This type of analogy is specifically called series contamination, where items that frequently are pronounced together in a series affect each other. Also happened in proto-Germanic with the ‘n’ in ten, from the ‘n’ in nine (and in the opposite direction of Latin novem and decem; cf. nonus “ninth” to see the original ‘n’).
Also probably at least partially responsible for the pronunciation of Wednesday seen in the meme, based on Tuesday.
Edit: “ten” is not an example. See below for explanation.
Ah, yeah, quite possibly, good catch. I suppose it could be either assimilation to final -t before it dropped off, or final nasal merger to -n, a la Greek. I’ll check and report back.
Edit: Yeah, you’re probably right that it’s not series contamination, and is instead just assimilation. Ringe’s “From PIE to Proto-Germanic”, pg. 81, has the regular outcome of syllabic sonorants as *uX, so -um-, -un-, -ul-, -ur-.
So, while we’d expect dekumt and humda for “ten” and “hundred” (and may indeed have gotten them at some intermediate stage), we instead get tehun and hunda, making this pretty clearly regressive nasal assimilation of place. Note, e.g. *smH- ‘summer’ giving OE “sumor” instead, so when not before another consonant we don’t see the assimilation.
That’s what I get for doing off-the-cuff etymology, I suppose.
This type of analogy is specifically called series contamination, where items that frequently are pronounced together in a series affect each other. Also happened in proto-Germanic with the ‘n’ in ten, from the ‘n’ in nine (and in the opposite direction of Latin novem and decem; cf. nonus “ninth” to see the original ‘n’).
Also probably at least partially responsible for the pronunciation of Wednesday seen in the meme, based on Tuesday.
Edit: “ten” is not an example. See below for explanation.
Is the n in Germanic words for ten not just a reflex of m in PIE *dekm̥t?
Ah, yeah, quite possibly, good catch. I suppose it could be either assimilation to final -t before it dropped off, or final nasal merger to -n, a la Greek. I’ll check and report back.
Edit: Yeah, you’re probably right that it’s not series contamination, and is instead just assimilation. Ringe’s “From PIE to Proto-Germanic”, pg. 81, has the regular outcome of syllabic sonorants as *uX, so -um-, -un-, -ul-, -ur-.
So, while we’d expect dekumt and humda for “ten” and “hundred” (and may indeed have gotten them at some intermediate stage), we instead get tehun and hunda, making this pretty clearly regressive nasal assimilation of place. Note, e.g. *smH- ‘summer’ giving OE “sumor” instead, so when not before another consonant we don’t see the assimilation.
That’s what I get for doing off-the-cuff etymology, I suppose.