हकारोत्तरानुनासिकाः

उच्चारणभेदः

केचन ब्राह्मण, वह्नि…, ग्रुह्णमि इत्यादीनां वचने हकारानुनासिकयोः क्रमम् परिवर्तयन्ति।

पाणिनीयकाले

As best as I know, historically all these clusters are h-initial clusters. The reasons are not hard to see. Just as -jñ- in forms like rājñaḥ is a contraction of -jan- in rājan, similarly -hn- in ahnaḥ is a contraction of -han- in ahan. - MD

अर्वाचीनकाले वर्णपरिवर्तने प्रभावः

  • शिष्टाचारः
  • कौण्डिन्यशिक्षा - हकारान् नणमा ऊर्ध्वाः पूर्वं तु प्रसरन्ति हात् ।
  • Indirect evidence cited by some
    • हकारान् नणम-परान् नासिक्यम् ॥ तैत्तिरीयप्रातिशाख्यम्, २१-१४ ॥ (After h, when followed by n, ṇ, or m, is inserted nāsikya.)
    • हकारान् नासिक्येन ॥ अथर्वप्रातिशाख्यम्, १-१०० ॥
    • Paniniyasiksha talks about an eighth place of pronunciation - उरः. Sistas say that the हकार in examples cited by you is the only letter which is औरस्य. हकारं पञ्चमैर्युक्तम् अन्तस्थाभिश्च संयुतम् । औरस्यं तं विजानीयात् कण्ठ्यमाहुरसंयुतम् ॥ पाणिनीयशिक्षा, १६ ॥ (Pāṇinīyaśikṣā distinguishes between urasya and kaṇṭhya pronunciations of -h- in different contexts, but does not strictly speak about metathesis of -h- )

However, the metathecised pronunciation sets in very early. It is certainly the norm in the Marathi region today, and most likely this is a very old phenomenon. This is also a common error in manuscripts. - MD … In my article “Linguistic Presuppositions of Pāṇini 8.3.26-27” [pp. 23-42, in Proceedings of the International Seminar on Pāṇini (held in July 1981), ed. by S.D. Joshi and S.D. Laddu, Publications of the Centre of Advanced Study in Sanskrit, Class E, No. 9, University of Poona, Pune, 1983], I have discussed Pāṇini’s treatment of the anusvāra before hm and hn, and have noted that Kātyāyana extends this pattern to hy, hv, hl. I have shown that Pāṇini’s treatment restricted to hm and hn, is due to specific linguistic environment of the northwestern frontier region, while Kātyāyana’s extension of the same treatment to hy, hv, hl, is most likely due to the Prakritized pronunciation of Sanskrit, where all h-initial clusters of Sanskrit undergo metathesis+++(=interchange)+++. This metathesis is widely documented in inscriptional and literary Prakrits from Ashokan inscriptions onwards. The fact that both Pāṇini and Kātyāyana formulate their rules (he mapare maḥ, napare naḥ, Kātyāyana’s Vārttika: yavalapare yavalā vā) as option rules probably indicates that the metathecised pronunciation of the h-clusters was still not universal, but only optional. I don’t have a scan of my article, but it should be available in libraries in India. - MD

सम्बद्ध-चर्चाः