Jean Fox Tree:
Embracing the spoken word

We don't process speech by listening for the next word and
slotting it into the formulas we learned in grade school--a noun
is followed by a verb, remember? We can comprehend speech
that breaks all those rules as quickly as we can understand
grammatical speech.


Jean Fox Tree is, like, a psycholinguist who specializes in, uh, the little stops and starts that, you know, pepper our speech. Dubbed "disfluencies and discourse markers," these pauses and tiny words account for as much as 6 to 10 percent of spontaneous speech, yet they've been largely ignored because most scholars consider them mistakes. Enter Fox Tree, an assistant professor of psychology whose work is challenging some long-held assumptions about speech comprehension.

Disfluencies include repetition (in the in the room), false starts (Do you have the--what time is it?), and speech fillers such as um and uh. Discourse markers are words like well, oh, I mean, you know, and, more recently, like totally.

A cognitive psychologist, Fox Tree began studying spontaneous speech as a graduate at Stanford University. She is convinced that these speech "undesirables" play specific roles in our communication, sending signals to the listener to do things like pay more attention, help the speaker find a word, or be patient while the speaker gathers his or her thoughts.

Indeed, Fox Tree believes these "mistakes" may not hinder comprehension at all--they may actually help us understand each other.

"What I'm finding is that we don't process speech by listening for the next word and slotting it into the formulas we learned in grade school--a noun is followed by a verb, remember?" says Fox Tree. "We can comprehend speech that breaks all those rules as quickly as we can understand grammatical speech."

For example, one of Fox Tree's studies showed that listeners understand "This lady walks into the gift the gift department of a store," just as quickly as they grasp "This lady walks into the gift department of a store."

Similarly, the sentence "That was a bit um confusing," takes no longer to comprehend than "That was a bit confusing."

What is unique about Fox Tree's research is that she uses spontaneous speech in her studies, spending hours recording actual conversations in the laboratory, complete with all the uh's, er's, and y'know's that riddle natural speech. By contrast, standard tests of speech comprehension use scripted speech that is recorded by actors who have rehearsed their lines until they are able to perform without a single pause, um, or false start.

Using her tapes and computer techniques that psycholinguists could only dream of ten years ago, Fox Tree has pioneered a technique of measuring listeners' comprehension of the same
sentence with and without the um, uh, or other disfluency or discourse marker.

It's an area of research that has already challenged fundamental tenets of speech comprehension, and Fox Tree is looking ahead to the impact her work will have.

"When I first introduced my work, a lot of my colleagues asked me why I had wasted years taping and transcribing actual conversations when I could've just used scripted speech like everyone else," recalls Fox Tree, who hasn't yet tired of explaining her work. "I told them, 'Because we don't know that an actor's speech is the same as spontaneous speech, and I have very good reason to believe that it's not.'"

That, um, convinced them.

Jennifer McNulty