The path by which TAMARIND arrived in English
Posted: 26 November 2010 12:53 AM   [ Ignore ]
RankRank
Total Posts:  64
Joined  2010-11-02

English dictionaries are unanimous that the word TAMARIND has its origin in Arabic. I agree with them, but I find them in error on the matter of when the word arrived in English, and from where. Specifically six of the following seven English dictionaries are in error in their summary of the path from Arabic to English.

Americant Heritage @ YourDictionary.com: Middle English, from Old French tamarinde, from Arabic tamr hindī : tamr, dates; see tmr in Semitic roots + hindī, of India.
Merriam-Webster @ M-W.com: Spanish & Portuguese tamarindo, from Arabic tamr hindī, literally, Indian date. First Known Use in English: 15th century
Collins English @ Dictionary.com: 16th Century English: from Medieval Latin tamarindus, ultimately from Arabic tamr hindī Indian date, from tamr date + hindī Indian, from Hind India
Random House @ Dictionary.com: English 1525–35; < ML tamarindus ≪ Ar tamr hindī lit., Indian date
Webster’s New World @ YourDictionary.com: Sp tamarindo < Ar tamr hindī, date of India
Concise OED @ OxfordDictionaries.com: late Middle English: from medieval Latin tamarindus, from Arabic tamr hindī ‘Indian date’
Chambers Dict @ ChambersHarrap.co.uk: 16th century English: ultimately from Arabic tamr-hindi Indian date.

The UMich Middle English Dictionary is able to quote use of the word TAMARIND in three different Middle English sources, dated 1400, circa 1425, and 1475: See them here. All three are translations into English of medieval Latin medical texts. The year 1400 English is a translation of a Latin medical text by Lanfranc of Milan; the 1425 English is a translation of a Latin medical text by Guy de Chauliac; and the 1475 English is a translation of a Latin medical text by Gilbertus Anglicus. Therefore, the English is directly from Latin.

تمر هندي ‘’tamr hindi’’ appears as an entry in the 11th century Arabic medical encyclopedia “The Cannon of Medicine” by Ibn Sina aka Avicenna (ref). That book was translated from Arabic to Latin in the later 12th century, and became, I’m told, the most widely read and influential medical book in 13th century Europe. Therefore, the Latin is directly from Arabic. So out of the seven summaries in the dictionaries, only the ConciseOED’s is correct.

I argue you can’t trust any of today’s ordinary English dictionaries to reliably summarize what’s known about the etymology of a word. In earlier posts I discussed the examples caliber, cork, alizarin, genet, almanac, lilac, hazard, massage, racquet, sandarac, massicot, scarlet, and those plus tamarind add up to 13 words. All 13 are listed as Arabic-origin words in the Wikipedia page List of Arabic loanwords in English. That page lists altogether around 160 words, not counting Arabic cusine words, Latinate botanical words, and a few other words. Thus I’ve found an error rate of 13/160 = 8% when only counting situations where (a) all or most of the dictionaries are in error, and (b) I was able to notice it myself, and (c) the error is not small. In the course of the exercise I noticed various inconsistencies among the dictionaries concerning others of the 160 words looked up, implying some aren’t right.

Profile
 
 
Posted: 26 November 2010 03:14 AM   [ Ignore ]   [ # 1 ]
RankRank
Total Posts:  44
Joined  2010-09-07

The contradictions are less than they might appear at first sight.  In the period under consideration, nobody’s native language was Latin.  So when Avicenna’s text was translated from Arabic into Latin, it was simultaneously translated in the mind of the translator also into French or Spanish or Italian or Engish, etc.  To to take the example of the American Heritage Dictionary,

Middle English, from Old French tamarinde, from Arabic tamr hindī

it is by no means “incorrect” to say that Old French tamarinde comes from Arabic tamr hindī.  As for Middle English, one can’t rely only on 3 citations in the Middle English Dictionary.  The word was clearly known in Britain at least a century before the earliest of these citations, specifically in the form of a number of Anglo-Norman glossaries, e.g., from the 13th century

tamarindes, anglice sorel

This and other citations can be found in the Anglo-Norman Dictionary.  So who is to say whether English tamarind comes directly from Old French (of the Anglo-Norman variety) or Latin?

Profile
 
 
Posted: 26 November 2010 04:09 AM   [ Ignore ]   [ # 2 ]
Administrator
Avatar
RankRankRankRankRank
Total Posts:  4752
Joined  2007-01-03

Another consideration is when these dictionaries were published. The MED and the Anglo-Norman dictionary are relatively new and it is no surprise that older dictionaries would not reflect the most recently found antedatings. Random House, for example, shuttered their dictionary division over ten years ago. The OED entry for tamarind, dating to the 1989 second edition and perhaps unrevised since the nineteenth century, gives a 1533 first citation. In most cases, the edition used in the free online version is not the latest and greatest. You really need to look at the most recent print edition, or the online paid subscription version, to get the best scholarship.

And if, as Madiera points out, the word appears in Anglo-Norman glosses of the Latin, it is a judgment call on the part of the etymologist whether the path is direct from Latin or via Old French, as any English writer would be more familiar with the French than the Latin, even if translating from Latin. Also, bear in mind that the dates in the MED are manuscript dates. In many cases the work in question is considerably older.

Profile
 
 
Posted: 26 November 2010 04:56 AM   [ Ignore ]   [ # 3 ]
RankRank
Total Posts:  64
Joined  2010-11-02

Thanks to madeira for the pointer to the Anglo-Norman dictionary, which I hadn’t been aware of.

The Anglo-Norman dictionary’s entry for tamarind states no date, and gives to context, but indicates the word meant ‘’sorrel”, with a medical application. The Arabic & Latin “tamarind” word meant a fruit ("date of India") and that is not compatible with the word “sorrel”. What might’ve been meant in the Anglo-Norman context was the medieval Latin word “tamariscus”, which had medical uses, and is a better fit with “sorrel”. Latin “tamariscus” is modern English Tamarisk, which in all the modern online English dictionaries, and is also in the Middle English Dictionary.

Profile
 
 
Posted: 26 November 2010 05:20 AM   [ Ignore ]   [ # 4 ]
RankRank
Total Posts:  64
Joined  2010-11-02

Middle English Dictionary entry for “thamarike” / “tamarisci” quotes from a Latin-to-English translation of a 4th century Latin writer as follows:

“of thamaryke [L gloss: thamaride; L tamarisci, vrr. tamarici, thamarici] and other flouris wylde”

Surely that’s the word in the Anglo-Norman dictionary. Not the Arabic-origin tamarind. It’s the Latin-origin tamariscus spelled wrong.

Profile
 
 
Posted: 26 November 2010 05:54 AM   [ Ignore ]   [ # 5 ]
RankRank
Total Posts:  44
Joined  2010-09-07
ᴚǝǝƶɐʍɐɈ - 26 November 2010 05:20 AM

Middle English Dictionary entry for “thamarike” / “tamarisci” quotes from a Latin-to-English translation of a 4th century Latin writer as follows:

“of thamaryke [L gloss: thamaride; L tamarisci, vrr. tamarici, thamarici] and other flouris wylde”

Surely that’s the word in the Anglo-Norman dictionary. Not the Arabic-origin tamarind. It’s the Latin-origin tamariscus spelled wrong.

There are references which clearly state that what is involved is tamarind.  See for example the “Particulars of Odinet the Apothecary” from the time of Edward II here, Plant Names of Medieval England by Tony Hunt, and Hunt’s Anglo-Norman Medicine Volume I and Volume II in which the botanical item in question is unequivocally identified as Tamarindus indica L.

Profile
 
 
Posted: 26 November 2010 08:37 AM   [ Ignore ]   [ # 6 ]
Administrator
Avatar
RankRankRankRankRank
Total Posts:  4752
Joined  2007-01-03

When dealing with glosses you must look at the manuscript itself. Looking at a dictionary excerpt from an edited version does not cut it. Who added the gloss? Is it the original scribe? A later commentator? The modern editor of the manuscript? The lexicographer? You can’t tell from this.

Profile
 
 
Posted: 26 November 2010 08:43 AM   [ Ignore ]   [ # 7 ]
RankRankRankRankRank
Total Posts:  3507
Joined  2007-01-29

Thanks to madeira for the pointer to the Anglo-Norman dictionary, which I hadn’t been aware of.

And you’re trying to do etymology?

Profile
 
 
Posted: 26 November 2010 09:53 AM   [ Ignore ]   [ # 8 ]
Administrator
Avatar
RankRankRankRankRank
Total Posts:  4752
Joined  2007-01-03

And you’re trying to do etymology?

The problem is not trying to do etymology. Everyone has to start from a place of ignorance; there is nothing wrong with that. And I for one am happy to discuss etymology on this forum with those who are not well versed in etymological methods and sources. The problem is the certainty with which ᴚǝǝƶɐʍɐɈ assumes his conclusions are correct. He says “the dictionaries are wrong,” when he should be asking, “what, if anything, am I misunderstanding about the methods being used by the dictionaries?”

Profile
 
 
Posted: 26 November 2010 02:32 PM   [ Ignore ]   [ # 9 ]
Avatar
RankRankRankRank
Total Posts:  1370
Joined  2007-01-29

Or, if he has an infallible argument for a dictionary’s mistakes, he should be asking: what can I now do that would correct this mistake? or simply contacting the editors of the dictionary in question (as I have said earlier) rather than constantly telling us that dictionaries aren’t to be trusted with Arabic etymology. 

For heavens’ sake, who do we rely on in the meanwhile till the dictionaries get it right? Fawazeer, obviously.

Profile
 
 
Posted: 26 November 2010 04:07 PM   [ Ignore ]   [ # 10 ]
RankRank
Total Posts:  64
Joined  2010-11-02
madeira - 26 November 2010 03:14 AM

So who is to say whether English tamarind comes directly from Old French (of the Anglo-Norman variety) or Latin?

I accept madeira’s point since he delivered once again a high caliber link to back it up: “tamarind ii unces et ii unces de myrobalans” in the Anglo-Norman tranlation of Roger Frugard. (I still think that “sorrel” stinks). However, since the earliest attestations in English are directly translations from Latin, it is appropriate to summarily say the English is borrowed directly from Latin.

[ Edited: 26 November 2010 04:12 PM by ᴚǝǝƶɐʍɐɈ ]
Profile
 
 
Posted: 27 November 2010 04:46 AM   [ Ignore ]   [ # 11 ]
Administrator
Avatar
RankRankRankRankRank
Total Posts:  4752
Joined  2007-01-03

However, since the earliest attestations in English are directly translations from Latin, it is appropriate to summarily say the English is borrowed directly from Latin.

No, that does not necessarily follow. The earliest attestation is almost never the earliest use. There is nothing to say the word was not already familiar to the translator, especially when there are contemporary French uses as well. In this case, I would say a borrowing from Latin is the most probable explanation (and most of the dictionaries agree), but I would not dismiss an French route out of hand.

The strongest evidence for a Latin origin of tamarind that I see is the Latin ending in the translation of Lanfranc in Ashmole 1396, thamarindorum. This is earliest manuscript that uses the word in English. But I’d like to know a bit more about the history of the translation. Is Ashmole 1396 the original? Or is it a copy of an earlier translation? Are the other early citations from copies of older documents, making them potentially the oldest use? What are the dates and circumstances of the manuscripts with French uses?

The point is you simply cannot glance at the abbreviated dictionary entry for a manuscript citation and draw firm conclusions about what it represents without doing detailed work on the history of the manuscripts themselves. And in cases involving glosses, you often must actually consult the manuscript itself (not a transcription or facsimile) to draw conclusions about when the glosses were added (based on examining the ink, etc.)

Profile
 
 
Posted: 27 November 2010 07:23 AM   [ Ignore ]   [ # 12 ]
RankRankRankRankRank
Total Posts:  3507
Joined  2007-01-29

Everyone has to start from a place of ignorance; there is nothing wrong with that. And I for one am happy to discuss etymology on this forum with those who are not well versed in etymological methods and sources.

Absolutely.  What I should have said was “And you’re presuming to look down on professional etymologists, who actually know the sources and methods?”

Profile
 
 
Posted: 27 November 2010 08:01 AM   [ Ignore ]   [ # 13 ]
RankRankRankRank
Total Posts:  829
Joined  2007-03-01

In earlier posts I discussed the examples caliber, cork, alizarin, genet, almanac, lilac, hazard, massage, racquet, sandarac, massicot, scarlet, and those plus tamarind add up to 13 words. All 13 are listed as Arabic-origin words in the Wikipedia page List of Arabic loanwords in English. That page lists altogether around 160 words, not counting Arabic cusine words, Latinate botanical words, and a few other words. Thus I’ve found an error rate of 13/160 = 8% when only counting situations where (a) all or most of the dictionaries are in error, and (b) I was able to notice it myself, and (c) the error is not small.

It’s small compared to some of the other “lists of English words of foreign origin” on Wikipedia, which are awash with folk-etymology and wishful thinking. Consider, for example, the page of “English words of Irish origin” . Although it states that “This is a list of English language words from the Celtic Irish language”, among the 41 words and phrases it contains are several that blatantly are nothing of the kind (e.g. boycott, craic, gob); several whose etymologies are debatable (kybosh, slapper), a good number that are at least as likely to come from Scots Gaelic as from Irish Gaelic (banshee, loch, bog, glen, galore, whisk(e)y), and a few that can hardly be said to be “English language words” at all (e.g. Shan Van Vocht).

This is hardly remarkable: indeed, given the popularity of folk-etymology in the internet it’s exactly what one should expect. The contents of any of these pages are completely irrelevant to your argument.

Profile