In continuation of the theme that you can’t trust any of today’s dictionaries to reliably summarize what’s known about the etymology of a word, consider how they summarize the etymology of MASSAGE:

Merriam-Webster @ M-W.com: French, from masser to massage, from Arabic massa to stroke. First Known Use in English: circa 1860
Webster’s New World @ YourDictionary.com: Fr < masser, to massage < Ar massa, to touch
American Heritage @ YourDictionary.com: French, from masser, to massage, from Arabic masaḥa, to stroke, anoint; see mšḥ in Semitic roots or massa, to touch; see mšš in Semitic roots.
Collins English @ Dictionary.com: Century19: from French, from masser to rub; see mass [NOTE: at ‘’mass’’, ‘’mass’’ is stated to be from Latin ‘’massa’’]
Chambers Dict @ ChambersHarrap.co.uk:: 19c: French, from masser to massage, from Greek massein to knead. [question: directly modern Greek? or ancient Greek along unspecified path?]
Concise OED @ OxfordDictionaries.com: late 19th century: from French, from masser ‘knead, treat with massage’, probably from Portuguese amassar ‘knead’, from massa ‘dough’
Random House @ Dictionary.com: 1875–80; < F, equiv. to mass ( er ) to massage (< Ar massa to handle) + -age

Notice the absence of uncertainty: The first three say French ‘’massage’’ is certainly directly from Arabic, the fourth says the French is certainly directly from Latin, the fifth says the French is certainly from Greek (modern? or ancient?), the sixth says the French is probably directly from a Portuguese root, and the seventh, if I interpret its notation correctly, says the French is probably directly from Arabic.

The official dictionary of the French language says the French verb ‘’masser’’ meaning to massage is first attested in French in 1779—ref: CNRTL.fr, sup2—and the noun ‘’massage’’ is first attested in French in 1808—ref: CNRTL.fr. This dictionary too says the French comes directly from Arabic ‘’massa’’ meaning to touch, and it adds the remark: “The fact that the [early French] word appeared chiefly in accounts of travels in the Orient [the Middle East] seems to preclude the hypothesis offered by some that it came from Greek.”

All the dictionaries above that go with the Arabic ‘’massa’’ proposition say correctly that in Arabic ‘’massa’’ meant to touch—not to massage. The practice of massage was common in the Middle East for centuries before it started to become common in the West in the mid-to-late 19th century. But the Arabic word for massage was not ‘’massa’’ or anything close to it. The fact that the early evidences of the French word appear chiefly in accounts of travels in the Middle East, and yet these did not use the Arabic word for massage, seems to preclude the hypothesis that it came from Arabic.

Leonhard Rauwolf visited the Middle East in 1573-75, and published a 300+ page narrative of his visit in 1582. He says that massage was very common (search his book for the string “ bath"), and he has a detailed description of massage and of the bath houses where the massage took place. His book in English translation is downloadable at Archive.org (a massage at a bathhouse in Lebanon in 1573 is described on pages 37-38 of the djVu electronic copy, which is pages 20-21 of the print). The Middle Eastern bathhouses entered Western Europe in the mid-to-late 19th century under the name “Turkish bath”. In the Europe of the later 19th century you could get a “massage” at a Turkish bathhouse. Before then, the word massage was confined to accounts in travelers in the Middle East plus some attestations in a medical context.

The Arabic word for massage is تمسيد ‘’tamsīd’’ (also تدليك ‘’tadlīk’’). This Arabic word تمسيد ‘’tamsīd’’ contains the rootword مس ‘’mas’’ meaning to touch, to physically feel (such as with your fingers). The Arabic مسحة ‘’masha’’ (pronounced mas-ha) is much the same thing as ‘’mas’’. The root ‘’mas’’ or ‘’mas-ha’’ is what the dictionaries quoted above are saying is the source for the French ‘’massage’’. But ‘’mas’’ and ‘’mas-ha’’ do not mean massage at all. They are not tamsīd! Here’s how one of today’s Arabic-English dictionary defines ‘’mas’’: [Verb]: befall, contact, feel, finger, graze, handle, impinge, palpate, tip, twiddle, touch. [Noun]: touch. Similary the dictionary defines ‘’mas-ha’’ as: [Noun]: trace, touch, bit, streak, smack, tang, suggestion, tinge, shade, tint, hue. Arabic also has the verb مسح ‘’mas-h’’ or ‘’masah’’ defined as [Verb]: wipe, mop, mop up, rub, clean, polish, sweep, sponge, swab, extinguish, scrub.

I can find no explanation of how the various French travelers in the Middle East took up that or one of those Arabic words, and turned it into French ‘’masser’’, bearing in mind that the Arabs themselves didn’t use the word in that sense. The dictionaries appear to be relying solely on the phonetic similarity between an Arabic ‘’massa’’ and French ‘’masser’’.

I don’t know anything about the Greek ‘’massein’’ = “to knead” hypothesis.

In Turkish today the word for massage is ‘’masaj’’—how old is that, anyone? If it has been borrowed from the French, I’d have to regard that as weird conduct on the part of the Turks, because the Turks have been into massage for centuries before the French got into it, as evidenced in Leonhard Rauwolf’s account of Turkish-controlled Levant in the 1570s.

Archive.org has a Turkish-English dictionary dated 1852 (ref, page 788) containing the word مسح transscribeable as ‘’mas-h’’ or ‘’masah’’ or ‘’mess-h’’ and translated as “a rubbing with the inner parts of the fingers; the canonical mode of performing certain parts of the smaller ablution. To rub in that particular way.” That Turkish word is certainly a borrowing of the Arabic rootword given earlier, well attested as ancient in Arabic. Does Turkish grammar let you form a meaningful ‘’masaj’’ from a ‘’masah’’, anyone? Probably not, and the ‘-aj’ in ‘’massaj’’ is a dead-ringer for French ‘’-age’’. From what evidence I’ve seen, it’s equally justified and equally meritless to hypothesize that the French ‘’masser’’ came from the Turkish ‘’mas-h’’, rather than the Arabic ‘’mas-h’’.

It’s interesting to note that the first reported use of masser, masseur, masseuse is in reference to India, not the Middle East.  See here (pp. 128-129).  Thus,

“On assure que cette opération est nécessaire dans l’Inde, & facilite la circulation des fluides ... le massement rend les membres plus souples et plus agiles.”

This procedure is said to be necessary in India, and facilitates the circulation of fluids . . . massement makes limbs more supple and agile.

As for massage, its first reported use is in a French medical encyclopedia published in 1808 (pp. 522-523):

“Le massage est une coutume usitée en Inde, & généralement dans le Mogol chez les personnes qui observent la loi de Mahomet . . .”

Massage is a custom used in India, and generally in Mogul [Mongol regions] among those who observe the law of Mohammed . . .

On the surface it seem unlikely that an Arab word would be used for what appears to have been a common custom in India (unless of course the word came along with the custom from Arab lands to India).

Another high caliber post from madeira. Thank you.

OED on the verb “mass” v.3:

Now rare.
[< French masser to massage (1779) < Arabic massa to feel, handle, palpate (compare MASSAGE n.2).]
trans. To massage. Also (occas.) used intr.
1788 Ann. Reg. 1786 Misc. Ess. 119/1 A servant..then masses, and seems to knead the body without giving the slightest sensation of pain.

OED on the noun “massage” n2:

[< French massage (1808) < masser MASS v.3 + -age -AGE suffix. Compare earlier MASSING n.3
N.E.D. (1905) gives pronunciation with stress on the second syllable. The usual British pronunciation with stress on the first syllable was attested by Encycl. Dict. (1884) for the noun and by its Supplement (c1904) for the verb, and confirmed by pronouncing dictionaries of the early 20th cent. The pronunciation with stress on the second syllable remains usual in North America.]
1. The rubbing, kneading, or percussion of the muscles and joints of the body with the hands, usually performed by one person on another, esp. to relieve tension or pain; an instance of this. Also fig.
1866 D. MAGUIRE (title) The art of massage.

The fact that the early evidences of the French word appear chiefly in accounts of travels in the Middle East, and yet these did not use the Arabic word for massage, seems to preclude the hypothesis that it came from Arabic.

Words may sometimes be imported with meanings not exactly the same as their meaning in the original language. Again, I suggest submitting your theories to the dictionaries with which you take issue.  I doubt you will find many people outside dictionary circles who are sufficiently qualified to validate or refute your theories on Arabic etymology, though I feel certain you will find many who disagree that all the dictionaries you mention are unreliable etymological resources.

MASSAGE: Perhaps there exists some relation with the Hebrew root mem-shin-het - mashah (infinitive: limshoah) meaning “to anoint with oil” (cf. mashiah = Messiah, “The Anointed One")? (or again, perhaps there doesn’t ;-). BTW - does anyone know what the Roman word was for massage, or a masseur?

I think all Middle Eastern languages are full of loan-words from European languages, and vice versa. Any Hebrew (or Arabic) speaker today would know what massaj means. Etymologies may take very convoluted paths. Not long ago I asked an Israeli Arab greengrocer - what is the Arabic word for “artichoke?”, and he told me: “ ‘artishok’; we don’t have our own word for it, we borrowed the name from you” (meaning, from the Jews). Artichokes were only recently reintroduced into Israel, after a lapse of probably many years, possibly centuries. “Artichoke” is an English word derived (possibly, though not certainly, via French artichaud) from the Spanish alcachofa, derived in its turn from the Arabic al-harshuf or al-harshaf, perhaps a thousand or more years ago, when artichokes were brought from the Eastern Mediterranean to the West (probably by Arabs). The name, like the artichoke itself, has come full circle.

I wouldn’t look too hopefully for certainty, in the etymologies of European words which may have originated outside Europe. And I think that reezawaL is right, to be censorious when a dictionary makes unfounded assumptions (or copies them from others) and appears to state them without reservation. Are we not entitled to expect professional rigour from a lexicographer, no less than from an astronomer, or a biochemist, or any other scientific investigator?

I do agree that reezawaL’s posts sometimes tend to sound as though he held this forum somehow responsible for various dictionaries’ shortcomings - but I find them enjoyable and instructive nevertheless. Keep ‘em coming, reezawaL

bearing in mind that the Arabs themselves didn’t use the word in that sense.

That is completely irrelevant.  Words are very frequently borrowed in forms that don’t correspond with the form or usage in the original language.  And you continue to write as if you were the only person on earth who knows what they’re talking about, and the professional etymologists are bumblers and fools, which is off-putting.

Words are very frequently borrowed in forms that don’t correspond with the form or usage in the original language.

A parallel case being shampoo, derived from a Hindi word meaning - massage!

In continuation of the theme that you can’t trust any of today’s dictionaries to reliably summarize what’s known about the etymology of a word, consider how they summarize the etymology of MASSAGE:

I am assuming that this is the point of all these recent posts.  However, I think the point is irrelevant.

Etymology is just a hobby for me and I am certainly no lexicographer, but I do know that logic does not apply when it comes to how a word enters a language.  In this case, you seem to be arguing that massage could not have come from Arabic because the Arabic word for a massage is not massa.  A logical argument, perhaps, but irrelevant.  One can witness the same illogical word borrowing where English words enter into other languages.  One that immediately comes to mind is the German word handy, which means a mobile or cell phone.  This word has been borrowed from English, yet nowhere in the English speaking world is handy used to mean a mobile phone.  I have also seen Japanese examples, though none come immediately to mind.  As the handy example illustrates, words can follow odd routes when entering a language.

lionello - 21 November 2010 11:01 PM

“Artichoke" is an English word derived (possibly, though not certainly, via French artichaud) from the Spanish alcachofa, derived in its turn from the Arabic al-harshuf or al-harshaf, perhaps a thousand or more years ago.

Here’s what’s about artichoke at Wikipedia which I believe is correct:

الخرشوف ‘’al-kharshūf’’ with the meaning artichoke was used by for example Al-Razi (died 930).[14] Early Spanish carchiofa (1423), Italian carciofjo (circa 1525) are reasonably close to the Arabic precedent and so are today’s Spanish alcachofa, Italian carciofo. It is not clear how the word was corrupted to French artichault (1538), northern Italian articiocch (circa 1550),[5] English artochock (1591), but all of the etymology dictionaries say it is a corruption.

The corruption to the form ‘’artichoke’’ from ‘’kharshuf’’ or ‘’al-kharshuf’’ has no acceptable explanation, in my judgment. I grant it that it’s very probably a corruption. But the evidence leaves some room for uncertainty about the source of the artichoke word. If you’re going to abandon hard evidence-based reasoning, and if you don’t create clarity for yourself about what you do and don’t know, then what you’re left with is a story, a yarn. Some significant minority of the conventionally reported etymologies are yarns with inadequate evidence, and are not tagged as such. It would be better to tag a case like artichoke with “probably” or if you prefer “very probably”. Thereby when somebody reads that the Spanish ‘’carchiofa’’ is from the Arabic ‘’kharshuf’’ with no “probably” tag, they can trust it’s safe and sound to a high caliber.

I suggest starting a new thread for a new word.

The corruption to the form ‘’artichoke’’ from ‘’kharshuf’’ or ‘’al-kharshuf’’ has no acceptable explanation, in my judgment.

See the OED. The explanation is extremely complex. Such phonetic based changes are often not obvious, but rely on well-established patterns of sound variation over time and between languages. This is what separates the professionals from the amateurs. It simply does not do to look at the word and at its etymon and say “I can’t see how they got here from there, therefore it must be wrong.” (Or the converse, which is probably more common, “they look./sound alike, therefore they must be related.")

In short, the OED explanation is that the word was borrowed into Italian in the sixteenth century from either the Spanish alcarchofa or from the Spanish Arabic al-karsufa. It Italian, the initial element was reanalyzed based on the archi- (arti- in northern dialect) into articiocco. English took the word from Italian, reanalyzing the second element into the familiar -choke. In French, this second element was reanalyzed as -chaud, which in turn influenced the English word as well.

Such phonetic based changes are often not obvious, but rely on well-established patterns of sound variation over time and between languages.

This is a very crucial point. People with different native languages may pronounce the same word very differently (I remember reading about the difficulties Chinese had, when transcribing and pronouncing a name like Khruschev). The foreign expressions overheard by British soldiers who serve overseas are a case in point. The Arabic obscenities picked up in wartime in Egypt and other Middle Eastern countries by British soldiers, don’t sound at all like the Arabic originals when they are spoken by a Londoner or a Glaswegian, and would quite probably be unrecognizable to many an Egyptian who didn’t know English. I’m fairly sure the same applies to non-obscene words, though these are less familiar territory to me ;-). What reezawaL calls “corruption” (others on this forum prefer “change") can occur very quickly, and be very considerable indeed. For example, I remember the Chilean slang word for “beachcomber” when I was growing up there. It was michicuma. I see nothing surprising, or hard to believe, in “artichoke” evolving (perhaps by stages, as Dave suggests) from “al-harshuf”.....  Look at how the Spanish language itself alters words of incontestably Arabic origin: almotacén, Guadalquivir, alcalde....

What reezawaL calls “corruption” (others on this forum prefer “change") can occur very quickly, and be very considerable indeed.

A recent addition to my household is a Great Dane puppy.  I named him “Stavros” (Greek for “cross").  A few of my co-workers have trouble pronouncing his name and occasionally ask me how “Starbucks” is doing.  Its a perfect example of what you’re talking about.

Incidentally, I’ve noticed that those that have trouble with “Stavros” are exclusively male.  Female coworkers pick it up quite easily.

Edit: sp