Crocodile vs. cocodrile: where does the "r" belong?

Question

We are all familiar with crocodiles. We know, love, and recognize them in many European languages:

• German: Krokodil
• French: crocodile
• Portuguese: crocodilo
• Russian: крокодил

But perhaps it comes as a surprise that other modern languages put the "r" in a different spot:

• Italian: coccodrillo
• Spanish: cocodrilo
• Catalan: cocodril

The OED gives us the following colorful etymology, which shows all these forms emerging from a corrupt Medieval Latin cocodrillus--and, at least in some cases, they were corrected back!

Middle English cocodrille, cokadrill, etc. < Old French cocodrille (13–17th cent.) = Provençal cocodrilh, Spanish cocodrilo, Italian coccodrillo, medieval Latin cocodrillus, corruption of Latin crocodīlus (also corcodilus), < Greek κροκόδειλος, found from Herodotus downward. The original form after Greek and Latin was restored in most of the modern languages in the 16–17th cent.: French crocodile (in Paré), Italian crocodillo (in Florio), Spanish crocodilo (in Percival).

I thought this completely answered my question, but I found at least one competing etymology that suggests that the "-drilo" does belong! The The Merriam-Webster New Book of Word Histories --which is echoed also by the etymology given in a Google word search--says the following:

[alteration (influenced by Latin crocodilus) of Middle English cocodrille, from Old French, from Medieval Latin cocodrillus, alteration of Latin crocodilus, corcodillus, from Greek krokodeilos, krokodilos, alteration of (assumed) krokodrilos, from krokē pebble + drilos worm]

(abbreviations expanded)

I have a lot of related quetions:

• Is the latter etymology just wrong about the supposed origin in Greek "krokodrilos"?
• If not, is it pure happenstance that the Medieval "corruption" of crocodilus happened to move the "r" to a place that it existed originally?
• Most importantly: Why shouldn't we start pedantically over-correcting to "crocodriles"?

Answer 1

The "pebble-worm" derivation seems rather doubtful. Of the etymological dictionaries, Frisk and Chantraine accept it, but Beekes says it "should be forgotten". I can't find any direct evidence of a word δρῖλος meaning "worm"; here's what I've found:

1. Hesychius lists a word δρίλαξ meaning "leech", which is pretty close.
2. δρῖλος itself appears to be a δὶς λεγόμενον: it appears once in an inscription at Amphissa to which I don't have access at the moment (Supp.Epigr. 2.353), and once in the Anthologia Palatina, in this rather opaque couplet:

ἤθελε ‘ΔΡΙΜΥΣ’ ἄγαν τὸ πρόσθ᾽ Ἱερώνυμος εἶναι:
νῦν δὲ τὸ ΔΡΙ μὲν ἔχει, ΛΟΣ δὲ τὸ ΜΥΣ γέγονεν.
"Drimus once wanted too much to be Hieronymus: now he has the Dri-, but the -mus has become -los".

Chantraine adds that δρῖλος is glossed in Latin as verpus, which L&S define as "a circumcised man". I'm not sure how you get from any of this to "worm" (though verpus is pretty close to verpa, which means "penis"). Maybe the Amphissa inscription would explain things.

In any case if the word really was originally κροκόδριλος then a dissimilatory loss of the second ρ would be very plausible, as would a subsequent unrelated metathesis in medieval Latin / Spanish etc.

To your last question I'd say the obvious answer is "because you'd be stared at like a crazy person".

Answer 2

My guess is the following: the appearance and disappearance of liquid consonants is very common among languages, for instance, latin parabola became palavra in portuguese, and mīrāculum became milagre, in both these cases by a process we call "metathesis". The word indeed came from krókē + drîlos and it changed by means of making its articulation easier. So krokódeilos became one greek form and it was borrowed by latin speakers and it became crocodīlus. A later latin form, changed by making the articulation easier, a spontaneous process, was *cocodrillus, and it generated the descendants you showed. The forms like crocodilo and crocodile were borrowed later from latin.