Core77 shared a “year in review”-type post highlighting curious design solutions. One, in particular, caught my attention: an “anti-gluttony” kitchen door from a 12th-century monastery. Esquire has details:

Measuring two meters high and just 32 centimeters wide, the door served a practical purpose for the monks, many of whom were reportedly on the heavy side. There’s even an inscription carved into the entrance to the refectory that translates to, “Consider that you eat the sins of the people.”

The monks were required to pass through the door to get their own food from the kitchen and bring it to the refectory to eat. If you couldn’t fit, you weren’t allowed to eat, which meant you were forced to fast until you lost the weight. Gluttony is a mortal sin, after all.

The idea sounds outrageous and inappropriate to 21st-century ears. It reminds me of an even more egregious structural constraint: Robert Moses’s parkway bridges, which were allegedly designed to intentionally exclude minorities.

As offensive as they are, these are examples of an important design principle: structural constraints modify behavior. Once you build a door 32 cm. (12.6 in.) wide, only some people will pass. The outcome isn’t just a narrow passageway, but different use patterns, power imbalances, shame, etc.

When cast in stone — as in Moses’s parkway bridges and the monastery wall — constraints can be long-lived. We build atop such structural elements. (E.g., we connect roads and railways to bridges.) Once they’re part of systems (whether in stone or code), fixed structural constraints limit our ability to iterate: they don’t just limit user behavior, but also the scope of action of future system stewards.

As a result, you must take great care when designing structural constraints. Intended behavior changes should have noble aspirations. (How do you know?) And beware of setting such constraints unintentionally! Systemic constraints define usage patterns — sometimes long after we’ve outgrown the mindsets that led to their creation.