manage to make them up on a regular basis and still have people understand them?”
Bauer says various scholars have suggested up to 100 different relationships between the elements in compound words like these, but these classifications are unable to capture the relationship between gravy and train in gravy train.
Take spaghetti and western in spaghetti western, where Bauer suggests the relationship “must be something like ‘the kind of western made in a country whose population is typically characterised as eating a lot of spaghetti’”.
“Examples like these suggest there is not a fixed set of relationships between the elements of compounds, but that speakers and listeners interpret compounds on the basis of their experience of similar forms, the nature of the things denoted by each element, and general cognitive principles of interpreting things in context,” explains Bauer.
Raising questions like these for discussion helps improve the description of language and the way in which people process it. In the long term, improved descriptions lead to better understandings of difficulties people might have learning a new language or because they have a problem as a result of, for example, a stroke.
“They also help people working in Artificial Intelligence program computers to deal with natural language in a way that makes them understand and produce appropriate language better, and help psychologists understand the workings of the mind,” says Bauer.
“By understanding the individual bits of language better, we hope eventually to reach a position where we understand the whole system, just as a jigsaw puzzle allows us to create a complex scene from small, individual pieces.”