How Job Titles Shape Our Expectations in Everyday Interactions
Imagine reading the sentence: “I need this truck moved, you’re making me late for work.” This simple complaint was presented to 300 American adults via the Prolific platform, revealed in small chunks as participants tapped the space bar to progress. Researchers at Yale University were less interested in the complaint itself and more focused on how readers processed the identity of the person involved. Specifically, they examined whether the role or job title associated with a person influenced how quickly and easily readers understood the situation.
In their study, the truck blocking the car was attributed either to a postal worker or a customer, and the person responding was either another postal worker or a different customer. Although the wording of the complaint remained unchanged, switching the job titles led to measurable differences in reading speed. Notably, readers took approximately 213 milliseconds longer to read the customer version compared to the postal worker version, with model estimates suggesting an even larger gap of around 262 milliseconds. This subtle but significant delay highlights how quickly and unconsciously people use job titles to form expectations about others’ behavior.
When One Postal Worker Replaces Another
The third study in the series explored the dynamics of role substitution within everyday interactions. Participants read scenarios where an object was borrowed from one individual and returned by another, such as a pen at a doctor’s office or a computer at a hotel front desk. One key example involved a blocked car at a post office.
Results showed that when a postal worker was replaced by another postal worker, readers processed the final part of the passage in about 1352.24 milliseconds. When the same customer remained throughout the interaction, it took slightly less time — 1327.88 milliseconds. However, when one customer was replaced by a different customer, reading time surged to 1565.66 milliseconds, an increase of over 50% compared to the study’s average reading time of 1024.69 milliseconds for similar text chunks.
The researchers concluded that readers experience “equal unsurprise” when a same-role substitution occurs or when no substitution happens at all, but genuine surprise arises when the person taking over lacks a relevant job role. This suggests that roles provide a cognitive scaffold that helps people anticipate how social interactions unfold — a salesperson stepping away mid-sale can be seamlessly replaced by another salesperson, because the role itself carries the interaction forward.
An exploratory analysis further supported this notion: nearly half of participants showed their slowest reading times during swapped-customer conditions, whereas slower times were less frequent when postal workers were swapped or no swap occurred. This pattern reinforces the idea that role consistency facilitates smoother comprehension.
Reading Times as Windows into Cognitive Processing
The methodology used — self-paced reading — is intentionally narrow but insightful. Passages are divided into regions revealed sequentially, with the duration between space bar presses serving as a proxy for processing difficulty. Participants simply read without any immediate task, so longer reading times indicate subconscious cognitive effort rather than deliberate judgment.
Surprise and confusion ratings collected after reading mirrored these timing patterns. In the first study, readers rated role-consistent scenarios as “not surprising at all,” situations where roles were broken as “pretty surprising,” and cases where an unrelated person took action as “very surprising.” In the third study, while 38.3% of readers still found swapped-customer interactions “not at all” surprising, 35% rated them as “a little” surprising, a statistically significant difference from other conditions.
Control conditions tightened the design. For example, a mechanic taking a cell phone (instead of car keys) slowed reading compared to a mechanic taking keys, ruling out simple object-based expectations. Similarly, a stranger mentioning a personal ulcer was more surprising than a doctor doing so, clarifying the role-based effect rather than mere mention of unusual topics.
Though some data handling decisions — such as excluding pauses longer than ten seconds — were adjusted after initial analysis to avoid discarding meaningful slowdowns, the authors transparently discuss these choices, underscoring the robustness and limitations of their findings.
Doctors, Ulcers, and Knowledge Attribution
The second study shifted focus from actions to knowledge attribution. Participants read sentences where a doctor or a fellow patient remarked on an ulcer in the reader’s stomach. Reading times averaged 896.47 milliseconds when the doctor spoke, close to the study’s baseline of 809.09 milliseconds. However, when a fellow patient made the same remark, reading time rose significantly to 1044.55 milliseconds. Interestingly, when the patient spoke about their own ulcer, reading times did not increase, suggesting that role and knowledge expectations influence comprehension.
This nuanced finding challenges simpler interpretations. The authors caution that readers might not explicitly represent others’ knowledge but instead form low-level expectations about what typical speakers say based on their roles. Even if so, these role-based expectations may scaffold deeper reasoning about knowledge when needed.
Memory Errors and Reconstruction
Beyond reading times, the studies assessed participants’ recall accuracy. Readers were asked to identify who performed specific actions or made statements. In the first study, 92% correctly named the role-holder when appropriate, but only 69% did so when the actor lacked a defined job role. In the second, 97% correctly identified the doctor, versus 74% when the line was spoken by another patient.
The authors propose an intriguing hypothesis: because readers could not revisit earlier text, they might have assumed they misread and reconstructed events to be role-consistent. Rather than poor memory, these errors may indicate an active cognitive process aimed at maintaining coherent narratives aligned with role expectations.
This contrasts with other research showing that unexpected events are often remembered better, highlighting the complexity of how surprise and expectation interact in memory formation.
Limitations and Future Directions
While the study offers compelling evidence for the cognitive weight carried by job titles, the authors acknowledge important caveats. All participants were based in the United States, and cultural differences in role perception may affect generalizability. Furthermore, the experiment explicitly labeled roles in text; whether visual or contextual cues alone would evoke similar rapid expectations remains an open question.
The authors also avoid overgeneralizing their findings, emphasizing that their results do not claim role-based reasoning is inherently faster than more complex social cognition processes like Theory of Mind. Instead, roles likely serve as an efficient heuristic in everyday social understanding.
Two key questions remain: how quickly these role-based expectations activate in real-world contexts, and the boundaries of role substitutability — for example, when a friend might be replaced mid-interaction versus when this is less acceptable. The authors share materials, data, and code openly (available here), inviting further research to explore these nuances.
This research was published on 19 August in Communications Psychology by Aaron Baker, Yarrow Dunham, and Julian Jara-Ettinger of Yale University. It sheds light on the subtle but powerful role that job titles play in how we process social interactions — quietly shaping our expectations and comprehension in milliseconds.
For further details, see the original article Here.
