In, on, at: why one Spanish «en» becomes three English prepositions

"I'm in the bus," you type into the work chat. "See you in Christmas," you tell your coworker in Denver. Both sentences land, and both give you away: the bus takes on and Christmas takes at.

Your vocabulary isn't the problem. Spanish covers with one word what English splits across three. The way out isn't memorizing more rules. It's trading the rule for an image and storing the rest as chunks. Casey Taliancich-Klinger, Lisa Bedore and Elizabeth Peña, who tested 148 Spanish–English bilingual children for the Journal of Child Language, name the collision plainly: "in English the prepositions at, in, and on can be expressed with one preposition, en, in Spanish." They also measured what it costs. At was the least accurate English preposition of all, at 42.1% correct, with on second at 48.2%.

One word of yours, three slots in English#

The authoritative description is spatial, not a list of exceptions. Cambridge defines all three this way: at marks a position you see as a point, on marks a position on a surface, and in marks a position inside a larger area. When you hesitate, ask yourself what you're picturing before you ask which rule applies.

Wrong: I'm in the bus. Right: I'm on the bus.

English reads public transportation as a platform carrying you, not a box containing you, so it takes on: on the train, on the plane, on the bus. Switch to the mode of travel rather than the vehicle and it changes again: by train.

Real situation: you're on the 22 heading to a Chicago office, running fifteen minutes late, typing into the team channel. I'm on the bus, five minutes goes unnoticed. In the bus makes somebody read the line twice.

Christmas is a point; July is a box#

Wrong: See you in Christmas. Right: See you at Christmas.

Big holidays work as a single point on the calendar, which is why Cambridge files them under at: at Christmas, at the New Year, at the weekend. Months, years and seasons do wrap around you, and there in rules: in July, in 1835, in the spring. Specific days rest on the surface of the calendar and take on: on October 25, on Friday.

Real situation: you're confirming the trip with your partner's family in Miami. We arrive on December 22nd and we'll be there at Christmas puts each piece where it belongs, one day on the calendar and one holiday as a point. Nobody has to ask you again.

"In the morning," but "on the morning of the 9th"#

Wrong: We landed in the morning of the 9th. Right: We landed on the morning of the 9th.

A morning with no date is a stretch of time that wraps around you, which is why in the morning is correct. Attach a date and that morning becomes a slice of one specific day, so it jumps to on. The same turn separates at night, nights in general, from in the night, one particular night.

Real situation: you're telling your Austin team on a call when you landed. On the morning of the ninth sounds like a fact. In the morning of the ninth sounds like a translation.

No rule covers you completely#

Native speakers don't agree with each other on which preposition belongs. In an experiment using 200 encyclopedia sentences, two native speakers filled in the missing preposition and matched each other only 76% of the time. Martin Chodorow, Joel Tetreault and Na-Rae Han, who build automatic detectors for this error, put it this way: "preposition usage is one of the most difficult aspects of English grammar for non-native speakers to master." Their paper cites an analysis in which prepositions were the largest error category, roughly 29% of the total, among 53 intermediate and advanced students.

Proficiency doesn't fix this. Penny MacDonald analyzed 304 essays by Spanish university students writing in English and found that picking the wrong preposition is the most frequent grammar error precisely at the advanced levels: 15% of grammar errors at C1 and 23% at C2. The cause she names is direct translation from Spanish.

Store chunks, not rules#

Teaching the spatial image pays roughly triple what teaching the rule pays. In a controlled experiment, 50 students learned the same ten prepositions — in, on and at among them — under two different methods: the group working with spatial schemas gained 6.96 points, and the rules-and-pictures group gained 2.08. A 2024 systematic review of 62 studies found the approach effective in 93.5% of them.

The rest of the work comes prefabricated. About half of fluent native text isn't built word by word at all; it's assembled from ready-made chunks. Nick Ellis and his colleagues report that estimate in TESOL Quarterly and quote John Sinclair: "a language user has available to him or her a large number of semi-preconstructed phrases that constitute single choices." On the bus, at work and in the morning are three of those phrases. You don't derive them on the fly. You either have them or you don't.

Could you say I'm on the bus out loud right now without routing through "estoy en el autobús" first? That gap between knowing the rule and having it available is exactly what Tabla trains: you say the whole phrase out loud, against the clock, until the preposition arrives without you thinking about it.

Say all three out loud: "See you ___ Christmas," "We landed ___ the morning of the 9th," "I'm ___ the train."

At Christmas, because English treats a big holiday as a single point on the calendar. On the morning of the 9th, because a morning with a date stops being a stretch of time and becomes a slice of one specific day. On the train, because transportation reads as a surface, not a box. If all three came out with in, you translated from Spanish, which is the exact pattern the corpora document.

The next time you're about to say en in English, stop for half a second and look at the picture: something inside a space, something resting on a surface, or a point. That question gets it right more often than any list of exceptions.

Try this today: record twenty seconds telling someone where you are and what you're doing tomorrow, forcing yourself to use on the bus, at work and in the morning. Play it back and check that none of the three turned into in.