This talk discusses the forms in -nd which occur besides forms in -ng in Middle Persian (3rd c. BC to 7th c. CE), Early New Persian (8th c. to 9th c. CE), and Classical Persian (10th c. to 18th c. CE). I will argue that these forms owe their (apparent) existence to misinterpreting the highly ambiguous script in which Zoroastrian Middle Persian (Pahlavi) was written. Examples include: piring/parand ‘silk’ aurand/aurang ‘throne, glory, beauty,’ kuland/kulang ‘pickaxe,’ dērang/dīrand ‘world, time.’ The Middle Persian phonemes /d/ and /g/ were written with the same grapheme in the Pahlavi script. One could differentiate them by diacritic marks, but scribes rarely make use of this. Forms like parand (which is in contrast with the original form containing -ng) could thus be due to a misreading by people who could read Pahlavi, but not speak Middle Persian fluently. The other abovementioned words could be explained in the same way. Although both forms in -ng ~ -nd are evident, no Persian grammar mentioned them. Moreover, dialectal forms are mostly in contrast with Persian forms, strengthening the idea that literary texts developed new forms (learnéd forms) by misinterpretations of the Pahlavi script. For example, Borhān-e qāṭeʿ (17th c. CE) described parand as a silk-made garment and parang as a sword’s light reflection. This study provides a new view by interpreting the Pahlavi script and attested forms in Iranian languages plus considering forms in Iranian contact languages to illustrate the aforementioned consonant clusters.