Family Matters: Language Transfer and Merging for Adapting Small LLMs to Faroese

language modeling performance, linguistic accuracy, and text comprehension. To address the lack of existing Faroese evaluation resources, we construct two new minimal-pair probing benchmarks, one for linguistic acceptability and one for text comprehension, and complement them with human evaluations conducted by native Faroese linguists. Our results show that transfer from related languages is essential, but the optimal source language is task-dependent: Icelandic improves linguistic accuracy, while Danish boosts reading comprehension. The choice of adaptation method likewise depends on the target task: LoRA yields stronger linguistic acceptability and marginally higher human evaluation scores, whereas full fine-tuning produces better comprehension performance and more robust downstream fine-tuning. Merging multiple related languages under full fine-tuning (but not LoRA) improves general language modeling, though its benefits in the linguistic acceptability and comprehension probes are less consistent.