This article proposes the "scalpel model," a new model of third and additional language (L3/Ln) acquisition. The model aims to identify and examine what happens beyond the initial state of acquisition and what factors may influence change from one state of knowledge to another.

The article briefly examines the currently proposed hypotheses and models and evaluates the existing evidence for their predictions. It highlights several cognitive and experiential factors affecting crosslinguistic influence that are not taken into account by the current models. These factors include: structural linguistic complexity; misleading input or lack of clear unambiguous evidence for some properties or constructions; construction frequency in the target L3; and prevalent language activation or use.

Findings of recently published research are discussed in support of the scalpel model. In particular, findings of differential learnability of properties within the same groups of learners suggest that L1 or L2 transfer happens property by property and is influenced by diverse factors.

The scalpel model explicitly argues that wholesale transfer of one of the previously acquired languages does not happen at the initial stages of acquisition because it is not necessary. It also argues that transfer can be from the L1 or the L2 or both, but it is not only facilitative.

The new model increases the explanatory coverage of the current experimental findings on how the L3/Ln linguistic representations develop.

The model emphasizes the importance of the cognitive, experiential, and linguistic influences on the L3/Ln beyond transfer from the L1 or L2. Thus, it aligns L3/Ln acquisition with current debates within L2 acquisition theory.