The third layer of loanwords evaluated in this work clearly represents the oldest. It testifies to a prolonged contact between Armenian and one or more unclassified, non-IE languages. This234 5. Evaluation and outlook contact event predated all or most Armenian sound changes. Crucially, most other Indo-European languages, with the exception of Anatolian and Tocharian, were to some degree in contact with the same stratum. These facts, taken together, suggest that this period of language contact must have begun relatively shortly after the dissolution of the Core Indo-European languages. Therefore, it most likely represents contact between speakers still residing near theIndo-European homeland, and speakers of those languages neighbouring them. It seems clear that Armenian, Greek, and Albanian remained in close contact with the same language(s) for the longest period of time. This is consistent with the data showing that these languages shared innovations on the basis of inherited material as well (Matzinger 2012, Lamberterie 2013, Olsen & Thorsø 2022).At the same time, there is also a considerable overlap between non-Indo-European vocabulary in Armenian and that found in Germanic, Italic and Celtic. Among these loanwords are terms for agricultural crops, like ‘barley’ and ‘some pulse’ (> Arm. ‘alfalfa’),indicating that Proto-Armenian existed within the core of Indo European languages whose speakers migrated Westward across the steppe and went through a gradual transition from a completely herding-based economy to a more sedentary culture with elements of agriculture, starting from around 3300 bce (cf. Kroonen et al.2022). Nevertheless, Armenian does not share as much foreign agricultural vocabulary with Germanic, Italic, and Celtic as these languages do with one another. Thus, there is reason to believe that its speakers did not take part in those population movements that later gave rise to the Corded Ware and Bell Beaker cultures in Europe. Again, given that population movements around 2000 bce are a plausible vector for the movement of Proto-Armenian speakers into the Caucasus, it is tempting to preliminarily locate these Proto Armenian speakers somewhere in the Late Yamnaya and perhaps in the Catacomb culture, which emerges from Yamnaya starting around 2500 bce. Future studies combining linguistic, archaeological, and genetic evidence will hopefully be able to confirm or reject this hypothesis.