From Defaults to Semantic Sensitivity: The Acquisition of Nominal Concord

This study investigates the development of nominal concord in Jordanian Arabic (JA) among children aged 2.0 to 10.11. Drawing on naturalistic data from the Child Language Corpus of JA, the analysis examined 441 obligatory noun–modifier contexts produced by 137 children. The dataset was restricted to contexts in which adult JA requires a marked concord form beyond default masculine singular agreement; noun phrases whose adult target was masculine singular were excluded. Accuracy in these marked contexts was lowest among children aged 2.0 to 4.11, remained similarly low in the 5.0 to 5.11 group, rose sharply in the 6.0 to 6.11 group, and exceeded 80% among children aged 7.0 to 10.11. The distribution of non-target patterns also changed with age. Early masculine singular overgeneralization gave way to mixed feature marking, followed by residual overgeneralization of plural concord in non-human plural contexts where feminine singular agreement was the established adult target in the relevant attributive construction. Within the mutually exclusive coding system, plural-concord overgeneralization was the most frequent residual pattern in the 6.0 to 6.11 group, whereas definiteness-only mismatches were the most frequent residual pattern in the oldest group. The study provides corpus-based evidence for differentiated development within the Arabic noun phrase while limiting its claims to the marked concord environments represented in the dataset.