Not clear that it comes from Arabic
According to my research, the Greek word ἐλέφας is not related, at least not clearly or directly, with Arabic الفيل . I looked up four or five etymologycal books and all of them said basically what is stated in this website. So, it is probable that it comes from a Semitic word but not necessarily Arabic al-fil.
In the etymology of this word there is as well a hidden commentary that goes: "<!--, Egyptian [[Ab]] ... how is pil/fil related to Ab?-->". The answer is "not to pil/fil" but to "ebur". According to my research, some connected the origin of Latin ebur (ivory) with Greek ἐλέφας but the connection cannot be established satisfactorily . It is ebur, and not ἐλέφας, the word which is related to Egiptian āb/ābu < Coptic ebou/ebu < Hebrew (shen-)habbim (tooth of) elephants < Sanskrit इभ (íbha[s]) (eventually back to an Indo-European root). What it is not clear is the fact that "ebur" and "elephas" share one common origin. So, I am going to hide the reference to Arabic and Persian and state that it probably comes from a Semitic language.--Piolinfax 01:58, 22 October 2006 (UTC)
- Hello, Germany calling! :) First of all, there is a very good small dictionary of German words of Arabic Origin - Nabil Osman: Kleines Lexikon deutscher Wörter arabischer Herkunft, Beck'sche Reihe 1982 (I use the 3rd edition of 1993). Osman is listing even words, which are archaic, rare or technical terms of Islamic studies. But "Elefant" is not mentioned there at all! This is a pretty good proof, that this Arab scholar in Germanistics came to the conclusion, that this word is not coming from Arabic.
- Second, I can report, what the Etymology Duden says about Elefant: Middle High German elefant, Old High German elpfant, elafant and also helfant, folk etymology from helfen (to help). (The elephant was considered as a helpful animal, which got its name from the tooth, which were much longer known, than the animal self) < Latin elephantus < Greece eléphas (Genitive eléphantos), < Egyptian âb(u) < Coptic eb(o)u > > Latin ebur.
- Can it be, that this folk etymology with "help" also applies to English, if this phenomen played a role? -- Arne List 20:04, 21 May 2007 (UTC)
Maybe it's "Aleph".
- You mean the (Central) Semitic word for "ox", *ʾālp. That's a possibility scholars have considered, too, but it doesn't explain the Greek ending. Perhaps the Greek word is a kind of contamination of the Egyptian word with something. But with what? It might be a compound with the Berber word, but such a bilingual compound would be very unusual. (Don't forget that the Greek word is already attested in Mycenaean, and originally seems to have had only the meaning "ivory". That and the following might make the suggestion sound somehow plausible, as some kind of "trade word", whatever that means precisely; but one annoying complication is that Berber is attested so late and we don't know what it was like in the Late Bronze Age: a Proto-Berber form could possibly be reconstructed, but that would probably take us back only to the 1st millennium BC, a bit late for our purposes.)
- Chantraine, interestingly, mentions Hittite lahpaš "tusk; ivory" and says this must itself be a borrowing. He rejects the connection with Sanskrit íbha- (which, according to Mayrhofer, may not be old, but a semantic specialisation of a Vedic word) suggested by others. I've heard of a tongue-in-cheek etymology "(s)he who says elé" (with elé supposedly an imitation of the typical "trumpeting" sound), which even kind of works and might actually have influenced the form of the Greek word at least as folk etymology. :-) --Florian Blaschke (talk) 02:40, 28 August 2013 (UTC)
- As for Arne's question, a folk-etymological transformation, either independently or inspired by Old High German helfant, does seem plausible on the face of it (but only for Old English, not Middle or Modern English), considering that the expected outcome would be *elfent – were it not for the fact that the Old English word is elpend, not *helpend, compare hǣlend (> healand) for the ending. But perhaps this form existed too, but was not recorded (or it was and we do not mention it; at least according to Bosworth-Toller, helpend is only recorded in the expected meaning "helper", not "elephant"), and turned into elpend by monks or other people familiar with the Latin source.
- Oh, and considering that Arabic fīl must be a post-turn-of-the-eras borrowing from Middle Persian pīl (probably not a direct continuation of Old Persian pīru- "ivory", but perhaps a borrowing from Sogdian or Syriac), compare Encyclopædia Iranica, you can forget it anyway. Never mind that fīl, even with the article, does not even resemble the Greek word – some hobby etymologists are really desperate. --Florian Blaschke (talk) 15:22, 28 August 2013 (UTC)
- There is a false lead which is, however, still helpful as a demonstration of the nature of serious etymology as an ultimately scientific endeavour, following scientific method. In a footnote in the Lexikon der Indogermanischen Verben, I discovered that there exists an Ancient Greek verb elepháiromai "I puzzle, riddle; I cheat", and this is explained as deriving from an unattested r/n-heteroclitic which would have resulted in *elephar in Attic. One might consider the possibility that a different derivation from the same noun (with a -t suffix) would have resulted in a word **elephant- in Greek (not sure about the accent, but **eléphant- does not look obviously wrong to me), which would have originally meant something like "riddler" or "cheater". This would imply a most interesting semantic shift, considering the culture-historical background which might underlie it. One might speculate that the word originally referred to a mythological being, a kind of trickster, or even specifically a sphinx, and that its meaning was eventually (perhaps because both could be associated with Egypt?) transferred to the animal. The formal morphological derivation is indeed flawless, as a competent friend has assured me.
- However, there is a complication: There are words in Sanskrit and Lithuanian which must be derived from the same verbal root, and these words start in v- in both languages. This entails a prediction about the Greek word for elephant: If the hypothesis (the derivation from the heteroclitic) just pondered is true, the word must necessarily have started with a digamma, i. e., with w-. (This sound has disappeared in Classical Attic, but is still found in other dialects and spelt as the letter digamma there, which looks like the Latin F, which ultimately derives from it, in fact.) The prediction can be tested: All we need is some direct (from dialects which have retained the sound) or indirect (other traces for word-initial digamma, which Homeric Greek often provides) evidence that the word originally started in w-. The word is actually attested in Mycenaean Greek, after all. We know that Mycenaean Greek has retained the w- word-initially and elsewhere, and that in the Linear B script which Mycenaean Greek was written with, it can be distinguished whether a word started with a w- or not. If the hypothesis is true, the word should be spelt **we-re-pa-..., otherwise it is refuted. The spelling we actually find is e-re-pa-te-jo "made of ivory", and this means that the hypothesis is disproven.
- This one single flaw is decisive: there is no way around it, the hypothesis is unsalvageable. Were it not for the missing w-, the hypothesis would still be defensible, at least on formal grounds, despite the fanciful semantic connection, and the unusual shift required. However, to bolster the hypothesis, it would have been necessary to look for parallels, similar semantic shifts, or associations made between elephants and what could be described as "riddlers" or "cheaters" in either the relevant or other cultures. But alas, the single formal flaw means that there is no use to pursue the idea any further.
- As for the hypothesis explaining the word as originally meaning "he who says 'elé!'", it does look more attractive. Granted, to my ears, the elephant's trumpeting does not sound very much like "elé!", but who knows what people back then heard, and what language was spoken by those who would have come up with the supposed onomatopoeia originally (if the word does contain such an onomatopoeia, of course). It need not have been the Greeks anyway, and the Greeks would not even notice if an elephant's trumpeting does not sound like "elé!" in the first place if (at least in the Bronze Age) they would never encounter the live animal anyway, only its tusks, which would also give a certain credence to the idea above that they may have thought of it as a mysterious (puzzling or deceiving) mythological being whose appearance they were not sure of. But that's a lot of ifs anyway.
- One weakness both explanations share is the assumption that the word originally referred to the animal in the first place, and not ivory, even though the Mycenaean word is only attested referring to ivory.
- Ultimately, therefore, it is more plausible that the word is a loan, quite possibly from the same source as the Hittite word, and perhaps this source is actually a compound involving elephant words from different languages, even though a hybrid compound like this would certainly seem unusual (though not impossible). --Florian Blaschke (talk) 01:08, 20 September 2013 (UTC)
Elephant in crinoline
Isn't there another victorian era meaning to "elephant": a woman, usually an elderly one, who looks over young lovers outdoors, so the unmarried couple does not engage in some extremely gross activity, like touching each other, God forsake kissing!