Why Children Forget Farsi: What First Language Attrition Research Actually Shows
Parents raising children outside Iran describe a familiar sequence: a child who spoke Farsi comfortably at four answers only in English at seven, and by ten struggles to retrieve words that were once automatic. The phenomenon is real and has been studied for four decades. It is also widely misdescribed in popular writing, which tends to assert precise age windows and exposure thresholds that the research does not support. This article sets out what first language attrition is, how it differs from the related but distinct phenomenon of divergent acquisition, what is actually established about the role of age, and what the evidence suggests families can do.
Written by the curriculum team at Online Persian School, London, Ontario.
What first language attrition is
First language attrition refers to the decline of a previously acquired first language under reduced use, typically following migration and immersion in a dominant second language. Schmid (2011) provides the standard book-length treatment. Attrition is not forgetting in the ordinary sense. Much of what appears lost remains accessible under the right conditions: retrieval slows, access becomes effortful, and production degrades before comprehension does — which is precisely why a child can understand a grandparent perfectly while being unable to reply.
The single most useful thing for a parent to understand: the comprehension your child retains is an asset that has not disappeared. It is the bridge on which speech can be rebuilt, and it is why starting at nine is not futile even though starting at four would have been easier.
"After trying so many different Farsi classes and getting no results, my daughter is finally starting to speak Farsi."— Baharak Solasi, parent, Facebook review
A distinction that matters: attrition versus divergent acquisition
Not every gap in a child's Farsi is attrition. A structure that was never fully acquired cannot be lost. Children who emigrate young, or who are born abroad, often receive input that is quantitatively reduced and qualitatively narrowed — restricted to domestic registers, delivered by a small number of speakers, and rarely written. The resulting grammar may differ from the homeland variety without anything having been lost at all.
Distinguishing the two has occupied the field for some time. Montrul (2008) examined the age factor systematically and argued that much of what looks like loss in bilinguals is better characterised as acquisition that proceeded differently under different input conditions. Schmid and Köpke (2017) argue that attrition is not a marginal curiosity but central to how bilingual development should be theorised generally.
Kupisch and Rothman (2018) push the point further, and their framing deserves to reach parents directly: difference is not incompleteness. A child whose Persian diverges from Tehran Persian has a real linguistic system, not a broken one. This is not a consolation offered in place of the facts — it is the current scholarly position, and it has a practical consequence, because children who are told their Farsi is wrong tend to stop using it altogether.
"Online Persian School was the place for our daughter to proudly stay connected to her cultural and linguistic Persian roots… She was nurturing a fun educational environment where Niki was able to learn and love to learn."— Sara Saberi, parent, Facebook review
Age: what is established, and what is asserted without support
It is commonly claimed in popular articles — including in an earlier version of this one — that attrition peaks within a specific narrow age band, and that a precise weekly threshold of dominant-language exposure triggers it. Those figures circulate widely and are not, so far as we can establish, traceable to a primary source. They should not be repeated, and we no longer repeat them.
What the research does support is more general and more useful. Age of onset of bilingualism matters, and the vulnerability of the first language is not constant across childhood. Bylund (2009) examined whether maturational constraints govern susceptibility to attrition, and the broader literature summarised by Schmid (2011) converges on the finding that first languages are considerably more vulnerable to attrition before puberty than after it. In adults, wholesale loss of a fully acquired first language is rare; in young children, substantial shift is common.
The practical inference for a family is therefore directional rather than numerical. The younger the child, and the more abruptly the dominant language takes over the child's day, the more actively maintenance has to be pursued. There is no cliff edge at a particular birthday, and a family that believes it has passed some deadline has not missed anything irrecoverable.
What actually predicts maintenance
Across the heritage-language literature the recurring predictors are not mysterious, though they are demanding (Polinsky, 2018; Montrul, 2016). Three stand out.
Quantity of input
Hours of meaningful exposure, sustained over years — not intensive bursts followed by gaps.
Quality and range of input
Language from more than one speaker, in more than one register, about more than domestic routine. A child who only ever hears Farsi used to give instructions acquires the Farsi of instructions. Range is what turns a household vocabulary into a language.
"She can read and write in Farsi, and also she is learning a lot about Iran's culture, which is amazing. She knows about cities and historical places in Iran."— Negar Dgt, parent, Google review
Literacy
Reading and writing open a channel of input that does not depend on a speaker being present, and they are strongly associated with better-maintained heritage grammars. This is also where a family can add input without adding hours of conversation they do not have. And it does not compete with school literacy: Cummins (1979) argued that proficiency developed in one language supports the other through a common underlying proficiency, which is the empirical answer to the persistent worry that Farsi will crowd out English.
"I finished Farsi Grade 1 in seven months and now I can read children's books and write simple sentences."— Mahkameh Mohsenin, student, Google review
The family is the decisive variable
The study of how families make and enact language decisions is now a field of its own. Spolsky (2004) established the framework of language policy operating at every level from the state to the household, and King, Fogle and Logan-Terry (2008) set out family language policy as a distinct area of inquiry: the explicit and implicit household decisions about which language is used, with whom, and in what circumstances.
The practical finding is uncomfortable and worth stating plainly. Attitude alone does not maintain a language. Families overwhelmingly report wanting their children to keep Farsi. The ones who succeed differ not in wanting it more, but in having converted the wish into ordinary, repeated, non-negotiable practice — a language with protected times, protected people, and consequences no more dramatic than a parent who simply does not switch.
What follows for parents
- Respond to Farsi in Farsi, consistently, without making the child's reply a condition of the conversation continuing.
- Widen the range of input: relatives, recordings, stories, other children, more than one adult voice.
- Add literacy early, in small amounts, because it is the one input channel that does not require your presence.
- Do not correct into silence. A child corrected on every utterance stops producing utterances — which is the outcome the correction was meant to prevent.
- Treat a weekly lesson as a scaffold for home use rather than a substitute for it. The lesson does not maintain the language; it makes home use easier to sustain.
Limits of the present evidence
Two caveats apply to everything above. First, Persian is markedly under-represented in this literature, which is dominated by Spanish, Russian, Chinese and Turkish; extending those findings to Persian-speaking families is a reasonable inference and not a demonstrated result. Second, much attrition research is cross-sectional — comparing different people at different ages rather than following the same children over time — which limits how confidently causal claims can be made about the trajectory of any individual child.
Neither caveat undermines the practical guidance. Both are reasons to distrust anyone, including a school, who offers you a precise number.
References
- Bylund, E. (2009). Maturational constraints and first language attrition. Language Learning. https://doi.org/10.1111/j.1467-9922.2009.00521.x
- Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49(2), 222–251.
- King, K. A., Fogle, L., & Logan-Terry, A. (2008). Family language policy. Language and Linguistics Compass, 2. https://doi.org/10.1111/j.1749-818X.2008.00076.x
- Kupisch, T., & Rothman, J. (2018). Terminology matters! Why difference is not incompleteness and how early child bilinguals are heritage speakers. International Journal of Bilingualism. https://doi.org/10.1177/1367006916654355
- Montrul, S. (2008). Incomplete Acquisition in Bilingualism: Re-examining the Age Factor. Amsterdam: John Benjamins.
- Montrul, S. (2016). The Acquisition of Heritage Languages. Cambridge: Cambridge University Press.
- Polinsky, M. (2018). Heritage Languages and Their Speakers. Cambridge: Cambridge University Press.
- Schmid, M. S. (2011). Language Attrition. Cambridge: Cambridge University Press.
- Schmid, M. S., & Köpke, B. (2017). The relevance of first language attrition to theories of bilingual development. Linguistic Approaches to Bilingualism, 7. https://doi.org/10.1075/lab.17058.sch
- Spolsky, B. (2004). Language Policy. Cambridge: Cambridge University Press.
If your child has stopped answering in Farsi
Most of the families we work with started exactly there — a child who understands everything and replies in English.
A weekly lesson will not do this work on its own; as the article above says, what happens at home matters more. But it gives a child a reason to speak Farsi to someone who is not a parent, and that is often what breaks the silence.
One week of lessons, free. No card, no commitment.
