Code-switching and identity: longitudinal research on bilingual selves

Lina Park is a translator in Seoul, mid-thirties, raised bilingual in Korean and English. She tells me — switching mid-conversation, in English — that her English self is more direct, more willing to argue with her father about politics. In Korean, the same argument feels physically harder to begin. "Not impossible. Just different muscle."

She is describing what bilingual researchers have studied for three decades under the umbrella of cultural frame switching: the well-documented phenomenon that bilinguals report different traits, values, and emotional textures depending on which language they're operating in. The phenomenon is real, the explanations have evolved, and the implications for how we think about self are not trivial.

1. The first wave: language as cultural primer

Hong and colleagues' classic 2000 study presented Chinese-American bilinguals with cultural icons (a Chinese dragon vs. an American flag) before asking them to make attributional judgments about a scenario. The icons biased participants toward culturally consistent interpretations (Hong et al., 2000). The finding established the basic frame-switching effect: bilinguals carry two cultural reference systems, and the active one shifts contextually.

Subsequent work showed the language itself — independent of overt cultural cues — could trigger the shift. Chen and Bond (2010) had Hong Kong bilinguals complete personality measures in Cantonese vs. English. Participants reported more extraverted, open, and assertive traits in English; more agreeable, modest traits in Cantonese. The same person. The same week. Different scores.

2. The second wave: emotional reach

Caldwell-Harris's lab took the question into emotion: how does a bilingual feel in their two languages? Their findings are consistent and slightly unsettling. Taboo words, emotional reproaches, declarations of love all have reduced electrodermal response (the body's autonomic stress signal) when delivered in the speaker's second language vs. their first (Caldwell-Harris, 2014).

This is not a small effect. Childhood swear words in L1 produce skin-conductance spikes comparable to a startle. The same content in L2 — even at native-like fluency — barely moves the needle.

Adult late-learners report this experientially. Many bilinguals describe being able to have difficult conversations — about death, money, sexuality — more easily in their second language. The emotional bandwidth is narrower. That can be a relief.

3. The third wave: structural explanations

What's happening underneath? Three converging hypotheses:

Encoding contexts. First-language memories were encoded in L1; retrieving them activates the same neural representations they were stored with. L2 retrieval activates a thinner, more conceptual encoding (Marian & Neisser, 2000).

Acquired social roles. Each language is learned in specific social contexts (childhood family, adolescent peers, professional work). Speaking in a language re-activates the social roles associated with that context.

Reduced limbic engagement. L2 routes through more dorsal, less limbic pathways. The amygdala participates less. So does the autonomic system.

4. What this implies about identity

The traditional self-concept literature has assumed a unified core self with surface variation. Bilingual research keeps challenging this. The bilingual mind isn't deceived into feeling like two people. It actually deploys partially separate selves — different value weightings, different emotional responses, different social positioning — based on the language frame.

For adult learners, this lands as a paradox. The L2 self is initially impoverished. Over years, a fully realized second self develops. People who reach this point often describe it as having added a room to the house, not as having become someone else.

The destination, in other words, isn't fluency. It's a second functioning self.