May 2011

Typically, when the origin of a word is disputed, as in the case of awning/auvent discussed in Origin Unknown Part 1 previously posted to this blog, the contenders for the parent form are all in some sense “real,” i.e., are words that have actually been attested or whose prior existence can be inferred. There are some exceptional cases, however. Consider the following from

Main entry: kemosabe
Part of Speech: n
Definition: faithful friend
Etymology: has various meanings in Native American language
Usage: used in this way in “The Lone Ranger” television show[]

Main entry: kemo sabe
Part of Speech: n
Definition: invented word meaning ‘faithful friend’
Etymology: 1933; fr The Lone Ranger[]

“The Lone Ranger” debuted in 1933 on the Detroit radio station WXYZ. According to Fran Striker, Jr., the son of the first writer for the popular series, “The Lone Ranger had nobody to talk to if he was a lone ranger, so it was suggested they create a sidekick for TLR. Script 11 introduced Tonto. And [he] was developed solely for the purpose of giving the Lone Ranger someone to talk to.”
[] Tonto customarily addressed TLR as “Kemosabe,” while the latter customarily addressed the faithful companion as “Tonto,” though Striker notes elsewhere  ( that “… in many of the early radio broadcasts, the Ranger calls Tonto Kemo Sabe AND Tonto also calls the Ranger Kemo Sabe.” Tonto does not appear to refer to TLR as “Tonto,” at least within earshot.

Fran Striker, Jr. reports that the program’s director, Jim Jewell, claimed to have borrowed the appellation “Kemosabe” from the name of a summer camp in Michigan, Kamp Kee-Mo Sah-Bee, which was established by Jewell’s father-in-law, Charles Yaeger in 1911 and ran until 1940. This much of the story is undisputed. What the name of the camp actually meant (if anything), and in what language (if any) is, however, rather murky. According to Jewell, the name meant “trusty scout.”

A number of possibilities have been proposed and have been nicely summarized by linguist John Koontz (, the most likely (if problematic) being from a Potawatomi form meaning something like “‘spy’—someone looking at a secret thing, ‘scout’—someone looking secretly at something, or even a ‘masked man’—someone with a secret identity looking out through the eyeholes of a mask.” A stretch, though Potawatomi would have been spoken in the area in which  Kamp Kee-Mo Sah-Bee was situated, something that cannot be said of Yavapai, a language spoken by an indigenous people of the southwest (sometimes referred to as the Tonto Apache) and in which, as Koontz relates, knymsave means ‘one who is white.’ 

What do you mean 'we,' kemosabe?

Tonto Speaks

And what of Tonto? Cecil Adams ( says “According to Jim Jewell, there was an Indian storyteller at Kamp Kee-Mo Sah-Bee who would get rowdy when drunk, leading the other Indians to call him ‘tonto.’” John Koontz elaborates: “It’s not known where [camp founder] Yeager himself got his word Kee-Mo-Sah-Bee.  However, he is reported to have employed a number of Indians at the camp, including at least one Potawatomi man, remembered as Chief Thundercloud…The Potawatomi man … was known as Tonto according to Jewell, who remembered that this meant ‘the Wild One’…Another possibility might be that there is a French form which was in circulation among the Potawatomi that resembles Spanish tonto ‘crazy’ … and which was a sobriquet of Mr. Thundercloud’s.” We may venture to guess that the French word in question is tonton (a common hypocorism for ‘uncle’).

De quel 'nous' parles-tu, tonton?

TLR Parle

The tonton proposal would certainly make at least as much sense as the derivation from Spanish tonto, which is more commonly used in the sense of ‘stupid’ than ‘crazy:’ Like bobo, tonto probably originally referred to someone with deficient language—and therefore mental—skills. (Bobo ‘stupid’ comes from Latin balbus ‘stammering’ while tonto is probably from Latin attonitus ‘dumbfounded’ literally, ‘thunderstruck,’ though Gómez de Silva in his Breve diccionario etimológico de la lengua española says that tonto is “a childish word [una palabra infantil] from common Romance.”) Why the Lone Ranger would address his transplanted Potawatomi (and possibly French-speaking) associate in Spanish is a conundrum, and it is a fundamental tenet of lexicography that just because two forms resemble each other doesn’t necessarily mean that they are in fact related (especially when there is no plausible semantic connection between them). Other proposals, such as that tonto represents the best the masked man could do to get his tongue around something like a truncation of Western Apache binii?e’dnende’ which Koontz glosses as “‘people without minds’, i.e., ‘crazy people’ or ‘wild people’”), however gratifying it might be to infer that the Lone Ranger’s aptitude for second language acquisition was no better than his faithful companion’s, take us even farther afield.

So, where does that leave us? The founder of Kamp Kee-Mo Sah-Bee and the original writers for “The Lone Ranger” are no longer with us, so we can’t ask them where the words kemosabe and tonto came from. Indeed, the opportunity to zero in on the creation of a word is often brief and fleeting. But not always, as we can see from the OED’s entry for the word limerence:

Forms:  19– limerance, 19– limerence.
Etymology:  < limer-, apparently an arbitrary element (compare quot.1977) + -ence suffix. Compare limerent adj.

In form limerance after –ance suffix.

The state of being romantically infatuated or obsessed with another person, typically experienced involuntarily and characterized by a strong desire for reciprocation of one’s feelings but not primarily for a sexual relationship. It has been suggested that this state results from fluctuations in the levels of various neurotransmitters, such as serotonin and dopamine.

1977    D. Tennov in Observer 11 Sept. 3/9,   I first used the term ‘amorance’ then changed it back to ‘limerence’.‥ It has no roots whatsoever. It looks nice. It works well in French. Take it from me it has no etymology whatsoever.”


As little kids, we first ask “why?”and later as we work our way through the other wh– interrogatives—who, what, where, when, and so on—we eventually come to “whence?” though most of us will typically ask where something or someone comes from rather than trot out the 25-cent word lest we be thought to be putting on airs. Whatever.

The quest for origins has given rise to two popular –ologies: genealogy (Where did Billy Smith come from, i.e., what is Billy’s family history?) and etymology (Where did  the words Billy and Smith come from, i.e., what is their linguistic history?). As with genealogy, so etymology: Just as we may know when Billy Smith was born and can sometimes trace his lineage, at least up to a point, making guesses about his distant forebears on the basis of his relatives’ DNA if we’re lucky enough to have samples, so we can often determine with some degree of precision when and how a given word entered the language, what words are related to it, and, using a variety of methods, take a stab at reconstructing its long-lost ancestors. Two other genealogical possibilities—that Billy is a foundling, and there are multiple candidates for Billy’s paternity—have their lexicographical analogs as well: We have no clue as to a word’s provenance, and there are competing hypotheses as to a word’s origin.

Of the 156449 “English” words glossed in the OED online as of this writing, 4024 are labeled as “origin uncertain.” Early in the list is awning: “A word of obscure origin, apparently at first only in nautical use. Probably to be referred (as by Wedgwood) to French auvent ‘a penthouse of cloth, etc. before a shop window, etc.’ Cotgrave, early plurals in Littré auvens, auvans, medieval Latin auvanna, auvannus, whence *auvan, *auwn, awn; the termination is of course English –ing suffix1.E. Müller refers it to Low German havenung, < haven harbour, in sense of ‘a shelter from wind and weather’; Skeat compares ‘Persian áwan, áwang, anything suspended, awangān hanging, awnanga clothes-line’; but neither of these is applied in its own language to an awning; in particular an oriental origin seems incompatible with the history. French auvent is itself of doubtful etymol. See Diez, Littré, Du Cange.”   

Dictionary etymologies are replete with qualifications and polite weakeners. In addition to obscure, popular disclaimers include uncertain, apparently, probably, perhaps, and doubtful–one of several polite ways of saying that a proposed etymology is at best controversial and at worst completely off the wall–as in “French auvent is itself of doubtful etymol.” Another form of disclaimer is signaled by the asterisk (*), signifying that a given form is hypothetical—it should have existed but nobody has ever seen it. Another way of kicking the can down the road is to send the reader to other sources (such as Diez, Littré, and Du Cange), assuming you can track them down.

So, let’s have a look French auvent to see what its DNA might tell us.  (Spoiler alert: As the OED warns us, the derivation of awning from auvent is an ignotum per ignotius, that

is, an explanation of that which is unknown by that which is even more unknown.) Here are the main contenders:

  • The author of says “the word auvent, designating a projecting part of a roof, comes from Vulgar Latin antevannum, surely derived from Gallic *andebanno. This term would then be a compound *ande-banno- signifying ‘front horn’ [corne de devant]. This roof projection being perceived by the Gauls as a horn.” We may for the moment overlook the fact that antevannum is given as an attested rather than a hypothetical form—and without a gloss at that. As Antoine Thomas puts it in his Mélanges d’étymologie française, “Auvent seems to be for anvan, a form preserved in Old Provençal, from *antevannum, a compound that should (must) have existed [a dû exister] in Vulgar Latin.”
  • Le Petit Robert derives auvent from Gallic *ande-banno, from ande- ‘forward’ and banno- ‘gable,’ sidestepping the question of an intermediate (*)antevannum.
  • Larousse étymologique gives as auvent’s progenitor *antevannum, “an obscure formation, perhaps from *banno, ‘horn’, in Gallic a protecting totem [totem protecteur].”
  • Bloch and von Wartburg, in their Dictionnaire étymologique de la langue française, say that auvent comes from Gallic *andebanno- “a compound consisting of the augmentative prefix ande and banno- ‘horn,’” adding by way of elaboration, “bull’s horns were used by the Gauls as a totem to protect their houses.” This is essentially the etymology favored by the Centre National de Ressources Textuelles et Lexicales (see, though, as with the other *andebanno derivations, a certain amount of hand-waving is involved.
  • Pierre Guiraud (Dictionnaire des étymologies obscures), seizing upon these and other flies in the lexicographical ointment,  presents several arguments against the *andebanno hypothesis, to “ruin the already inherently very fragile semantic conjecture concerning mythical totemic horns.” Instead, he proposes deriving auvent from “a form of au d(e)vant” (‘in front’), itself derived from a possible (if unattested) Latinate construction meaning something like “in front of the in front,” and not from Old French anvant, which he derives from a related but different construction, once again reminding us of  the Monty Python pet shop sketch in which the shopkeeper proposes to derive a cat fom a dog:

Shopkeeper: I’ve got a lovely terrier.
Man: I want a cat.
Shopkeeper: Listen, tell you what. I’ll file its legs down a bit, take its snout out, stick a few wires through its cheeks. There you are, a lovely pussy cat.
Man: It’s not a proper cat.
Shopkeeper: What do you mean?
Man: Well, it wouldn’t meow.
Shopkeeper: Well, it would howl a bit.

Two final meows: (1) Du Cange (whose 19th-century Glossarium mediae et infimae latinitatis the OED suggests you see) has an entry for the word auvanna ‘a wooden sheltering projection [umbraculum ligneum projectum] appended to a window or workshop,’ citing Gallic auvent, as he does for antevanna (“umbraculum ex assamentis”). (2) Antevannus appears in its accusative singular form (antevannum) in Paolo Canciani’s late 18th-century Barbarorum leges antiquae in a passage describing where ship owners are and are not allowed to create rooms (camerelas) on a vessel. It is not clear what, exactly, an antevannus was but, for what it’s worth, a vannus, according to Auguste Val’s 1848 Glossaire Nautique was ‘a room at the poop,’ from Latin vānus ‘empty.’

The front page of the Mexico City newspaper Reforma recently featured an article announcing the publication of a handbook promoting the use of nonsexist language under the eye-catching headline “Lanzan manual contra machismo” (roughly, ‘They’re Launching a Manual Against Male Chauvinism’). The “they” in question are the Office of the Secretary of Governance (Segob) and the National Commission for the Prevention and Eradication of Violence against Women, the manual is entitled Manual para el Uso No Sexista en el Lenguaje [‘Manual for Nonsexist Language Usage’], and its initial distribution will be to “all functionaries in the Federal Government.” The article begins:

 “In order to eradicate verbal violence against women, the Federal Government yesterday introduced a manual to counter male chauvinism (machismo) that proposes, among other things, the avoidance of the usage of such words and phrases as “vieja” [‘old (female),’ or, as undoubtedly intended in this case, ‘old lady’] and “si quería trabajar, ¿para qué tuvo hijos?” [‘if you wanted to work, why did you have children?’].”

Though the writer does not elaborate on these two examples, they encapsulate a number of the most daunting of the sociolinguistic obstacles that speakers of Spanish will inevitably encounter on the way to gender equality—never mind the way to gender blindness/agnosticism, which isn’t even on the horizon at this point in the game: We are talking about progressing from something like lady policeman to policewoman, not from policewoman to police officer. (Not that as sensitive new-age speakers of English we should be too smug: Hands up if it immediately occurred to you that “If you wanted to work, why did you have children?” might have been addressed to a dad?) The basic linguistic problems stem from the association of a grammatical feature called gender with a biological feature called sex (gender in reference to humans nowadays being used by many speakers of English to designate a culturally rather than strictly anatomically defined attribute not applicable to the other members of the animal world).

To some extent, speakers of Spanish can blame history and the Whorf-Sapir hypothesis (that your language shapes the way you think) for their sociolinguistic troubles: Like the other Romance languages, Spanish evolved from Latin in which all nōmina (nouns and adjectives) belonged to one or another of five classes called declensions, distinguished from each other by their endings.  As a rule, nōmina from the first Latin declension that survive in Spanish end in –a; those that come from the second declension end in –o; those that come from the third declension typically end in a consonant or –e; the tiny handful of fourth-declension relics end in –o; and the one holdout from the fifth declension (día ‘day’) ends in –a.  So far so good. However, to complicate matters, every Latin nōmen had associated with it a grammatical gender—masculine, feminine, or neuter (i.e., ‘neither’), though in theory they could just as well have been named van, choc, and straw, were it not for the fact that those nōmina of the first declension that refer to animate creatures generally (though not always) refer to anatomical females while  those nōmina of the second declension that refer to animate creatures almost always refer to anatomical males. 

The upshot in Spanish, to grossly oversimplify the story, was to generalize from the above that, except for inheritances from the Latin third, fourth, and fifth declensions, nouns and adjectives referring to females end in –a and those referring to males end in –o, so if you want to coin a term to refer to a creature of one sex or the other, you can basically take a root and stick –a or –o on the end of it, and you’re home free. Indeed, this is a strategy suggested by the manual, following mutatis mutandis a template provided by the Romans themselves. By implication, this strategy can be employed to avoid another historical quirk, the use of the masculine form when referring to a mixed group of males and females such as the hijos (‘children’) in the example quoted above. True, “¿para qué tuvo hijos o hijas?”  (“why did you have sons or daughters?”) is a mouthful and a gender-neutral single word substitute (like English children or kids) does not readily come to mind, but there are other subtler difficulties, one raised by the previously cited word vieja and another (not mentioned in the article) exemplified by the polite words for the police.

Vieja and viejo are the feminine and masculine forms, respectively, of a word whose basic, canonical meaning is ‘old.’ They are the direct descendents of Latin vetula/vetulus ‘elderly,’ the diminutive form of veta/vetus ‘old, aged.’ When used substantively (i.e., when the adjectival form is used as a noun), vieja can have a pejorative sense (‘old lady’), a wrinkle shared by a number of feminine/masculine pairs in Spanish in which the masculine form is neutral or positive in connotation while the corresponding feminine form is tainted—the manual cites zorro/zorra (literally, ‘male fox/female fox’ but figuratively something like  ‘clever guy/foxy lady’) and hombre público/mujer pública (literally, ‘public man/public woman’ but in general usage  ‘prominent man/ prostitute’). The manual suggests substituting non-loaded terms for such pairs as zorro/zorra and if you can’t say anything nice, don’t say anything at all.

But sometimes the challenge would seem insurmountable. ‘The police’ is la policía, grammatically feminine.  ‘A policeman’ is un policía, grammatically masculine (despite the –a ending). And ‘a policewoman?’ Not, as one might expect, una policía, but either un policía, or if further precision is called for una mujer policía (literally, ‘a woman policeman’) or una agente feminina de policía (‘a female agent of the police’). Well, as the Reforma article concludes, “Segob’s subsecretary of Human Rights, Felipe Zamora, said that it will take work to reduce the gender gap in diverse areas.”