Code-Switching Is a Narrative Device

My first draft of a family dinner scene switched between French and Darija fifteen times in three pages, exactly the way real bilingual dinners actually sound, and by the second page a reader who does not have both languages had stopped noticing the switches entirely.

That was the problem. In real life, code-switching often carries no special weight, a bilingual speaker moves between languages the way anyone shifts weight from one foot to the other, unconsciously, dozens of times an hour. Reproduced faithfully on the page, that same frequency reads as noise, because a reader parsing a written scene does not have the real-time social cues, tone, hesitation, the specific relative in the room, that tell a listener which switches matter and which are just weather. Fifteen switches in three pages taught my reader that switching means nothing, exactly the opposite of what I wanted a switch to mean when it actually counted.

The code-switch budget

I now set a budget before drafting any scene with more than one language in it: a fixed, small number of switches allowed per chapter, each one assigned in advance to a specific narrative job before I write a line of dialogue. A switch can mark a lie, the character retreating into the language they learned to deceive their mother in as a child. It can mark a threat, dropping into Darija to say something a French-speaking authority figure in the room will not follow. It can mark sudden intimacy, two characters briefly excluding everyone else at the table. What it cannot do, under the budget, is happen by accident, decoration, the writer reaching for a switch because the scene needed texture and this was the easiest way to add some.

Rewritten under a budget of three switches for the same dinner scene: one when the father lies about where he has been, one when the grandmother threatens the youngest son in a language his French wife does not speak, one when two sisters share a joke at the table that excludes their brother entirely. Three switches, each carrying a distinct and specific function, and every one of them now lands with the weight the fifteen accidental ones had spent into nothing.

Why scarcity is what makes the signal legible

A device only communicates if its absence is also meaningful. If characters switch languages constantly, a switch tells the reader nothing, because switching is simply the background condition of the scene, the way constant rain tells you nothing about today's weather. Once switches are rare, every one of them becomes a decision the reader can read, exactly as a single unexpected silence in a noisy room tells you more than the noise itself does. The budget is not there to make the scene less realistic. It is there to make the device readable to someone who is not living inside the bilingual household the scene depicts.

The obvious objection

You could object that this sanitises something real, that actual bilingual speech does not ration itself for a monolingual reader's convenience, and that a budget imposed for legibility is a kind of quiet apology for how people in these households actually talk, smoothing away the constant, unglamorous switching that is the real texture of the life. I think that objection is right about the facts and wrong about the conclusion. Real dinner conversation does switch fifteen times in three pages, and a transcript of it would be accurate. A novel is not a transcript. It is an instrument built to make a reader feel something specific at a specific moment, and an instrument that reproduces every frequency at equal volume produces only noise, regardless of how faithful that noise is to the room it was recorded in.

What the budget protects is not politeness toward a monolingual reader. It is the device itself, which stops being a device the moment it becomes ambient. I would rather under-represent how often my characters actually switch languages and keep the three switches that matter legible, than represent the true frequency accurately and lose all three inside a page a reader has already learned to skim past.

Setting the budget before you know the scene

The hardest discipline is setting the number before drafting rather than after, because a switch always feels justified in the moment you are writing it. I now decide the budget for a chapter during outlining, alongside decisions about point of view and tense, precisely so the number is fixed before any single scene can lobby for an exception. A budget you set after seeing the draft is not a budget, it is a rationalisation of whatever you already wrote, and it will not protect the one switch that actually needs the silence around it to be heard.