Novel labels provide feedback that may enhance categorical alignment between interlocutors. However, the nature of this feedback may not always be linguistic. Lupyan (2008) has demonstrated the effects of labels on individual categorization, and even non-word labels have seemingly produced greater consistency in sorting strategies (Lupyan & Casasanto, 2014). We extend this to alignment by demonstrating that arbitrary labels can increase sorting consistency to bring people’s categories closer together, even without dialogue. Importantly, we argue that increased alignment is not always due to labeling in a linguistic sense. Results suggest that it is not the content of the non-word labels driving the alignment effects, but the very presence of the labels acting as ‘anchors’ for category formation. This demonstrates a more general cognitive effect of arbitrary labels on categorization.