How words affect us, from the personal to the political

A young woman leans against a painted brick wall, resting her cheek in her hand. She is looking down at her phone.

How words affect us, from the personal to the political

McMaster University’s Victor Kuperman uses technology to help gauge the impact of language on individuals and society
February 11, 2015
An eye as pictured through an eye-tracker.

Undergraduate student, Emma Bridgwater’s eye as pictured
through an eye-tracker. Researchers at McMaster University
use sophisticated video cameras to track the eye movements
of people reading, which can help diagnose comprehension
problems and point the way toward effective treatment.
Credit: Noor Al-Zanoon

Reading and writing are uniquely human activities, and in the internet and social media era, more text is being produced than ever before in history. This flood of written data affects us on many levels, from the personal to the political. That’s why psycholinguist Victor Kuperman and his team at McMaster University have set out to measure a wide range of experiences with the written word, from the ability of an individual to parse words to large-scale trends in the language used to describe people and ideas. “What we want to be able to do is transition from very small units of measurement to very big ones, from a single person to a whole community,” he says. “We cover both individual language and language of a society.”

One of the most personal aspects of language is reading comprehension. In Canada, about a third of people aged 16 to 25 are functionally illiterate; they can read words, but they can’t follow a set of written instructions. In order to better understand the barriers to comprehension,  Kuperman is using funding from the Canada Foundation for Innovation (CFI) to buy an eye-tracker, which uses sophisticated video cameras to pinpoint exactly how someone’s gaze moves over a piece of text. This data is then correlated with their performance on other cognitive or verbal tests to see if there is a link.

For example, in one recent study, the team found that a person’s ability to tap out a steady rhythm was a strong predictor of their ability to comprehend written language. “There is a shared ability to coordinate the movements of the finger and the movement of the eyes” says Kuperman. This means that finger-tapping can be used to quickly diagnose reading problems in children. In low-literacy adults, tapping tests can help determine whether the problem is due to lack of educational opportunities or a deeper cognitive impairment. Kuperman is also purchasing a portable eye-tracker that can be brought into communities where reading comprehension is low.

On the other end of the scale, Kuperman’s team is making use of “big data” to study society-wide implications of language use. A CFI-funded suite of high-performing computers allows them to analyse millions of tweets, along with metadata on where and when each one was posted. Mining this rich vein of data can help researchers study the emotions associated with particular ideas, including brands, politicians or even countries. For example, by analysing the adjectives used near words like “Chinese,” “Russian” or “American,” the team found that countries with high living standards tended to be viewed in a more positive and less emotionally-charged way than poorer ones. “There must be something shared by countries like this,” Kuperman says. “We’re looking at census data and world statistics to figure this out.”

Big data can also study perceptions at various time scales, for example, by comparing tweets posted during a certain event to those posted months or years later. “We’re able to go from milliseconds, with the eye-tracker, to decades and centuries through textual analysis,” says Kuperman. “There aren’t too many labs that can do that.”

WATCH: Find out more about the Canada Foundation for Innovation.