Happy to chat! On zoom (https://zoom.us/j/116970507, Friday's poster session), or on twitter (@MKeshev)! **Abstract:** Recent studies suggest that comprehenders engage in rational noisy-channel inference to accommodate the possibility of production and perception errors. We examine two possible factors that may contribute to “noisy” representations of linguistic input: Low-level orthographic errors, and linguistic errors in morphosyntactic features. We test this within a subject/object relative ambiguity in Hebrew, and replicate previous findings suggesting that readers opt for a noisy SR analysis over a faithful but rare OR. Yet we failed to observe significant effects of featural and of orthographic distance on the tendency to choose a noisy analysis over a faithful one. ---------- **Want to read more?** This poster follows up on our study *Noisy is better than rare: Comprehenders compromise subject-verb agreement to form more probable linguistic structures* Preprint (under review): https://psyarxiv.com/6rdw2/ Data&code: https://osf.io/v6y8n/