This page is a critique of the historical content of two books by François Cardarelli: Scientific Unit Conversion (1997) and its successor Encyclopaedia of Scientific Units, Weights, and Measures (2003). Both books cover units used in science now and in the past, and also units that have at one time been used for commercial or other nonscientific use. The books are widely cited for Wikipedia articles, but questions have been raised as to the factual accuracy of various statements made in one or other. The current purpose of this page is to assess its appropriateness as a source for articles in Wikipedia, particularly for historical and other non-SI/metric units.
Since I discovered that large amounts of Cardarelli (including the Japanese howlers) are copied more or less verbatim from the earlier book by Washburn, and since there are a number of Wikipedia pages whose only sources are Cardarelli, Washburn, and an even older book by Clarke (including a quite different confusion of Japanese units), I am expanding this to try to assess all of these references.
I do not have a copy of the book itself, and do not plan to buy one. (As I write this in January 2015, the cheapest copy at abebooks.com of the newer book costs $124 plus postage.) However, Cardarelli generously makes many sample pages available on his website, and a couple of other Wikipedians have copies of one book or other. I refer to the newer book in the customary way by the name of its author: "Cardarelli". Where there are errors, I have no way of knowing whether these were introduced by others; so no comment on "Cardarelli" (the book) should be taken as relating personally to Cardarelli the man.
If you can contribute, please edit this page. If you think all or part is Entirely Wrong, or have any other comments, please use the talk page. Of course, any positive information is also extremely welcome. It would not be surprising, for example, to discover that the description of old French units is both comprehensive, accurate and useful. More green bits below are my notes to myself, or cues/requests for help in filling this out.
Thanks. Brian Chandler (Imaginatorium)
The second appears to be the successor of the first, even though the titles are different, the two containing for example some of the same howlers. The publisher is (now) Springer Science+Business Media, ultimately the successor to the venerable Springer Verlag, and formerly[1] highly reputable. "M.J. Shields, FIInfSc, MITI" is credited as translator on the title page of the earlier book, but there is no obvious sign of an original publication in French or Italian (either of which might be Cardarelli's native language).
You can read some sample chapters on Cardarelli's own website, www.francoiscardarelli.ca [1]
The back cover of Scientific Unit Conversion reads in part: "The pocket book format has been chosen for its compact size", but this book would require a sturdy pocket: height 235 mm, width 124 mm, thickness 20 mm. The book has xvi+416 pages; by far the largest single part of this, running from pp. 91–349, is of "Units in Alphabetical Order". This consists of a table, with the column headings "Unit (synonym, acronym)", "Symbol, "Physical quantity", "Dimension", "Conversion factor (SI equivalent unit)", "Notes, definitions, other conversion factors", "System". A single example (with slashes "/" substituting for line breaks) should give the idea: "Unit (synonym, acronym): dyne per square centimetre (barye, microbar, barrie)", "Symbol: dyn.cm−2", "Physical quantity: pressure, stress", "Dimension: ML−1T−2", "Conversion factor (SI equivalent unit): 1 dyn.cm−2 = 10−1 Pa (E)", "Notes, definitions, other conversion factors: Obsolete cgs derived unit of pressure and stress. / 1 dyn.cm−2 = 1 barye (E) / 1 atm = 1 013 250 dyn.cm−2 (E)", "System: cgs". (I [Hoary] suppose that "(E)" denotes "exact", but can't find this explained anywhere. The earliest appearance of "(E)" that I notice is on p. 23.)
The back cover of Scientific Unit Conversion also tells the reader that "Accurate metric equivalents and conversion factors are given for more than 10,000 scientific units with detailed descriptions of over 2,000." The notion that there are over 10,000 scientific units is surprising.
This [Encyclopaedia of Scientific Units, Weights and Measures] is a huge book, "with more than 5,000 units and 35,000 conversion factors" according to the blurb. Both titles include the word "Scientific", and a great bulk of the book is a recitation of current standards for SI units.
There are plenty of independent reviews, which are generally very positive.
This suggests that it is reasonable to assume that except for possible typos and awkwardnesses of English from a non-native speaker, the bulk of the book, on current SI and other scientific units is reliable, moreso since this material can be transferred from any number of currently available sources. But the reviews then typically just mention the presence of "some delightful 'obsolete units'" which might be a "godsend for compilers of crosswords". None of them look critically at this part of the book, and see how it matches up with the OED, for example. It can be said with some certainty that no reviewer with a knowledge of Japanese who looked properly at the book would be able to give a glowing report.
Pending
The problems come on the fringes. The book also covers non-SI units, most obviously the imperial and US customary systems, but also obscure and obsolete units. Let's start with the worst single example I have yet found.
None of the material in this section of Cardarelli has made it into Wikipedia, so this is not a direct problem. But the fact that this contains some complete nonsense was what first alerted me to the problems with Cardarelli as a reference. It emerges that the problem text is copied verbatim from a much earlier book: Washburn, E W: International critical tables of numerical data, physics, chemistry and technology (1926)[2] (link not currently working). To give some more background, I also list the (completely different!) confusion in a 19th century text: Clarke, F W: Weights, measures, and money, of all nations (1875)[3], and also the very accurate version by Basil Hall Chamberlain published in 1905[4].
This table compares the names in Cardarelli Table 3-262, Old Japanese units of weight with what is said in the articles Japanese units of measurement and Japanese numerals.
Note: The Cardarelli version is reproduced exactly, including the line-breaks.
Wikipedia | Cardarelli | Japanese | Notes | ||
---|---|---|---|---|---|
Romanized | Kanji | momme | |||
糸 | 0.0001 | Shi | shi | 糸 is the modern way to write a character whose standard form used to be 絲. In common with mō, rin and fun (see immediately below), shi was a multiplier within either of two, incompatible systems. Relevantly here, it was a multiplier of 1/10 000. (In the second system, it was a multiplier of 1/100 000. Within this second system, mō, rin and fun too were multipliers of one tenth the values stated below as primary.) | |
毛 | 0.001 | Mô | mō | 毛 is a traditional character whose uses are irrelevant here; it is also the modern way to write a character whose standard form used to be 毫. This, pronounced mō, was a standard 1/1 000 multiplier (but also a 1/10 000 multiplier). Koizumi's Zukai: Tan'i no rekishi jiten[5] says that it was used as a unit of mass. "Mô" is not standard (Hepburn) romanization (though use of the circumflex for the Hepburn macron is understandable, unambiguous, and commonplace). | |
厘 | 0.01 | Rin | rin | 厘 is both a traditional character and a modern way to write 釐; both 釐 and 厘 were used for rin, a standard 1/100 multiplier (but also a 1/1 000 multiplier). Koizumi's Zukai: Tan'i no rekishi jiten says that it was used as a unit of mass, and also that 釐 or 厘 had alternative pronunciations. | |
fun | 分 | 0.1 | Candareen (fun) |
fun, bun or bu | The Japanese pronunciation depends on the context. Candareen is from Malay; it is not a Japanese term. |
momme | 匁 | 1 | Mommé | monme / momme | The accented "é" is a nonstandard but widely used way to emphasize that the letter is pronounced (that the word does not rhyme with French homme). Whether written as momme or monme (flavours of Hepburn differ), there is no [n] sound. |
4 | Niyo | ?? | No obvious meaning. (Neither niyo nor niyō appears in Koizumi's Zukai: Tan'i no rekishi jiten.) Looks a bit like ryō (両), which is 10 momme (but, confusingly, was also a unit of length). | ||
hyakume | 百目 | 100 | Hyaku-mé | hyakume | Not standard romanisation (though a common and easily understandable way to avoid the risk of having the word mispronounced to rhyme with "plume"). This is literally "100 [me]", using the character 目 (=eye), and is a form used for 100 monme cited to Daijirin. |
kin | 斤 | 160 | Kin | kin | |
kan or kanme | 貫, 貫目 | 1000 | Kwan | kan or kanme / kamme | Kwan is an old pronunciation of kan. Kanme (written either 貫目 or 〆) was a variant intended to avoid confusion with various homophones of plain kan. Depending on the particular flavour of Hepburn romanization, perhaps written kamme instead of kanme; either way, there is no [n] sound. |
7000 | Ninsoku- ichi-nin |
人足一人 | Ninsoku is an Edo-period word for a labourer, so this could be translated "One labourer". There is no evidence of this regarded as a unit. (No term starting either ninsoku or ichinin appears in Koizumi's Zukai: Tan'i no rekishi jiten.) | ||
16,000 | Kiyak-kin | hyakkin | Recognisable misprint for hyakkin (=100 kin). This is not a unit: the word for 'hundred' (hyaku) combines with the following 'kin' to form a geminated consonant. (None among kiyakkin, kyakkin, hyakukin or hyakkin appears in Koizumi's Zukai: Tan'i no rekishi jiten.) | ||
18,000 | Karus hiri- ichi-da |
Karushiri? ichi da | (gibberish) Japanese phonology does not permit a syllable to end in /s/; if the space between "karus" and "hiri" is closed up, the resulting karushiri is at least possible according to Japanese phonology. Flirtatious one/market/position "it is" (informal copula). (Nothing starting karu appears in Koizumi's Zukai: Tan'i no rekishi jiten.) | ||
40,000 | Komma- ichi-da |
Koma? ichi da | (gibberish) Sounds like "It's comma-one", i.e. 0.1, since a continental decimal comma is sometimes used. (Nothing starting koma or kon and resembling this appears in Koizumi's Zukai: Tan'i no rekishi jiten.) |
(Coming soon)
It turns out that this section of Cardarelli is copied more or less verbatim from the earlier book by Washburn. (references to follow)
This even older book by Clarke has a quite different confusion of Japanese units. I cannot make any sense of the lengths "inc" and "sals", neither being remotely possible phonology, but the volumes are interesting. The unit is the gō (合): ichigō (一合) is just "one gō", so the pleonastic "5 itchigau" means 5 gō, which in (slightly old-fashioned) Japanese is gongō, again being "1 gonghau" (perhaps using the 'h' to avoid a "soft g" prounciation). Similarly, again, "2 ghonghau" makes 10 gō, which is called shō (一升), and "ischiomassé" is probably isshōmasu, or a 1-shō measure. Unfortunately, though, 1 shō is about 1.68 litres, not 1 gō. The larger unit is unclear, except that "itto" might be confusion with tō meaning "ten". Well, this is significantly wrong, but working a century before Cardarelli, with probably no relevant reference works at all, we can begin to forgive Clarke.
For comparison, here is the section on Weight units from Basil Hall Chamberlain's Things Japanese: Being notes on various subjects connected with Japan for the use of travellers and others, first published in 1890; this is from the fifth edition ("revised") of 1905.
(Deleted: AfD discussion)
This page as created contained the following sentence, and conversion factors referenced to Cardarelli:
Salmarazd (French name: salmanazar) was a UK bottle size for wine and champagne.
There certainly is a large wine bottle size named after an Assyrian king named Shalmaneser, but as shown in Wine bottle sizes, it is normally known as "Salmanazar". So we might first ask: what is "Salmarazd"?
The DuckDuckGo search engine produces 24 pages including the word Salmarazd on the web. These break down as:
This casts serious doubt on the very existence of the name. It is particularly interesting not to find any trace of this as an alternative rendering of Salmanazar.
Hoary also pointed out in the AfD that one would expect to see instances of "drank a salmarazd", "poured a salmarazd", "shared a salmarazd", "celebrated with a salmarazd", or similar; but that a search in Google for "a salmarazd" (thus after any verb or preposition) brought no hits.
Thanks to User:Hoary, we have another list incuding the same form: another dictionary from Springer. But this is a later edition: the second edition (1964) which I have does not include this section; Appendix 3 ends with "Troy Weights".
This is a table of bottle sizes, headed Table 3-33 UK units of capacity for wine, which can be seen on page 47 of this. Overall this list is similar to the list in Wine bottle sizes; these are named after ancient middle-Eastern kings, and inevitably such names have acquired variously-adapted/mangled versions in different European languages. But the heading to the column of bottle names is the rather odd "French name (English word)" – odd, because in no sense is anything an English word. It is difficult to avoid the suspicion that something written originally in French ("Nom - anglais" perhaps) was converted to this form by an editor or translator who was not properly familiar with the content. In any event, the problem is that the parenthesised "English" version is not reliably the version normally used in English. For example, Nebuchadnezzar ("Nabuchodonosor" in French) is rendered as "Nabuchadnezzar". The totally anomalous "Salmarazd" is thus not an isolated case.
In creating this entry, the editor saw fit to pick out just one isolated term. It is not entirely clear whether the original author (Cardarelli) even intended the English form to be seen as the name of the bottle, rather than as a gloss, which (if correct!) would remind us of the usual English form of the historical character even when (as in this case) the bottle is known by the French form "Salmanazar".
A combination of unreliable source information, inadequate editing of the publication (Cardarelli), and Wikipedia editing without at least a native speaker grasp of the subject have resulted in an entry which is almost certainly complete nonsense.
This is an analysis of the basis for the WP article Deal (unit). Cardarelli page 52 is cited as reference for this article, but this page in its entirety consists of the following:
(As far as possible this reproduces the wonky typography: although the convention exists in some cases of omitting the full-stop after abbreviations including the last letter, the standard traditional notation would be "12 ft. 6 in.")
These four footnotes constitute the only appearances of the word "deal" in this Section (3.1). I assume that the "(E)" means "exact", though no definitive confirmation of this has yet been found.
Comments:
In conclusion: these are footnotes, of unspecified relevance. There is no evidence that any "deal" is or ever was used as a unit for measuring firewood. Note that the single line for "rick" is equally doubtful, but appears to be the entire basis for the WP article rick (unit).
In the table of "UK Units for Beer, Wines, and Spirits" is the entry "Stuck (hock) ... 1182-1206 litres". In more than four years since the article was created, no support for this supposed English unit has emerged, hence the AfD (discussion). It appears to result from omitting the umlaut in the German word Stück; in the AfD discussion several editors found instances of reference in books about wine, almost invariably italicised and parenthesised, implying indeed that it is a foreign term. It looks, therefore, as though it could be a simple error, since the word Stück is usually simply a counter, meaning "per item" or "per piece". But in fact there is a German wine cask size known as Stückfass (WP:de), commonly abbreviated to Stück. So this is just inaccurate, and as usual overprecise: Cardarelli gives 1182-1205 litres (actually cubic decimetres, more spurious precision!) while WP:de gives 1000-1200 litres.
units.wikia.com/wiki/Special:Search?search=cardarelli&fulltext=Search&page=2&ns0=1&ns14=1 [3]
These are often listed as though they were all part of a standard system, when in fact they were of specialised use and/or variable definition, and in some cases no support can be found even for their existence. A common problem is the possible misinterpretation of qualifying words or phrases — it should be remembered that Cardarelli appears to be a native speaker of French or Italian, and this book is cited as having a translator. It is likely therefore that some of this material has been copied from unreliable sources (of unspecified antiquity) in French or other romance languages. For example, Cardarelli lists "Land" as another unit equalling (statute) "Mile", yet this is surely a misunderstanding of "Land mile" as opposed to "Nautical mile". In many cases standard sources such as the SOED refer to "varying" values in different times and places, where Cardarelli gives a false impression of precision by mentioning only one.
Cardarelli Section 3.3.1.1.1 lists the following as "Other units of length which have been used in the UK".
Cardarelli | Comments | |
---|---|---|
lea | 360 ft | SOED: "A measure of yarn in varying quantity..."; worsted 80 yds, cotton and silk 120 yds |
skein | 360 feet | Specialised usage in the wool industry (?) |
wrap | 240 ft | Specialised usage in the wool industry (?) |
bolt | 120 feet | Specialised usage in the wool industry (?) |
shackle | 90 feet | |
rope | 20 feet | |
ell | 45 inches | This is the English ell; the Scottish one was 37.2 in. |
pace | 30 inches | |
span | 9 inches | |
nail | 9/4 inches | |
finger | 7/8 inch | |
barleycorn | 1/3 inch | |
button | 1/12 inch | |
iron | 1/48 inch | This is a unit used for measuring (only) the thickness of leather used to make the uppers of boots and shoes. This conversion chart [4] from someone selling leather in the US includes a conversion to ounces [sic]. |
calibre | 1/100 inch | |
point | 1/144 inch | |
mil | 1/1000 inch | |
mil | 1 thou |
Still at the "notes" stage...
The heading and first two rows of Table 2.13 Non-SI discontinued metric prefixes are as follows:
Prefix | Symbol | Value |
---|---|---|
hebdo | H | 107 |
lacta | L | 105 |
(Notice the implausibly offered "symbols" (itself an SI term) H and L, for units not in SI.)
The term hebdomètre was certainly proposed for the pole-to-equator distance, and this is surely worth a mention. But there is no evidence for any independent prefix "hebdo-", and no obvious basis for the "symbol" H.
Several people independently suggested that this might somehow be a confusion with lakh, the numeral for 100,000 used in Indian English.
In a private communication with another editor, the author François Cardarelli has said (hearsay) that this "lacta-" was an error, and should be "lakh", citing a confusion because he was using Italian sources. There is a clear reference in question: Lange's Handbook of Chemistry page 2-16, where under a table evidently of prefixes for powers of ten there is a note which reads (in full): The prefix "myria" is sometimes used for 104 and "lakh" for 105. This footnote has disappeared from later editions, so must have been written sometime between the first edition of 1934 and the 1967 edition in the Google books link; it is surely clear that the author is trying to be helpful, in an era before the internet, by mentioning a couple of obscure multipliers. Well, "myria-" may be a prefix, but "lakh" is not, it's a numeral. In Indian English it is widely used for counting things or money: in other words, you say "1 lakh people" or "30 lakh rupees". It is not normally a unit multiplier, so can you say "1 lakh miles"? Perhaps, but it's not going to be "1 lakhmile" (singular!). But we know all about the lakh numeral, and its use in Indian English, so the entry from Lange's book is not now a useful source. (See article lakh)
Meanwhile, what of lacta-? The only evidence we now have of its existence is that the author Cardarelli says that using the Italian translation[6] of Lange's book, he wrote it by mistake for lakh. No-one has yet managed to produce the Italian text of the note under the Table, nor any other evidence for its existence. I have found no trace of it in monolingual Italian dictionaries on line, the Italian wikipedia entry on the Indian numbering system does not mention it; bilingual dictionaries do not mention it Dizionari Corriere gives Lakh (Ind. Eng.) = centomila, the normal Italian word for 100,000. I tried searching for the putative prefix attached to basic units, in both singular and plural; for a reality check I compared with the number of hits with the prefix etto-, which is the Italian form of "hecto", with the following results:
ettogrammo 2000 | lactagrammo 0 | gram (singular) |
ettogrammi 4000 | lactagrammi 0 | gram (plural) |
ettometro 2000 | lactametro * | metre (singular) |
ettometri 3000 | lactametri * | metre (plural) |
ettolitro 34000 | lactalitro 0 | litre (singular) |
ettolitri 186000 | lactalitri 0 | litre (plural) |
The numbers for hecto- hits are very approximate, but the result for lacta- is zero, leaving no support at all for the existence of this prefix. And it could be noted that even the clearly existing prefix etto- does not appear in en:WP, while of course it has an entry in it:WP etto, and at Wiktionary: etto.
The original article created by User:Shevonsilva (listed at User:Johnuniq/sandbox3#Lacta) was a one-sentence statement, "Lacta is a Non-SI, obsolete metric prefix (Symbol: L) for 105." with a reference to Cardarelli.
The lacta- prefix claim was then added to Metric prefix#Obsolete metric prefixes by User:Matthiaspaul; discussion became dispersed between Talk:Metric_prefix#Lacta? and Wikipedia:Redirects for discussion/Log/2016 August 18#Lacta-.
I was taught (grammar school, England, 1960s) to use the abbreviation "gm." for gramme, and recall our physics teacher poking fun at people who wrote "grm.", while pointing out the undesirability of using "gr." because of confusion with "grain". But the point is that these were invariably called "abbreviations", not "symbols". When the new-fangled "g" appeared (around the 1970s??) it was notable for two things: not using a full stop, just being the bare letter, and being termed a "symbol" rather than an abbreviation. This distinction is not observed anywhere that I have found in Cardarelli, so that old English units are given "symbols" (or codes? the book is not explicit about this), more or less abbreviations without full stops. Many are very questionable, often including an incongruous "UK" prefix: e.g., Wey has the symbol "UK wy".
The book is not written in native-speaker-informed English. The following quote, for example, in Section 3.3.1.2 Imperial Units of Area:
It is important to note that in North America and Britain, the prefix sq. (an abbreviation of the word square) is sometimes used before an area unit instead of raising the unit to the power of two.
I do not know where or when the custom arose of writing a power after the unit, but it was certainly unknown in England before SI units. So the "sometimes" in the above quote is odd: this was the only way to write units of area: sq. in., sq. m., etc. In particular, this helped to avoid confusion with expressions like "4 inches square", which means a square of side 4 inches, and thus an area of 16 sq. in.
It seems that the author does not realise that in English the unit "square foot", for example, naturally abbreviates to "sq. ft."; remember that the original book is said to have been translated, from some unnamed language. If this is French mètre carré or Italian metro quadrato, it would make sense to explain the abbreviation, so this might might just be thoughtless translation.
Cites Kotaro Honda (ja:本多光太郎) as correspondent in Japan
Bibliography: 日本化学誌 Proceedings of the Physico-Mathematical Society of Japan
Sample chapters from Cardarelli's website:
There's a table of contents somewhere, and I can't find it... (The Google Books view[7] currently - 8 January 2015 - shows it.)
In the back of his earlier book, Cardarelli presents impressive lists of sources: more than three hundred in all. The great majority of these have titles suggesting concentration on SI, very limited areas of application, or dry presentation of conversion factors, or seem unlikely to appear other than in the largest of libraries, or are otherwise hardly worth bothering with here. But among the titles that more obviously suggest usability for historical and encyclopedic purposes are:
Other (older) books
Online resources
C: Cardarelli 2003, W: Washburn 1926, -c: Clarke
This list is intended to help to determine whether a "units" website (there are many of them) was copying from Cardarelli. These terms are ones for which no genuine basis has been found.
Doursther, Horace: "Dictionnaire universel des poids et mesures anciens et modernes", Bruxelles 1840 Google books