Most of us like to believe we eat for rational, biological reasons. We get hungry, we eat, we stop when we’re full. Simple. Psychological research, however, has spent decades showing that this story is mostly fiction. One of the most famous demonstrations comes from a deceptively simple experiment involving soup, bowls, and a hidden tube mechanism.
In what is sometimes known as the “bottomless soup bowl experiment,” participants were invited to eat tomato soup until they felt satisfied. Unknown to some of them, their bowls were secretly refilled from underneath using a concealed pump. These participants consumed substantially more soup than those eating from normal bowls (about 73% more, on average), yet they did not report feeling more full afterward and often failed to notice that their bowls never emptied (Wansink, Painter, & North, 2005). Rather than relying on internal signals of satiety, people appeared to use an external cue to decide when to stop eating, which in this case was simply the visual emptiness of the bowl.
The implication here is critical: fullness is not purely physiological. Instead, people rely on environmental signals to tell them when they’ve had enough. When those signals are manipulated, eating behavior changes subconsciously.
This idea became a cornerstone of research from Wansink’s lab out of Cornell University, which focused on what he termed mindless eating. Across multiple studies, the lab demonstrated that portion size, container shape, and visual norms systematically influence how much people eat. Larger plates and bowls lead people to serve themselves more food and consume more calories without reporting greater satisfaction (Wansink & van Ittersum, 2013). Short, wide glasses cause people to pour and drink more liquid than tall, narrow glasses, even when they believe they are consuming less (Wansink & van Ittersum, 2003). In each case, perception, and not necessarily hunger, is driving behavior.
Similar findings have been replicated well beyond one lab. Portion size effects are among the most robust phenomena in eating research. When portion sizes increase, people reliably consume more food, regardless of hunger levels, body weight, or nutritional knowledge (Rolls, Morris, & Roe, 2002; Zlatevska, Dubelaar, & Holden, 2014). This effect appears in children and adults alike, suggesting that it reflects basic learning and conditioning rather than conscious decision-making.
Social cues matter just as much as visual ones. People eat more when dining with others than when eating alone, a phenomenon known as social facilitation of eating (de Castro, 1994). Individuals also tend to match the eating pace and quantity of those around them, often without awareness (Herman, Roth, & Polivy, 2003). In these contexts, eating becomes synchronized with social norms rather than bodily need.
Time and routine are powerful cues as well. People frequently report hunger simply because it is “time to eat,” even if they consumed food recently. Over time, clocks, locations, and familiar meal patterns become conditioned stimuli that trigger appetite automatically (Woods, 1991). Hunger, in this sense, can be learned (think: Pavlov’s dogs).
Beliefs about food also shape physiological responses. In a widely cited study, which I’ve discussed on WTFreud before, participants consumed the same milkshake on two occasions but were told it was either indulgent and high-calorie or sensible and low-calorie. Those who believed they were drinking the indulgent shake showed a greater reduction in ghrelin (a hormone associated with hunger) despite consuming identical nutrients (Crum, Corbin, Brownell, & Salovey, 2011). Expectation alone altered the body’s satiety response.
Taken together, these findings challenge the comforting idea that eating is guided primarily by internal biological signals. Hunger matters, but it competes with a powerful set of conditioned cues: portion norms, visual signals, social contexts, routines, beliefs, and beyond. Much of eating happens automatically, guided by an environment that quietly tells us how much is “normal.”
One important note is worth acknowledging. In recent years, several publications from Wansink’s lab were retracted due to data handling and reporting issues. However, the core phenomena discussed here are supported by a broad body of independent research across multiple laboratories. The bottomless soup bowl captured public attention, but the underlying conclusion does not rest on a single study.
The real lesson is not about soup. It’s about design. Our brains evolved to rely on external cues in environments where food was scarce and portions were stable. Modern food environments exploit those shortcuts. When willpower “fails,” it is often because conditioning, not hunger, is in control. Bon Appétit!
Leave a Reply
You must be logged in to post a comment.