fMRI of the lectulus tuberosus, the brain's passivity center
Reprinted from What’s New at Cornell? Press release from the University News Bureau
The human brain is hardwired to prefer the passive voice. A definite predilection for passive constructions has been found by a team of neuroscientists led by Elaine Bao Weiss and W. Strang-Ng, postdoctoral researchers at Cornell University’s Neurosyntax Imaging Laboratory.
“This was totally unexpected,” Bao Weiss said of the findings. “Generations of writers have been advised to prefer the active to the passive, but that’s not how the brain works.”
According to Bao Weiss, the research explains why writers have to be told over and over to prefer the active: their brains expect the passive and respond more quickly to it. Pointing to evidence from the scholarly literature, she observed, “Even experts like Strunk and White and George Orwell use the passive voice when telling writers to use the active.”
Many a tame sentence of description or exposition can be made lively and emphatic by substituting a transitive in the active voice.
William Strunk, Elements of Style, 1919, p. 24.
[T]he passive voice is wherever possible used in preference to the active…
George Orwell, “Politics and the English Language,” 1946, p. 256.
In a series of double-blind experiments, Bao Weiss and Strang-Ng tied passive syntax to an area of the brain called the lectulus tuberosus, the passivity center. Researchers showed test subjects (n = 143) images ranging from simple dipolar object renditions to complex mulitvariate graphical (CMG) tokens. Subjects were chosen from students in introductory psychology courses, controlling for gender (M = 71; F = 72) and age (mean = 19.3 yrs). They were told they were being tested for associative creativity skills. Seated at computer terminals, subjects were asked to describe what was happening in pictures that appeared fleetingly on the screens (t = 0.5s), and their responses, recorded on webcams, were coded by research assistants who had been told the experiment was designed to test for lexical choices along a concrete-abstract axis. The coders were also asked to mark forms of the verbs to be and to get, as well as occurrences of prepositions. This was explained as a control for accuracy. Then a second team of analysts isolated those specific lexical items associated with passive syntax--the true data being sought--and ran a regression to determine passive validity.
A second group of research subjects (n = 133, M = 60; F = 73; median age 20.1 yrs) was shown blank cards and asked to write a sentence no longer than 15 words describing what would be on the cards if in fact the cards had pictures printed on them. Their written responses were similarly coded and analyzed. Subjects who produced passives did so more quickly than those producing actives (tpassive = 0.6ns; tactive = 0.87ns).
A subset of the two test groups (n = 14 and n = 15, respectively) was administered fMRIs while their responses were being composed. Evidence from the fMRI’s showed rapid onset of electrochemical activity in the lectulus tuberosus corresponding to the presence of passive syntacticity (t = 0.003 ns). In contrast, the LT did not evidence significant activity for students producing active tokens in speech or writing. Hence a correlation between passive thinking and passive syntax is confirmed in the test subjects (p < 0.05).
fMRI imaging: the lit-up area shows that the lectulus tuberosus, the brain’s passivity center, lights up when subjects use the passive voice.
So strong was test subjects’ preference for the passive that the University’s Human Subjects Board recommended that the experiment be terminated prematurely. Test subjects were sent home after being admonised to use the active voice. Strang-Ng commented, “We could not in good conscience allow these subjects, who are students after all, to continue using the passive, even if they were predisposed to do so, because that would put them at risk for other problems down the line, like split infinitives, sentence-final prepositions, and, of course, low grades on English papers.”
Bao Weiss and Strang-Ng’s findings were announced last week at the 2014 ACES conference in Las Vegas. Many conference attendees expressed skepticism. “I will wait to see if these results can be replicated, but I do not think they may be” said Lynne Truss Ramamurthy, who holds the chair in the neurology of usage at the Max Planck Institute. But E. W. Gilman, of MIT’s biolexicology group, who was not involved in the research, was cautiously optimistic: “These findings reinforce recent evidence from palaeohistory and go a long way to explaining some aspects of human behavior that have previously been resistant to analysis. If these data on passive preferentiality can be borne out, then I think we have some important clues into the connections between language and mind.”
The paper, “An Element of Style: Neurotypical passivity in stimulus-generated syntactical structures,” appears online in the prepublication section of the current issue of Neurolinguistica.
Other Web of Language Posts you may like:
Grammar Sticklers May Have OCD
The Great Language Change Hoax
The Book, the Scroll, and the Web
House Passes Bill to Ban Texting in Spanish