Sensitive periods, a more cautious term than critical period

The notion of a critical period comes from mid-twentieth-century ethology, notably work on imprinting in precocial birds, where a narrow time window determines attachment to a moving object. Konrad Lorenz described the phenomenon in greylag geese in the 1930s, and his work brought him wide renown.

Its transposition to humans was rapid and widely publicised. The field has progressively substituted the term sensitive period, which describes a window of lower learning cost rather than a closure.

Binocular vision provides the most solid example: uncorrected early deprivation produces lasting effects, which justifies systematic screening for squint and amblyopia in young children. The work of David Hubel and Torsten Wiesel, conducted on animals in the 1960s and rewarded with a Nobel Prize in 1981, established its physiological basis.

The phonological domain provides a second: discrimination of contrasts absent from the language of exposure declines across the first year, a robust result described notably by Janet Werker and Richard Tees.

Age of acquisition is associated with the level reached, especially for accent and for certain grammatical regularities. The existence of a sharp threshold, by contrast, has been much debated.

An analysis published in 2018 by Joshua Hartshorne, Joshua Tenenbaum and Steven Pinker, on a very large sample recruited online, places the inflection in acquisition capacity later than earlier models assumed, while confirming that an inflection exists. That result concerns grammar as measured by an online test, and says nothing about accent or oral fluency, which may follow other trajectories.

Follow-up of children who experienced severe institutional deprivation, notably the British study of children adopted from Romania, documented lasting effects associated with the duration of deprivation. That study, conducted by Michael Rutter's team, followed the children from their arrival in the United Kingdom into adulthood.

That work also shows substantial catch-up, particularly for children who left the situation early. Both results coexist and are regularly cited separately according to the thesis being defended.

Speaking of a window that closes leads to definitive readings, which do not match the state of the data for most domains studied.

Speaking of a sensitive period describes what is established: a time when learning is faster or less costly, without what was not acquired then becoming inaccessible.

The vocabulary can also err in the other direction. Invoking flexibility without further precision would erase the cases where a delay has lasting consequences, such as vision, which the term sensitive period also strives to avoid.