Thursday, May 23, 2013

Let's map Urheimats!

Above is a map of something I've never seen before: a map of all the ancient language families of Europe and Asia Minor. What does that mean for the layman and woman? Simply put, it is the ancestral homeland of every language family as far back as we can trace it. So, for instance, on the map we see Hurro-Urartian. In antiquity, there were two kingdoms - the Hurrians and the Urartians - that spoke two related languages. By comparing the two languages we can reconstruct a "proto" language. But sometimes the daughter languages migrate and move away from where their ancestral proto language was spoken. In the case of Hurrian and Urartian, they both moved away from their homeland. This map shows their original linguistic land, back in the Armenian mountains, and not where they ended up, which was further south and east. We can do this with a lot of languages of the world. Unfortunately most language maps show where living languages are currently spoken or where extinct languages were last spoken, kind of like a burial marker. But these languages came from somewhere, right? Well I put together a map of all of those homelands. How do we know how to do this?

Sometimes a group of languages are demonstrably related, and linguists can compare the languages and reconstruct the original tongue the group diverged from (like Proto-Indo-European, Proto-Uralic, Proto-Hurro-Urartian, Proto-Northwest-Caucasian, Proto-Northeast Caucasian, Proto-Kartvelian and Proto-Tyrrhenian). Other times, a language is isolated - that means it is unrelated to any others - but linguists are able to compare the internal differences among dialects to reach a proto-language as well (like Proto-Basque). Other times, we are simply stuck with a single language that we cannot recover earlier versions of (Iberian). Finally, sometimes we can look into living languages and discover that the language contains traces of ancient dead languages that whisper to us like echoes of a time long past (like in the case of Pre-Germanic, which was spoken by a people before the Germans arrived, as well as Pre-Irish and Helladic).

For the regular joe, what do the language families represent?
  • Proto-Indo-European: Their nation(s) was called the Kurgan civilization (5000 BCE or so). They broke into a lot of tribes. If you speak a language in the following sub-families then your language comes from Proto-Indo-European: Germanic (including English, Italic (including Romance languages and Romanian), Indic (like Hindi, Urdu), Iranian (like Pashto, Farsi, Persian), Anatolian (Hittite, Luwian), Greek, Albanian, Armenian, Tocharian, Balto-Slavic (Russian, Lithuanian, Latvian, Slovak...), and Celtic.
  • Proto-Basque: Ancestral language of Basque, the last living language of the paleo-European tongues. When the Indo-European and Uralic tribes invaded, all of the aboriginal languages disappeared excepting Basque.
  • Proto-Uralic: Ancestral language of the Finno-Ugrics (Finnish, Hungarian...) and the Samoyed.
  • Iberian: Dead language. Language isolate in Spain. It neighbored Basque and the two were closely linked by trade and culture, but their languages were unrelated. 
  • Pre-British: Dead language. When the Celts came to the British Isles, they came to a land already inhabited. Some of this "Pre-British" language survives in the Celtic languages.
  • Pre-Irish: Dead language. When the Celts continued on to Ireland, the settlers there had their own tongue as well. It may have been related to Pre-British, but the traces of their language in Old Irish are distinct enough that we can't say it was the same language.
  • Pre-Germanic: The Germans split from the Kurgan civilization and settled the Scandanavian terrain, intermarrying with the aborigines who taught them various maritime techniques and agricultural terms. This term isn't exactly fair, because it apparently existed in areas that were conquered by the Proto-Balto-Slavs and the Proto-Celts, so that we see cognates in their languages as well. In fact, we may be seeing some loans appear in Finno-Ugric as well (but that's debatable). Pre-Germanic simply spanned a greater geographic area than the actual German tribes that settled the majority of their land.
  • Helladic: Dead language. More than 1500 words exist in Greek from a predecessor language that was not necessarily Tyrrhenian (the other language group that interacted heavily with Greek).
  • Proto-Tyrrhenian: Extinct family. The homeland in westernmost Turkey disappeared when they were conquered by the Phrygian invasion in 1200 BCE. The survivors sailed west to Greece and Italy, becoming the Etruscans and Raetics.
  • Minoan: A language written in the Linear A alphabet. It died along with the civilization, replaced by the Mycenaeans, an early Greek tribe. Some aspects of Minoan hint at a relationship with Tyrrhenian and Anatolian but it's generally thought of as an isolate.
  • Sumerian: Dead language. Language of Sumer. One of the first written languages.
  • Proto-Northwest-Caucasian: About 1,700,000 total speakers of all its offspring languages, but none of the languages are well-known.
  • Proto-Northeast-Caucasian: Famous for being the family of Chechen, among others. Some linguists believe it to be related to Proto-Northwest-Caucasian, with whom both share an extremely large phonological inventory.
  • Proto-Kartvelian: Language family of Georgian.
  • Proto-Hurro-Urartian: A language family that died out a looong time ago.
  • Proto-Afro-Asiatic: Its homeland is barely seen here (the bottom-most part of Egypt). Ancestral language of Egyptian, Akkadian, Hebrew, Phoenician, Chadic, among many, many, many others.
  • Paleo-Sardinian: Dead language. A language that would have been unknown had not traces survived in the dialect of Italian spoken on northern half of the island Sardinia. It was also spoken in Corsica.
  • Elamite: Dead language. An isolate, though possibly a Dravidian language which would place its homeland further south and outside of the map. It was the primary language of Iran from 2800 to 550 BCE.
  • Pictish: Dead language. A language once spoken in present-day Scotland. It's weakly attested, meaning that its classification is still uncertain. It may get dropped from the map.
  • Tartessian: Dead language. Another candidate to be cut. It may be Indo-European but displays enough mystifying features to keep linguists guessing.
  • Hattic: Dead language. The language of Anatolia before the Hittites came (an Indo-European tribe). Swallowed up by Hittite.
  • Kaskian: Dead language. A language along the Black Sea's southern coastal mountain range. Possibly related to Hattic and it suffered the same fate.
  • The Tree Language: Dead language. Tricky language to map! A language so called because it gave the incoming Kurgans the names of plants. In truth, I had to do some guesswork as no one has mapped it before (to my knowledge). It had to be close to Pre-Proto-Celtic, because a few words crop up in that language; it had to be directly in the line of the Proto-Italic migration, because a huge number appear in their language; and it needed to be central enough that words appear in Proto-Germanic. In truth, while the "tree language" is an acknowledged substrate in Indo-European languages, its geographic location is uncertain.
  • The Bird Language: Dead language. A pre-Indo-European language in the North Balkans that gave us words for avians and helped us name the exotic birds we were unfamiliar with.
  • Pre-Proto-Celtic: Dead language. Distinct from the other languages that lent words into Proto-Celtic, this language is responsible for a small but important share of the Proto-Celtic lexicon. The topics range from war to agriculture. Like Pre-Germanic, it has a high frequency of geminates (a "long" consonant sound).
  • North Picene: Dead language. A sparsely attested language in Italy.
  • South Picene: Dead language. Called Picene because it was so close to its northern neighbor, but there's little reason to believe the two were related apart from proximity. Again, sparsely attested.
Not all languages have been mapped yet. I decided to map Pictish and Tartessian... for now, but I have my reservations about including them. Their linguistic heads could be on the chopping block at any moment. I have no idea if the bird and tree languages will stay. Their impact upon living languages makes their existence near-certain, but their location in Europe is anything but certain.

Bear in mind: this map is not a snapshot in time. There can be thousands of years between two language families on this map. The Kurgan civilization, which spoke Proto-Indo-European, is about 1000 years older than Sumerian, 3500 years older than Proto-Tyrrhenian, 5000 years older than Proto-Basque and Pre-Irish. This is a snapshot of the ancient language homelands (called Urheimat).

This map is a work in progress. Things may be inaccurate. Many things need to be added. Some regional boundaries may be off. If you have knowledge to contribute, please feel free to chime in!

By the way, this is the same map of language families but today. As you can see, Indo-European and Afro-Asiatic pretty much dominate. Uralic's hanging in there. White areas are a new language family, Altaic (mostly Turkish in this map) that was introduced via the Mongol and Turkish invasions. The Caucasian languages are hanging on for dear life and Basque (its territory greatly reduced) now has to compete with Spanish and French.

Tuesday, May 21, 2013

of gods and the name

Sometimes in the Indo-European pantheons, gods would get special names. Many of these epithets are mysterious: they point out attributes of the deities that we otherwise would not have known of. Some of these epithets are funnier than others. Here are some of the stranger titles.

The deity: Hera
Known as: The wife of Zeus, goddess of women and marriage
Her title: βοῶπις
Meaning: Cow eyes. 

The deity: *Torh2nt- (Thor in Old Norse, Tarhunt- in Hittite, Torann in Old Iranian)

Known as: The thunderer.
His title: *perkwuno-
Meaning: Rocky.

The deity: Agni

Known as: God of fire
His title: Apām Napāt
Meaning: Descendant of the waters.

Why is the Hindu god Agni, a god of fire, described as the descendant of the waters in the ancient Sanskrit annals? Why is the Thunderer described as "the Rocky Thunderer?" Pretty ballin' title though. I won't even touch Hera - though judging by her epithet, I doubt many did.
Matasović, Ranko. A Reader in Comparative Indo-European Religion. University of Zagreb. 2010.
Greek reflex from: Robert Beekes. "βοῶπις" in: Etymological Dictionary of Greek. Indo-European Etymological Dictionaries Online. Edited by Alexander Lubotsky. Brill, 2013. Brill Online. May 22, 2013.

Sunday, May 19, 2013

animals and the warriors

There is an ancient story from Greece about two brothers, Aigeus "the goat" and Lykos "the wolf." Athens was ruled by Pandion II in those days, but the throne was usurped by the brothers. After exiling Pandion II, it is recorded by Herodotus that Aigeus drove Lykos out of Athens (and two other brothers as well) and claimed the seat for himself. The story is assumed to have been manufactured by Athenians out of political motives, but the probing work of Reyes Cebrian has shown that the story may also be a reflection of a pre-historical tradition of enmity between goats and wolves.

In fact, as Cebrian points out, the Greek tradition of animal warriors and animal heritage stretches far back in time - before the Greeks were ever "Greek" or even lived along the Mediterranean coast. It's a tradition that goes back thousands of years to around 5000 BCE, back when the Greeks ancestors were part of an ancestral tribe we call the Indo-Europeans. The Indo-Europeans were an advanced and powerful nation that lived in the Pontic-Caspian Steppes, just above the Caucuses. They were horse-riders and chariot-builders at a time when most tribes were still wandering on foot and eatin' berries and lame-ass shit.

The Indo-European culture split over time, as all nations do. The first to leave were the Anatolians, who founded the Hittite and Luwian empires in Turkey. Later to leave were the Italo-Faliscans (some of whom became the Italians), the Celts, the Germans, the Indo-Iranians (who further split into the Persian kingdom, the Aryans of India, and many other tribes), the Tocharians (who settled in Western China), the Albanians, the Armenians, and the Balto-Slavs (Russia, Lithuania, Bulgaria, etc...). The Indo-European path forked. Every tribe took its own way and began to innovate and evolve. Over time, each tribe lost parts of its ancestral language and culture, while retaining other parts, so that no single tribe paints a perfect picture of its past. 

But much can be recovered.

In the writings and speech of every nation resound the echoes of yesteryear, waiting to be discovered like ancient puzzle pieces and put back together. One of those lost items is the culture surrounding animals and the warrior spirit. The tribes venerated the animals and turned to them as a source of strength.

The Germans revered the wolf and the bear. In fact, the true name of the bear was taboo - never to be uttered. Instead the Germans created a euphemism for its name, which lives on in our languages. English bear, Old Norse björn, and German Bär all come from Proto-Germanic *beron "the brown one." The fearsome Norse berserker was literally "bear-shirter," a mad-man adorned in bearskin. Although the berserker gets all the attention, the more common crazy Norse fighter was the ulfhedinn "wolf-coat." The wolf appears in many of the early Germanic names: Beowulf, Wulfingas, etc...

The Greeks turned to the wolf and the goat. Lykos went on to found the Lykians "the wolf people." Zeus and Athena wear the aegis, the goatskin shield. While Artemis is a goddess over wild beasts, her brother Apollo is master over but one animal, the wolf. All this may point to an earlier tradition of shapeshifting. Berserkers and ulfhedinn probably believed they were becoming the spirit of the bear and the wolf, just as in early Greek tradition, warriors were often depicted in wolf and goat hide. The Saga of the Volsungs describes Sigmund and Sinfjotli putting on wolfskins and being transformed into super-charged wolves. These glimpses into early Germanic and Greek life are windows into their ancestral period, their Indo-European period.

I think it's an exciting possibility and Cebrian's paper is worth a read.
Cebrián, Reyes Bertolín. "Some Greek Evidence for Indo-European Youth Contingents of Shape Shifters". Journal of Indo-European Studies. Vol. 38, Is. 3/4. 2010: 343-357.

Tuesday, May 14, 2013

Peculiarities of the English language #1

Here are some odd facts of our language.

    bet, the house
    'alef, the ox head
  1. The word alphabet literally means "ox-house" in Phoenician. The first alphabet was created by the Phoenicians, who labeled each letter with a word that began with the letter's sound. It's the same concept behind the NATO phonetic alphabet ...Papa, Quebec, Romeo, Sierra... So the Phoenicians were like, four million years ahead of their time and probably made this up between flying space cars. Phoenicians labeled the first letter the "ox" because it looked like an ox head; the second letter was the "house" for the same reason. Of course, sounds change over time. The original sound of alpha was closer to 'alef (the ' indicates a glottal stop) and bet was more like... okay it was like bet too. If that doesn't get you excited then get out of my face.
    Click this image... for infinite pleasures.
  2. The word bookkeeper is the only unhyphenated English word with three pairs of letters.
  3. Words like dreamt, comfort and warmth have a hidden 'p' sound between 'm' and 't'/'f'.
  4. The full line of the old quote 'Ignorance is bliss' is: "Where ignorance is bliss, 'tis folly to be wise." Which of course is complete nonsense cause we just got some serious knowledge bombs here and we're all the more blissful for it. 
'Till the next blissfultude, folks.

Tuesday, May 7, 2013


Where did the word silver come from? The ordinary enthusiast may find the Online Etymology Dictionary's entry and be content. But the OED is beset with a singular problem: its sources are very old and so the entries do not always reflect the most recent opinions of scholarship. Of course, it's not as if the OED is wrong, but that it does not contain the full picture. Today's entry will do that.

So where does the OED leave us? Ah, yes. The Dictionary traces silver back to Proto-Germanic *silubra-. (The asterisk indicates that the word was never directly recorded, but that it was reconstructed by comparing the Germanic languages and their lexemes). The Dictionary also gives us a number of cognates, to which I will add even more via Boutkan and Siebinga (2013).

Germanic family
Old English: seolfur
Old Saxon: siluƀar
Old Norse: silf-r
Old High Germanic: sil(a)bar
Gothic: silubr
Proto-Frisian: *sēluvir
Proto-Germanic: *silubra-

Balto-Slavic family
Old Church Slavonic: sĭrebro
Russian: serebro
Lithuanian: sidãbras
Latvian: sidrabs
Old Prussian: siraplis
Polish: srebro
Proto-Balto-Slavic: ??

So that's a pretty fair list. From the get-go, we can see that whatever "proto" word we obtain will have a form akin to silub-r, silab-r, sileb-r or the ilk. The vowels are uncertain but we have a short list to choose from.

Now, where did the Germanic and Balto-Slavic tribes get this word? When most European tribes gave a name to the new metal, they picked a word from their native inventory. Not surprisingly, they usually called silver "the white (thing)," based on the Indo-European root *h2erǵ- "to shine," "white." That was true for the Celts (PCeltic *arganto-), the Greeks (Mycenaean a-ku-ro), the Armenians (PArmenian *arcant-), and the Italics (PItalic *argento-). 

Because both the Balto-Slavs and the Germans have a very similar word for silver, it is suggestive that they borrowed the word from a tribe that traded silver with them. But whence? Three telling hints come to us from the Iberian peninsula.

1. Basque zilar < Proto-Basque *zirar. Löpelmann (1968) believed the word to be a loan from Germanic, but the famous Vasconist, R. L. Trask, notes that the direction of German-to-Basque could not be firmly established (2008). Further, what if it was a third-party loan into German and Basque? Nevertheless, the word gives us a firm relationship to the Iberian peninsula.

2. Celtiberrian silapur. Celtiberrian was a Celtic language spoken in the center of Iberia. This word cannot be traced back to a Proto-Celtic form, rather it stands alone in the Celtic family, again indicating a loan from an unknown source (Matasović 2013).

3. Iberian sarif, salir. Could this be the source-tongue we were searching for? Maybe. Alone this example stands as a word that was not necessarily borrowed. Note that iltiŕta-śalir-ban appears on coins. Could śalir-ban be our quarry? It is not a stretch to see the morpheme -ban misinterpreted by listeners as salirba-, a single word. Because ban also meant the numeral 'one,' could the word have been used in Iberian in the same way Basque uses their numeral bat 'one?' The Basque can utilize bat as a definite article which appears in the same position of a sentence as Iberian -ban has appeared. Thus the phrase may mean "an iltiŕta silver coin." On the other hand, the vowels are metathesized from an i-a pattern, a pattern we see in all the other languages, to a-i. So perhaps we finally found the donor language (Polome 1989), but in my opinion the jury is still out. It's good to see, however, the complete tale available in one page online somewhere.

  • Boutkan, Dirk and Sjoerd Michiel Siebinga. "selover" in: Old Frisian Etymological Dictionary. Indo-European Etymological Dictionaries Online. Edited by Alexander Lubotsky. Brill, 2013. Brill Online. May 8, 2013.
  • Löpelmann, Martin. Etymologisches Wörterbuch der Baskischen Sprache. Berlin: De Gruyter. 1968.
  • Matasović, Ranko. "arganto-" in: Etymological Dictionary of Proto-Celtic. Indo-European Etymological Dictionaries Online. Edited by Alexander Lubotsky. Brill, 2013. Brill Online. May 8, 2013.
  • Polome, Edgar C. “Preparing an Etymological Dictionary of Proto-Germanic.” Journal of the Dictionary Society of North America. No. 11. 1989.
  • Trask, R. L. "zilar" in: The Etymological Dictionary of Basque. University of Sussex. Edited by Max Wheeler. 2008.

Sunday, May 5, 2013

What's the hardest language in the world?

Of the six thousand languages of the world, which is the most difficult for an English-speaker to learn?

It's a question I've long asked myself. Part of the reason I'm so fascinated by the question is that other linguists and enthusiasts do such a poor job of answering the question. Some hide behind question-retorts like, "Well, what do you mean by hardest?" or "What's the criteria for hardest?" While those questions are important, they also kill the fun by ending the conversation in a long-winded explanation that all languages are equal in content delivery. Yes, I think to myself, but that doesn't mean that the underlying structures of delivery are equally apprehensible by an English speaker.

So for this blog post and perhaps other posts to come, I will explore what I believe are the hardest languages to learn for a native English speaker. Previous attempts tend to focus on politically important languages: Japanese, Polish, Mandarin, Hungarian. Languages that are widely spoken and are globally important. Sure, Mandarin has a nightmarish writing system and Japanese has a painful content delivery system, but what about Northeast Caucasian languages that have an enormous inventory of phonemes and a fearsome system of vowel harmonies. Japanese and Mandarin are discussed because people bother to learn them. I intend to consider all languages.

This, of course, unfortunately means that many hundreds of qualified candidates will be overlooked because there is simply too little information or (more likely) I haven't read more than a sentence or so about them. My apologies in advance. If you know a good language to consider, please let me know in the comments. A handful of other very considerably difficult tongues must be skipped because our knowledge of the language is incomplete. From what we know of Hurrian and Sumerian, they are frustrating ancient languages, worthy of any list like mine. But our knowledge of the languages is grossly incomplete and therefore I condemn the tongues to the soil. Sorry, guys!

Today's candidate for most difficult language is an old favorite:


Basque is a language isolate spoken by ~300,000 people in northern Spain and the far western reaches of France. It is part of a category of Paleo-European languages that are unrelated. Paleo-European languages are languages that were spoken before the invasion of Indo-European and Uralic tongues. The Germanic languages like English and German; the Italic languages like Spanish and French; the Uralic languages like Finnish and Hungarian; Balto-Slavic languages like Russian, have all done a very good job at eliminating all the old tribal languages that were once spoken thousands of years ago. Among all of the ancients, only Basque survives. It is unrelated to any other language on earth, though we know a related dialect of Basque called Aquitaine existed during the time of Julius Caesar. So why is it so difficult?

1. Ergativity

Basque a completely ergative language, in fact it is the only totally ergative language in the world. All other ergative languages are actually "split-ergative," able to state things in either ergative or nominative cases, dependent on tense, structure, or class. What's an ergative sentence? It's where a subject "behaves" like an object. Wikipedia has a good example. While in English we can say "I moved her" or "she moved," ergative speakers would have to say the equivalent to "I moved her" or "her moved." But a better example might come from a short list of ergative verbs that exist in English. The verb to break is an ergative verb in English.
"The window breaks." 
 "The window breaks me."
 Note that the verb breaks is not behaving the same in both examples. In the first, the window receives the action because the verb is behaving ergatively. Compare the same structure but with a non-ergative verb.
"She loves."
"She loves me."
In this example, the woman is doing the 'loving' in both sentences. Unlike the window that breaks, the woman directs her love to an object, regardless of whether the object is named or not. By placing 'me' at the end of a verb, we can quickly hear that the verb is ergative. "The window breaks me" sounds very funny and it should be "I break the window" (but not "I break"). Most verbs in English are not ergative.

2. Those damn cases

Cases are markings at the beginnings, middles, or ends of nouns that signify meaning. English has a few cases, most notably the genitive case, indicated by -s. In English, all cases can be replaced with a preposition or a postposition. "Flannery's house," a sentence with the genitive -s, can just as easily be read "The house of Flannery" where the of contains the meaning of the case marker -s. Latin had 5, 6 or 7 cases, depending on which era of Latin you are studying and whether you are counting archaic and relict cases like the Locative Case during the Silver Age of Latin.

Now, are you ready to blow your mind? Of course you are. Basque has seventeen cases with four forms to each case. That's 68 different forms for every noun.

3. The unrelated lexicon

Because Basque is a language isolate, the vocabulary is unrelated to English. That means that unlike French or Norwegian, you don't get to rely on hunting for cognates. To be fair, Basque has borrowed scores of Spanish words which makes it a bit easier, but that doesn't mean that a huge number of words sound totally different.

4. The verb

The verb is probably the most difficult aspect of the language. Verbs can either be finite or non-finite, which means... well... God, I'll just let Wikipedia explain if you really want to ruin your day.

So those are some reasons why Basque is a high-level candidate for most difficult language for an English speaker to learn. There are some caveats: the sizeable number of loans from Spanish and the very simple phonology make Basque a good deal more palatable.

Anyway, if you agree/disagree or if you have a suggestion of your own, leave me a comment below.

Thursday, May 2, 2013

The Maxim of Sweat

"Horses sweat, Men perspire, and Women glow."

You may have heard this maxim before. It is an attempt to relegate particular verbs into categories of social politeness. The raw, gross nature of sweat is too crude, even for a man. It comes from an earthy Germanic root, much unlike the scientific and placid perspire, a verb of Latin origins. Perspiration, unlike sweat, is devoid of its grosser implications. It contains a medical nuance and does not evoke the mental images of wet skin and soaked shirts. But women (ever the delicate creature, eh?) deserve a verb of noble sensibility: glow. The virtue of women is to be respected to the point that the only appropriate description of sweat on a lady is a euphemism.

Apart from some amateurish detective work on the phrase, I have not seen anyone trace its history with any rigor and depth. It seems to date back to some point before 1883, wherein The Medical Age (pg. 268) we find the earliest written mention of the phrase in print,
The teacher of a female seminary in Philadelphia instructs her pupils, touching the function of the skin, as follows: "Horses sweat, men perspire, and young ladies get in a glow."
There's a lot we can learn here. First, based on the date of the quote (the 1880s) and the social constructions of the maxim, we can safely confirm the phrase was born out of Victorian ideals. Second, the age of the ladies was distinguished by the teacher (a male?), and the age of a lady seems to be latched onto the verb glow. For instance, in an 1884 publication of Puck (1884, Vol. 15, pg. 300), we see the age of a lady to be of importance,

A stock-yards urchin thoughtlessly ejaculated, "Oh, dear, how I sweat!" quite to the horror of a youthful and very precise aunt, who reproved him for making use of so inelegant and impolite a term.
"Oh, yes, I know all about it," he replied in an impatient manner and petulant tone: "horses sweat, men perspire, and giddy young things like you only glow."
The article claimed to be quoting the Chicago Sun - though I could not find their citation. It may be that the Chicago Sun's article pre-dates The Medical Age. The stock-yards urchin's response was sarcastic and petulant, implying aunt's correction was a silly and ostentatious bit of culture and erudition. It may reflect a division between normal society (represented by the urchin) and Victorian sensibility (the aunt). Joel Chandler Harris, writing a review of Puck in 1907 in American Wit and Humor, remarks that the magazine ran stories far too coarse and unrefined for typical society. Perhaps Puck ran stories that were a reflection of real American people, rather than the ideals they set themselves up to mirror.

There seems to be predecessors of the maxim. A 20 July 1881 publication of The daily Cairo bulletin. of Cairo, Ill., ran a story reported from the Catskill mountains, that ladies in the region hunted in full make-up and well-to-do clothing, and that (due to the heat) the ladies "thanked God that ladies could glow but not perspire." Well, that's odd. Is glowing not perspiration or sweat? The Oxford English Dictionary records that earlier uses of 'glow' indicate the burn of bodily heat from emotion, passion, or physical exertion, such as Walter Smith's poem Kildrostan in 1884, "Girls, all glowing with the flush of life." This, of course, is less interesting, because this particular use of glowing is quite ordinary and regular.

It seems, then, that the great maxim does not imply that ladies sweat at all. In fact, their skin merely flushes. Of course, this is a bit untrue, and it's difficult to imagine people truly believing it. Based on the stock-yards urchin's awareness that the maxim was already fixed into society's standards ("Oh, yes, I know all about it," he said, implying that the concept was well entrenched even in 1884), we can assume that his sarcastic reply meant that women obviously sweatted, and that the glowing was a lie "ladies" tell themselves.

We may also mention geographic location of the phrase. The instances center around the middle states of America: The Medical Age records Pennsylvania, Puck and The daily Cairo bulletin. record Illinois. Perhaps the phrase emerged from the central state as established descendants of earlier settlers formed an upper-crust class.

But the precise origins of this phrase remain unknown. Just another mystery of the English language and American custom for the books.