Most of us use the wrong word or misspeak from time to time, saying “squirrel” when we mean “chipmunk,” swapping sounds to utter “Yew Nork” instead of “New York” or calling a partner by a child’s name. Such slipups are more than just a quirk of human language, says Nazbanou “Bonnie” Nozari, PhD, a cognitive psychologist and assistant professor of neurology at the Johns Hopkins University School of Medicine. They’re also valuable tools for understanding the normal processes of speech.
“We have the ability not only to produce language, but to catch our errors when we make them. How do we detect those errors, apply corrections to them and prevent them from coming up again?” she asks.
Nozari aims to answer those questions as founder and head of the four-year-old Language Production and Executive Control Lab at Johns Hopkins University, where she studies the cognitive processes that monitor and control speech. The cognitive control of language production is surprisingly understudied, Nozari says. “My hope is that my work will help return language to those who have lost it.”
When speech misfires
One branch of Nozari’s research focuses on how we catch ourselves when we misspeak. Traditionally, researchers believed that the brain mechanisms involved in understanding language (the comprehension system) were responsible for recognizing and correcting slips of the tongue. While Nozari acknowledges the role of comprehension in detecting speech errors, her work suggests the brain mechanisms involved in generating speech (the language production system) play a key role in the process. She and her colleagues showed black-and-white drawings of objects to people who suffered from aphasia, or language impairment, after a stroke. The researchers recorded whether the participants named the objects incorrectly and, if so, whether they caught and corrected their mistakes. They found that each participant’s ability to detect errors in his or her speech was better predicted by that person’s language production skills, as opposed to his or her comprehension skills (Cognitive Psychologyopens in new window, Vol. 63, No. 1, 2011).
“There is no doubt that some part of self-monitoring happens through comprehension, but there are internal mechanisms within the production system itself that actually help catch and repair its own errors,” she says.
More recently, she and her colleague Rick Hanley, at the University of Essex in England, extended that theory to children. The research team tested 5- to 8-year-olds with the “moving animals” task, in which the children watched cartoons featuring nine familiar types of animals and described the events to the experimenter. Older children were better than younger children at catching and correcting their own semantic errors, such as calling a dog a cat.
Nozari, Hanley and their team also measured the maturity of each child’s language production system using a separate picture-naming task that required the child to identify the objects in a series of black-and-white drawings. By tallying the kids’ semantic errors (those related to meaning) and phonological errors (those related to sound), the researchers were able to estimate the strength of each child’s language production system using computational modeling. In particular, they showed that this strength was a key predictor in how well the children detected their errors in the moving animals task. This finding mirrored what Nozari and her colleagues found in individuals with aphasia, adding support to the theory that the language production system has its own built-in ability to catch verbal slipups, in children as well as adults (Journal of Experimental Child Psychologyopens in new window, Vol. 142, No. 1, 2016).
More recently, Nozari’s lab has begun to explore what happens after monitoring, specifically looking at how monitoring processes may help regulate and optimize speech production processes. So far, their work suggests that cognitive control processes, such as inhibitory control, play a key role in our ability to produce fluent and (mostly) error-free language, Nozari says.
A wolf in sheep's clothing
Nozari can pinpoint the moment she first became interested in speech errors. After receiving a medical degree from the Tehran University of Medical Sciences in her native Iran in 2005, she went to London to study people with Alzheimer’s disease. In a routine screening test for dementia, one of her research participants was shown a picture of a sheep and asked to name the object. First, he said “wolf.” He tried again: “steep.” Then, “sleep.”
“I was fascinated that these were not just random errors,” Nozari recalls. “‘Wolf’ is related to ‘sheep’ in meaning, ‘steep’ is related in sound, and ‘sleep’ in both meaning and sound. I was blown away by this phenomenon, and I started reading all about language.”
Her research on the topic led her to the work of cognitive psychologist Gary Dell, PhD, at the University of Illinois at Urbana-Champaign, who later became her PhD mentor. In 2014, she joined the faculty at Johns Hopkins, where she studies language production in healthy adults and children as well as older adults who have language deficits following a stroke.
These stroke victims are motivational for Nozari as she studies the monitoring and control processes that allow people to produce and understand language. She hopes her research will lead to new ways to restore language to those who have lost it. “One of the most rewarding parts of my job is working with participants with brain damage,” she says. “There is nothing more inspiring than seeing the effort and hard work they put into regaining lost function after a stroke.”
Much of Nozari’s research involves participants recruited from the Snyder Center for Aphasia Life Enhancement, an aphasia support and community center in Baltimore. In one project with those participants, she and her colleagues took a fresh look at how to reteach words to people after a stroke. Traditionally, these patients are taught words organized in semantic themes—learning fruits in one session, animal names in another. But all of us, with or without aphasia, are more inclined to mix up words that are similar to one another, Nozari says. “If you make a slip of the tongue, you’re more likely to confuse a fruit with another fruit than you are to confuse a fruit with an animal.” Nozari predicted that language therapy arranged by semantic themes might actually be less effective than therapy that retaught words in semantically unrelated blocks.
To test that idea, she and her colleagues performed a small pilot study with two people who had post-stroke aphasia. Each participated in six training sessions to relearn object names, with the words arranged within semantic groups (such as a block of fruit names) or in semantically unrelated groups. While grouping words by theme helped one participant remember them better in the short term, both participants had better long-term retention of the words they learned in unrelated groups. Nozari and her colleagues presented the results at the 2017 annual meeting of the Academy of Aphasia.
The findings could also have implications for second-language education. In one study being prepared for publication, Nozari and former graduate student Bonnie Breining, PhD, and their Johns Hopkins colleague Brenda Rapp, PhD, taught neurotypical adults an artificial language. They showed that participants were better at learning new labels for objects if they were trained in semantically unrelated blocks.
More recently, Nozari and her lab manager Jessa Sahl are completing a version of the language training experiment among Baltimore schoolchildren. Sahl taught 7- and 8-year-olds French vocabulary words, arranged in related or unrelated blocks, for several weeks. She revisited the students to test their recall of the words three weeks and six weeks later.
So far, the results suggest that children, too, learn words better when taught in unrelated groupings, Nozari says.
“It’s harder to learn something when it is presented along with similar things. Sometimes difficulty in learning can be a good thing because you put more effort into learning. But difficulty is undesirable if you cannot overcome it at the time of learning.”
While these findings are preliminary, Nozari hopes such research could point to ways to improve language instruction, leading to better learning outcomes for both students and people with language deficits.
Becoming a mentor
Nozari’s appointment is in the medical school’s neurology department, which doesn’t have a dedicated PhD program. Though she typically hosts a postdoctoral fellow and occasionally co-mentors graduate students from the department of cognitive science, most of her team includes undergraduate students and paid research assistants, who typically have bachelor’s or master’s degrees. She pays those assistants with help from both internal university funding and grants from sources such as the National Science Foundation and the National Institutes of Health.
Nozari embraces a direct approach to helping her students set deadlines and establish schedules. “It’s often hard for young students to get a grip on how to manage their time, while still doing quality work,” she says.
When it comes to her mentees’ research interests, though, she’s relatively hands off. Her students and research assistants often pursue projects of their own, as long as they fit into the theme of the lab. “I really believe students have to choose their direction,” she says. “I can give them some help, I can nudge them, but ultimately they have to come up with what they want to do, or they will not have vested interest in the research,” she says.
While most of Nozari’s work to date has focused on spoken language, she’s excited about the many possible directions her research can go. In the last couple of years, she’s collaborated with colleagues such as Rapp who have expertise in other modalities of language production, such as written language. Svetlana Pinet, PhD, a postdoctoral fellow in the lab, has a background studying the cognitive mechanisms at play when people type words rather than speak them. “Our backgrounds all touch on language production, so we can all understand each other and contribute,” says research assistant Chris Hepner. “But it’s a diverse enough group that we can bring different perspectives to the table.”
Going forward, Nozari hopes her team’s work will encourage other psychologists and scientists to see human language in a new light. “There has sometimes been a tendency to view language as so special that it is somehow disconnected from the rest of cognition,” she says. “The goal of a number of psycholinguists, including myself, is to situate language within the broader picture of cognition.”
“Lab Work” illuminates the work psychologists are doing in research labs nationwide.
