Our Magnificent Bastard Tongue by John McWhorter

One of the benefits of being an armchair linguist is that I have absolutely no qualms about veering from, say, Baugh and Cable’s A History of the English Language or the nominally rebellious but practically canonical works of David Crystal to less academic but infinitely more pleasurable works of dedicated amateurs like Bill Bryson. Our Magnificant Bastard Tongue falls into the latter category (though McWhorter sometimes resembles Crystal in tone), not only because McWhorter is a sort of nuovo-linguist, the sort who would wear sneakers before tweed jackets, but also because this particular book was intended to be a shorter and more information introduction to McWhorter’s sphere… essentially a 250-page brochure for modern linguistics.

Celtic

McWhorter’s particular niche is creole, but this isn’t a book about creole, as it described a time long before any creole we’d be familiar with. In the context of English’s origins—as in, the displacement of native Britons by Anglo-Saxons and Nordic peoples—his driving passion is a defense of the influence of Celtic. There seems to be a resurgence of linguists jumping on this wagon—rightly, I think—lately, among them David Crystal, but because McWhorter is primary a linguist (think grammar) rather than an etymologist, most of his defenses have to do with grammatical and syntactical points than morphological ones. In particular, he cites the pervasive use of the “meaningless ‘do’” in English and the existence of the present participial form (words ending in -ing) as features seen only in Celtic and English and nowhere else, including all the other Indo-European languages which form the vast majority of English’s vocabulary and grammar. For example, in modern English we might phrase a question as

Do we derive any language features from Celtic?

But in every other language which touches English, the question invariably becomes something like:

Derive we any language features from Celtic?

Note that this latter form is grammatically correct English as well, although it sounds old-fashioned, like something out of Shakespeare or a bad actor from Medieval Times. More to the point, however, this latter form communicates the same thing without the odd insertion of “do”. So what does do… do? Not much, but because it happened to be a basic feature of Celtic, it found its way into Welsh and Cornish. Consider McWhorter’s examples from Welsh (English on the left, Welsh on the right, obviously):

Did I open? Nes i agor
I did not open. Nes i ddim agor
I opened. Nes i agor.

There are others, of course, but I won’t describe them all. The “meaningless do” is McWhorter’s favorite, and I think his most compelling argument for rethinking Celtic’s linguistic influence on English. Most histories of the language seem to assume that when mainlanders invading the British Isles, the existing Celts were either wiped out, or integrated so quickly and completely into the invaders’ culture that little trace was left to the posterity of language aside from a smattering of Celtic place names. McWhorter points out, rightly, that this notion is unlikely, the former because of recent DNA evidence which seems to indicate that the British don’t share as much genetic legacy with their neighbors across the North Sea as a genocide/displacement would necessitate, and the latter because historical records seem to indicate the subjugated wealhs (Welsh) as a subservient but nonetheless categorical group. The dark time in America’s history when African blacks were brutally subjugated did not see African culture dissolve, but rather the culture was absorbed to some degree into the host culture.

Semitic

After coming out strong with an excellent section about Celtic, and additional interesting reading about the Nordic influences on English which are neither new nor controversial, the book sags a bit in the middle by switching tacks entirely to talk about the Sapir-Whorf hypothesis, and why its largely bunkum. You might remember this from Guy Deustcher’s Through the Looking Glass, which is a fuller and more effective treatment. McWhorter’s examination lasts only a chapter, and seems like an odd interlude between his exposé on Celtic and his final chapter on the influence of Semitic (the language family that gives us Arabic and Hebrew) on Proto-Germanic.

This latter notion is new to me, and admittedly it isn’t nearly as compelling as McWhorter’s views on Celtic, but it makes for interesting reading nonetheless. He notes a couple of coincidences. First, the Proto-Germanic feature wherein consonants remain constant while vowels change to case, and he asserts the same pattern holds true in Semitic languages as well. It might be a misunderstanding on my part, but based on what I read in And God Said, this seems like an oversimplification of Semitic word construction, and I’m leery of accepting it on face value. To make a long story short, McWhorter’s hypothesis is that Semitic influenced Proto-Germanic by way of the Phoenicians (and that it was at least partially responsible for the shift described by Grimm’s Law). It’s a bit of a wild swing, though admittedly McWhorter understands this and doesn’t phrase this chapter as boldly as his first, instead opting for a lot of “Wouldn’t it be interesting if…” and “What a strange coincidence that…”.

It’s a rather limp-wristed and unfortunate way to end an interesting little book. Though I like a bit of speculation as much as the next guy, I feel like McWhorter had an excellent 150 pages about Celtic and Nordic influences and wanted to ride that wave to his guesswork about Semitic. And inserted a chapter on Sapir-Whorf just for the hell of it? Maybe to get a round 256 pages? In any case, Our Magnificent Bastard Tongue was 50% excellent and 50% simply OK. I will say that it was refreshing to get a purely grammatical treatment of English’s history instead of the usually etymological approach, and I think it shows off McWhorter’s skill as a linguist and a writer. Perhaps one of his other books, such as The Power of Babel, is a more thoroughgoing affair.

Murder is a word which is familiar to just about every English speaker in the world; it gets used every day in newspapers and television, especially given the glut of crime shows on the air recently. Every Law & Order or CSI uses it, though mostly in the criminological sense of an intentional killing (“malice aforethought” is, I believe, the part official definition for murder in the first degree).

Murder, Saxon-style

Murder is an old word, and even though our modern “murder” technically comes from the Latinate murdrum, the word has a long history in Germanic languages as well. Our first recorded use of the Germanic form is in Beowulf, the famous Anglo-Saxon epic, as morðor. It’s important to note that the Old English use of the word didn’t refer to killing in general, but to killing which is unlawful or wicked. These were more barbarous times, remember, so there were a number of socially-acceptable situations where killing somebody else was just fine (and therefore not morðor)). David Crystal notes an additional facet, which is that murder was carried out in secret; public killing, even if unlawful, was not a crime so much as a personal slight which could be remedied by further killing or by payment.

Thus is the Proto-Indo-European *mrtro- (“killing”, from *mr-, “to die”) comes to give us both branches, as well as eventually giving is mortal[ity] as well. And in both branches, the notion of clandestine killing was there. Those readers with an interest in fantasy literature may have also noticed the close resemblance of morðor to the legendary and evil land of Mordor in J.R.R. Tolkein’s Lord of the Rings trilogy. Though the ð in the former is not a d, but rather an eth, pronounced like the “th” in “them”, the similarity is indisputable…. especially when we remember that Tolkein was learned of Old English and Beowulf in particular. An occult, secretive land associated with death is right on target with what we know from the books.

We have Old English myrþrian (“to murder”), the Gothic maurþr, the Old Norse morð, and Frankish *murþra. This last item lead to the Old French murdre, the Latin murdrum, and the Anglo-Norman murdre. Middle English’s murder, murdre, mourdre comes from an alteration of an early form, murther, which appears to have come from the Old English form. It may have been influenced by the th→d switch of the French/Latinate route, though Middle English sometimes performed this change without any help from the French.

Taking -cides

On the aforementioned crime shows, you’re likely to hear the word “homicide”, which combines homo- (Latin for “man”) with the suffix -cide, which is a construction of Old French inherited from the Latin suffix -cidium, which meant, roughly, “the act of killing”. The suffix ultimately comes from PIE (*kae-id-) and has to do with “striking”.

“Homicide” as we know it is a straight orthographic import from Old French, in which it appeared as early as the 13th century.

We have plenty of other constructions that use the form, either directly from Latin or backformed to match the pattern:

• “Suicide” is from the 17th century, from the Latin suicidiumsui- (“of oneself”) + -cide.
• “Fratricide” is mid-15th-century, from frater-(brother)
• “Uxoricide” was probably formed by analogy, since it wasn’t recorded until the mid-19th century. uxor is Latin for “wife”.
• “Genocide” was definitely formed by analogy, mostly likely by Raphael Lemkin in 1944. It comes from the Greek root genos (“race” or “kind”), where whence genus as a taxonomic classification.

Murder, Hebrew-style

It would be impossible for any self-respecting etymologist to talk about the word murder without referencing the King James translation of the fifth commandment—Thou Shalt Not Murder.

The actual Hebrew word used is r’tsach, and is variously translated as “kill” and “murder”, though these are two very different ideas, at least in antiquity.

As Joel Hoffman notes in And God Said, the book of Numbers provides much more in the way of context. Killing was legal, in a sense, except that in the Hebrew tradition of “blood redemption” (blood redeemer is go’el hadam), one’s male family member(s) would be morally obligated to hunt down the killer and kill him in turn, much like in later Anglo-Saxon societies.

According to Hoffman, there are four “kill” words that need considering. Cain haraged Abel, Moses hikahed the Egyptian taskmaster, and Abraham asks Yahweh not to heimit any righteous people in Sodom. Finally, ratasch, a form of the earlier r’tsach, is used both in the Ten Commandments as well as in the codification of honor killing described in Numbers. The wordhikah seems to mean “strike”, either fatally or not, though the KJV tends to imply the latter because it frequently translated it as “smite” or “smote”. Heimit seems clearly to indicate a general causing of death, without specification as to mean or result, and is used interchangeably with harag during the story of Joseph (Genesis 37:18-20).

To ratsach, as Numbers tells us, sometimes happens if you hikah someone and he dies. You can also ratsach someone accidentally, implying that to ratsach someone is to kill them, possibly by accident. The blood redeemer, however must heimit the ratsacher. It any case, the term used means “illegal killing” (though one might think that Cain’s haraging of Abel would count, too), or “murder” or “manslaughter”, since other forms of killing (such as when a blood redeemer heimited somebody) we considered, well, kosher.

Through the Language Glass: Why the World Looks Different in Other Languages by Guy Deutscher

Those of you who know grammar snobs (I am admittedly one, myself) may be tempted to think that we are all William Safires, wrinkling our noses at slang and treating publicly-dangled participles like a fart in church. In fact, linguists (as distinct from grammatical pedants) are a pretty liberal bunch, or at least those of us studying since the second half of the twentieth century.

I add that last clause because the study of linguistics—especially its intersection with sociology or anthropology—has had a rather dialectical history. It was not altogether long ago that the study of languages and what they did was the sole province of rich white males, who did what rich white males have (generally) historically done, namely sneer at cultures different from their own. When colonial Europeans found their way into Africa and encountered tribes with languages consisting of clicking noises, for instance, the conclusions reached were that indigenous languages were primitive languages for primitive people. In other words, language reflects its speakers; to a similar degree, language affects its speakers. It was not until sixty or so years that the trend in academia became the mantra that “There are no primitive languages”, which is more or less true. But it also engendered a form of strict relativism that stifled the study of the relationship between language and culture.

Though Deutscher covers several topics, his longest and most important is about the language of color, and he begins in 1858 in London, England. William Ewart Gladstone, though otherwise unremarkable, publishes a huge (1700+ pages) tome entitled Studies on Homer and the Homeric Age, which was both widely read (and thrusting Gladstone into some degree of literary prominence) and widely panned by critics. Amidst its enormous bulk, however, was a jewel of a section concerning Homer’s use of color words. They are all wrong, you see; the famous “wine-dark sea” (literally “wine-looking sea”) doesn’t make much sense unless you allow for significant poetic license, and that’s only one of many examples. There were three problems, which I crib here from Deutscher:

1. The use of the same word to denote colors which, according to us, are essentially different.
2. The description of the same object under epithets of color of color fundamentally disagreeing one from the other.
3. The slight use of color, and its absence from certain cases where we might confidently expect it.
4. The vast predominance of the most crude and elemental forms of color, black and white, over every other.
5. The small size of Homer’s color vocabulary.

Gladstone’s proposed explanation? The ancient Greeks were largely color-blind, except for perhaps red,, and the facility of color vision was a recent evolution in human eyesight. This is, of course, not the case (not quite, anyway), though one must credit Gladstone not only for his meticulous argument but for his foresight as well.

I will spare you the intervening story (which is quite good) and skip to the general conclusion, namely that while the biological mechanism of sight was not deficient in the ancient Greeks, and has not changed appreciably since their era, it turns out that there is an identifiable pattern to the way that cultures categorize and name colors. Much of it has to do with the necessity of differentiating colors, which follows (to some degree) the spectra in ascending order (e.g. red is first) and has much to do with the availability of the color in vivid and/or available artificial forms. Though a plausible (initially) hypothesis is that the eyes of “evolved” societies had evolved the capacity for higher wavelengths of light, the most likely scenario is that colors like red are more readily found in nature, not simply in the form of bright red blood but in materials for dyes. Though even primitive cultures can see a large blue object (the sky), blue and purple dyes are much more rare, and their existence in nature less vivid. It may come as no surprise, then, to find a culture which lumps English-speakers’ “yellow” and “green” together under one color, even though it seems clear to us that they are distinct are merit distinct names.

To illustrate why this is a fruitless conceit on our part, Deutscher gives the example of the Russian синий (siniy) and голубой (goluboy), which are Russian words for dark blue and light blue, respectively. An English-speaker, if pressed, could offer the “navy” for dark blue, but Russian is distinct in that the division between dark and light blue is as clear within their language as green and yellow for us, and our perception of visual information (not the biomechanical processing of light in the eyes, but our internal categorization of colors) directly affects brain response during experiments. In other words, Russians don’t see blue any different than Americans, but because the Russian language and Russian culture differentiates and categorizes siniy and goluboy, how Russian speakers perceive shades of blue is therefore affected. Perhaps we, with our single word for both light and dark blue, appears to Russians as linguistically-impoverished savages.

Of course, none of this is particularly controversial except to suggest that the language we speak affects how we think. It stands apart from the old canards that Deutscher mentions in his intro, such as the orderliness of the German language indicating the orderliness of the German mind; the silky romance of French indicates the silky romance of Frenchmen; and so on and so forth. All it means, of course, is that language both reflects and influences our taxonomy of the world—nothing less, nothing more. The same principles at work in the section on color also work for the Aboriginal language that uses only cardinal direction, not speaker-relative direction (e.g., “south” instead of “left (if you’re facing west)”). But such differences are narrow in their focus and not tied to larger notions of culture and ill-informed cultural conceits.

The central theme of Through the Language Glass is not groundbreaking; it’s not even particularly surprising once you think about it. Significantly, though, it’s an area one is not likely to think about, since native language is so apt to produce blind spots. Even simple hypothetical scenarios posited by Deutscher are enough to make one realize just how absurd and arbitrary our notions of foreign languages (and their hypothesized effects) really are. In that respect, it’s a startling and refreshing book.

A coworker mused aloud just the other day, “Why is it we have bedroom, dining room, living room, bathroom and….. kitchen.”

Why indeed? Of course, we are simplifying things a bit too much, excluding even current room names like basement, foyer, and office, and more archaic room names like boudoir, parlor, and study. But nonetheless, why the preponderance of -rooms and the rather unique “kitchen” in our modern household terminology?

«Kitchen» by Vincenzo Campi, c. 1580

We’ll begin at the beginning. The Proto-Indo-European language had a root, *pekʷ-, which referred to cooking. As I mentioned several weeks ago in my article about numeral names, the Latin language had a tendency to take any PIE roots which contained an initial p- and internal kʷ and assimilate the p into another kʷ. In this case, the Latin gives us coquere (to cook) and coquina (kitchen); the French borrowed it as cuisine (which, as you probably noticed, we in turn borrowed from the French as something slightly different); the Germanic languages borrowed it as *kocina, and by the time it was appropriated into Old English, it had become cycene, and by the time of Middle English it had become kichene, recognizable as a recent relative of our own modern form.

It’s important to note that the word was in place in the English language well before the 12th century, which may have something to do with its longevity. As we will see, our other modern -room constructions are relatively recent replacements of more complicated things. The word “room” itself, by the way, is adapted from the Old English rum (← Proto-Germanic *ruman ← P.I.E. *rew- (“wide” or “open”), and originally referred to any space, but came to refer to spaces within houses sometime in the mid-15th century when it replaced the Old English cofa (our modern “cove”).

Take the modern “bathroom”, which only dates back to 1780, a period well after the greatest flux in English was already done. Clearly a composite word (and an Americanism, at that), it replaced the more traditional lavatory, a 14th century borrowing from Latin lavatorium. While lavatory initially referred to a washbasin, it later gained the sense of a “washroom”, and finally took to mean a toilet in the early 20th century. The word “bath” itself is another Anglo-Saxon word: the Old English is bæð, from a PIE root meaning “to warm”, which also gave is the modern word “foment” via Latin.

“Bedroom” dates back a little father, having cropped up in the early 17th century when it replaced “bedchamber” in the popular tongue. The latter, clearly, is also a composite word which could have only dated from the early 12th century, as the word “chamber” didn’t enter English until then via French (chambre) from Latin (camera) and Ancient Greek (καμάρα, or kamara, a vaulted chamber). The more important part of that composite word, “bed”, is from the Old English bedd, a workhorse word which referred to bed both in the sense of a place for sleeping as well as a place for gardening—e.g. a “bed of roses” (the P.I.E., *bhedh-, referred to digging and clearly meant the latter).

“Living room” in the sense of a general room for entertaining or residing is the most recent coining yet, not appearing until 1825, and probably formed by analogy with its older composite cousins “bedroom” and “dining room”. “Living” in the sense of dwelling or residing in a particular place, comes from the 14th-century Old English lifiende, from the same P.I.E. root that gives us the modern “leave”.

But speaking of dining rooms, they cropped up in the language around the same time as “bedroom” (early 17th century), another easy composite of “dine” and “room”; the important root here came from the Old French disner and therefore very obviously from the Vulgar Latin *disjejunare. What you might not realize is that the same Latin root that gives us the modern French loanword jejune, which refers to metaphorical lack of substance but apparently got its meaning from a more literal sort of hunger.

That’s largely it for -room rooms, which doesn’t seem like a lot until you consider that such rooms are also the most common and frequently mentioned rooms of a house, which may account for some of their notoriety. The general trend seems to be that our modern usage favors recent constructions or popularizations, and the times at which such constructions entered the language seemed to favor Anglo-Saxon words, with the notable exception of the dining room.

But what of the other notable components of a house? How did we come to refer to a garage and a basement and a parlor? A study? A foyer?

Most are easy: a “study” in the sense of a room filled with desks and books is an early 14th-century coinage, unsurprisingly from French (estudier, from the Latin studiare connoting diligence, from the P.I.E. *(s)teu-, which meant to “knock” or “beat”), since Old English would still have been at that point the tongue of the lower class, and any person who could both afford and use an office would likely be both wealthy and learnéd and therefore intimately familiar with French and/or Latin.

Though its largely fallen out of the language now, “parlor” also came from French—specifically parleor—and by the late 14th century had come to refer to a specific room of the house meant for private conversation. By a number of twists and turns, the word comes from the Latin parabola (“comparison”), which in one branch came to refer to speech and thus gives us “parley” and “parlance” and “parable”: the parabola of geometry is a different idea entirely.

One word that was never adapted from the French, but rather borrowed directly, is foyer, which we now use interchangeably with “entryway” but which originally referred to an off-stage area for actors. Curiously, its literal definition is “fireplace”, inheriting that meaning from its Latin origin word of focus. If that confuses you, because you think of “focus” as a point of visual convergence, that’s because the modern definition didn’t come around until the aspiring astronomer Johannes Kepler neologized it in 1604. The change in meaning for “foyer” seems to have come about in the sense that this off-stage area could also be the lobby of many theaters, and so the sense of a foyer as both a hearth and a lobby came into popular notion. Only the latter sense, to my knowledge, survives today in American English.

“Basement” is an 18th century coinage, and yet another composite word combining the root base (another Latinate word via French, and the same origin of our “basis” and “basic”) with the suffix -ment. This is an interesting case, because though -ment is a common Latinate suffix use in English to transform verbs into nouns (e.g. punish→punishment, banish→banishment, argue→argument), Doug Harper places the room name in 1730 but the verb of “base” in the sense of “to place on a foundation” doesn’t arise until 1841, meaning that the indicated construction must have arisen from an existing verb form or be something else entirely.

The garage is the most recent construction of all, which is no surprise considering that until the advent of automobiles, there was no such structure (just stables, I suppose). The word, as well as its pronunciation if you use the soft internal “zh” (i.e. “guh-razh”), is a transplant from French, which was generalized as a shelter of any sort and at one point referred to the docking of ships. Though most French words are Latinate, this particular verb came from the Frankish *waron and the P.I.E. root *wer-, meaning “to cover”. This is the same root, by the way, which gives us “warrant” and “warranty”, a word that was imported into English again as guarantee, a linguistic slight of hand very common for initial g and w, since neither originally existed (written as a “yogh” or ȝ until systematically replaced by Norman scribes).

I end on the garage because it inevitably begs the age-old question intoned by smartasses everywhere: why do we drive on a parkway and park on a driveway? The simple answer, of course, is that one didn’t originally park on driveways, since they tended to be long stretches of private road leading from the edges of an estate to the garage; they were, in fact, literal what their names suggested.

A parkway, on the other hand, is a very public thoroughfare, but the word originally referred to special road which was heavily landscaped and meant for pleasure cruises (read: like driving through a park); somewhat more generally, it can also refer to a highway with a landscaped median. This sense left quickly as the pace of road development turns such things into strictly utilitarian places, much to our great loss. In fairness to the word’s original meaning, “park” in the sense of an enclosed grassy space is 13th-century; the verb form that we use today with cars doesn’t show up for another 500 years or so.

Gore Vidal once famously quipped that “‘Politics’ is made up of two words, ‘poli,’ which is Greek for ‘many,’ and ‘tics,’ which are blood-sucking insects”, and that’s perhaps one of the nicer things said about the practice and its practitioners. It seems as though politics has always been reviled, even back to the earliest statesmen.

You’ll notice that I wrote “politics has” instead of “politics have“; politics, despite the terminal s, is a singular noun, though popular usage has effaced that tradition somewhat. You still hear this form in the phrase “All politics is local”. Another interesting characteristic of the word is that the noun was formed from the adjective (often it’s the other way around) politic in the 16th century. The adjective came from the French politique and the Latin politicus, which simply referred to civil or state matters. The Latin itself was modeled from the Greek politikos (πολιτικά), which, despite Vidal’s musings, actually came from the word polis, or “city”. This Greek root can be traced all the way back to the Proto-Indo-European *p(o)lH-, which indicated an enclosed spaced on high ground. This same root also gives is policy via a slightly different lineage, and from which our modern police split off in the 16th century.

The world of politics has given birth to a whole host of new words and phrases, especially the latter. Many of them are borrowings from the Latin/Greek traditions just like our judicial system: Congress talks about a quorum, for instance, which is a direct borrowing from the Latin that refers to the minimum number of representatives necessary to effect a vote.

But there are other words with slightly less obvious origins. Our modern “mayor” comes from the Old French maire of the same meaning. Ultimately, the term is derived from the Latin major, related to magnus, which gives us our modern magnum both in guns and in prophylactics.

Senator comes via French from the Latin senator, itself from senex, or “very old man”, which makes a lot of sense when you consider that most senates have been historically comprised of old white men. Not surprisingly, this same root (from the PIE *sen-, or “old”) gives us the word “senile” and “senior” as well.

Along those same lines, the less recognizable “alderman” comes from Old English aldormonn, the West Saxon ealdormann, a composite of ealder (patriarch) and mann (man). The the W.S. suffix is a cognate of our modern “elder”, and so once again the word literally means an old man, a fixture of the authority-by-seniority that marked that era.

There are far too many terms to cover them all, but they’re all interesting reading. Take “governor”, for instance, which is a 14th-century absorption from the Latin gubernatorem (director or ruler, though originally it referred to a pilot or steersman). This origin is why we talk about electing a governor during a gubernatorial race; the mutation of /b/ into /v/ was one of the many changes that marked the transition from Vulgar Latin’s diaspora into various and sundry Romance languages. That original Latin, by the way, is from the Greek kubernētikos (κυβερνητικός), which is cool for three reasons.

1. The shift from the initial k to g, according to Doug Harper, probably happened by influence from Etruscan on Greek, which is an explanation I hear for most strange things happening in Greek.
2. The Greek kubernētikos is what gives us our modern cybernetics and therefore cyborg.
3. The Governor of California is most famous for playing a cyborg know as a Terminator.

Awesome.

You may or may not know that the system of numerals (from the Middle French numéral ← Latin numerālis ← numerālis (“number” or “quantity”) ← PIE *nem (“assign, distribute, allot”)) we use is the sole province of very smart people in the near East. Though our language and literary culture is dominated by the Greeks and Romans and later by western Europe, the early days of math owes just about everything to such tongue-twistable personages as Aryabhata (positional notation), Al-Khwarizmi (algebra), and Brahmagupta (zero). What’s more, our numerals are relatively straightforward transplants from the Hindu-Arabic system of numerals. Such was the influence of the East on early math and numerical theory that there was never any real competitors in Europe—the unwieldy Roman Numeral system simply couldn’t compete, lacking any real form of positional notation or zero, and being almost impossible to work with in algebra.

The general trend in the history of English is for areas of law and learning to be Latinate (either directly from Latin in the case of judicial matters or via French in most everything else) while everyday things like swearing came from the lower-class Ango-Saxon. One would expect that, since numbers and math constitute a rather high-brow pursuit, our words for numbers should be Latinate as well. In an unexpected twist of history, this is not so, although many of our numeral names have a common ancestor for both the Latin and the Anglo-Saxon forms.

A note on vocabulary. When we talk about numbers, they can take a variety of grammatical forms. The numeral itself, e.g. “one” or 1, is known as a “cardinal” number. When using the number to describe a position in a sequence, e.g. “first” or 1st, it is an “ordinal” number (think ordinal→order). The term “multiplier” refers to the form of the number used when (wait for it) multiplying, e.g. single or double. A form which only exists for the first three numbers is the adverbial, as in “once”.

One (1)

“One” is one of those numbers that makes no sense to children, since it seem counterintuitive in it’s pronunciation: by analogy with ones like “done” it should be pronounced simply “uhn”, or perhaps “own”. In fact, it used to be pronounced as the latter, until the 14th century, when it underwent a spontaneous mutation in western England. A popular Biblical translation, the Tyndale Bible, spells it won. It comes to us from the Old English ān, the Proto-Germanic *ainaz, and the PIE *oinos, all with the same meaning of singularity. You can see the root at work in the modern German ein as well as the Latin unus, from whence “unity” and “universal”.

On a completely different etymological track, the ordinal form of one, “first”, comes from the Old English fyrst, the Proto-Germanic *furisto, and ultimately from the PIE root *pro-.

The adverbial form of one is the straight-forward “once”, from a grammatical declension of the Old English. The multiplier, single, is yet another separate etymology. It comes from the Old French sengle, the Latin singulus, a derivation of simplus, which has the same meaning and forms the root of our “simple”.

Two (2)

Another word that doesn’t quite look like it sounds is “two”, which comes from the Old English twa, a form of twegen from whence our rather archaic twain (e.g. “cleft in twain”). It’s from the Proto-Germanic *twai and ultimately the PIE *duwo, all with the same meaning (number meanings tend to stay constant over time). The migration from the original dʰ or d is another example of Grimm’s Law (see footnote), which is why Germanic languages ended up with the dental t and non-Germanic languages like Latin retained d words like duo and gave us dual. The reason why modern German has zwei instead of two is due to a second Germanic consonant shift known as the High German consonant shift; modern German descended from OHG, but other Germanic branches such as Dutch (twee) or Swedish (två) did not.

Two’s adverbial form is twice, the etymology of which is as straightfoward as it suggests. Its multiplier, double, comes from the Old French doble, from the Latin duplus, all the way back to the aforementioned duo. Duplus, by the way, still gives us duplicity and duplex, among other words.

The ordinal form of two is “second”, from the identical Old French, and finally from the Latin secundus, which means “to follow in order”. The same root which underlies this also gives us sequel and sequence

Three (3)

As Schoolhouse Rock once crooned, “Three is a magic number”. It’s the last in a well-known grouping (sort of like the modern day “A, B & C”, which is why it’s the last number to have an adverbial form (“thrice”), though even this has become archaic.

Three comes to us from the Old English þreo, from the Proto-Germanic *thrijiz, and the PIE *trejes. Once again, Grimm’s Law is at work: the original dental t of PIE gave way to a th (or θ/þ). Hence, we have three; due to the aforementioned High German consonant shift, the th morphed into an initial d, hence the German drei. This was the last phase of the High German shift, occurring in the 9th and 10th centuries, and the only part of it to work its way into Dutch as well, hence the Dutch drie.

Three’s ordinal form, third, comes again from the Old English, this time from a metathesis (rearrangement of sounds within words) of þridda; in fact, it was not uncommon to hear thrid well unto the 16th century. Ultimately, it comes from the PIE root *tritjos, which gives us tertiary (cousins of primary and secondary) from Latin.

The multiplier of three, triple, is from the Latin triplus, a combination of the tri- prefix and -plus, which means “fold”—literally “threefold”, a construction still used today. Note the discrepancy between triple and double; the absorption of this multiplier words into French often changed the internal p into a b, and so these words were effectively absorbed twice. One can still find “duple” used (especially in musical vocabulary, which was heavily influenced by the Italian, which kept the internal p), just as one can also find treble used both in music (oddly enough) and as a synonym for triple.

Four (4)

Four marks a change in our sequence of numbers; we’ve passed the more common one, two, and three and our etymologies get proportionally more complicated. The lineage itself is easy: the Old English feower comes from the Proto-Germanic *petwor- and the PIE *kwetwer-. This earliest root explains the Latin quattuor (and hence quaternary and the modern multiplier quadruple; beyond that, as Douglas Harper notes, the evolution of the Germanic is as-yet unexplained. The usual function of Grimm’s law would turn the *kʷ of the PIE root into a hw sound, which is doesn’t. The change from the Proto-Germanic p to the later Old English f, however, does appear to follow the established pattern. Modern German and Dutch vier appear to buck the trend until one realizes that the letter V (as opposed to the phoneme V is often pronounced as F in Germany words such as vier or Volk (“folk”).

In the case of four, its ordinal form is a relatively simple affair, attaching the -th suffix to the base word. This particular suffix, which applies to all numbers except the first three, is derived from the Old English -þa, itself a derivation of an Indo-European superlative suffix—that is, serving a function similar to Modern English’s -est. The multiplier maintains a similar pattern, pairing the Latin root qua[t/d] with the -uple suffix, which is from the aforementioned -plus (“fold”) suffix.

Five (5)

Five is an adaptation of the Old English fif and Proto-Germanic *fimfe, ultimately derived from the PIE *penkʷe. You can still see the PIE influence in non-Germanic branches such as the Greek pente, from whence comes our well-known prefix of penta-—e.g. pentagram, Pentecost, Pentateuch. How the Latin quinque was managed from *penkʷe is a little strange and not extensively documented. It appears, however, that in circumstances when the PIE has a *p and a following kʷ, the initial p was often assimilated, so that *penkʷe became, effectively, *kʷenkʷe… or quinque.

The ordinal “fifth” is from the Middle English fift, which was altered by analogy with the aforementioned fourth, beginning the trend that will mark the rest of the ordinals we’ll cover.

Six (6)

Six invariably makes schoolchildren giggle, since its Latinate root is sex. It gets that root from the PIE *seks, which gave rise simultaneous to the Latin as well as the Proto-Germanic *sekhs and eventually the Old English siex. The Latin gave rise to the name of the sextant, an early navigational device. In case you’re wondering, sex in the sense of coitus or gender is from a different Latin source (sexus, related to secare, “to cut”).

The ordinal sixth comes from the Old English syxte and eventually altered by analogy with the now-established pattern.

Our other well-known prefix for six, hex- (e.g. a hex-head bolt), is from the Greek ἕξ, which is Epilson and Xi, but the diacritical mark above the Epilson indicated a rough aspiration, or “H” sound.

Seven (7)

Seven is the only base numeral that is not monosyllabic. It comes from the Old English seofon, the Proto-Germanic *sebun, an the PIE *septm. In the Latinate branch, we get septem, from whence septuple (e.g. septuplets). On the Germanic side, it follows a similar evolution to our modern “heaven”, which was heofon in Old English.

Despite the similar appearance, the Latin septem is not related to the septum (the wall between the nostrils) which is from the Latin word for “fence”; nor is it related to a septic system, which is a Latin derivative of an earlier Greek word meaning “to rot”. On the subject of Greek, that language’s word for seven, hept- is one against a substitute of an initial s- for an h. You can see this same phenomenon on other words like semi/hemi (“half”), but I can’t for the life of me find any reason for it, other than a suggestion that the prehistoric Greek maintained the S.

Update: it appears that the /h/ to /s/ switch is even more regular than I implied, and that one can see the same idea at work in Old Persian and Armenian (according to this source). Why Sanskrit appeared responsible for this switch is still a mystery, but may be an effect of pronunciation on orthography.

Eight (8)

Eight demonstrates a number of interesting etymological phenomenon all at once, which is why I chose it as the star word for this entry. Originally eighte, it comes from ehte, from Old English æhta, from the Proto-Germanic *akhto(u) and finally the PIE *okto. One can easily see where the Latin root octo- came from, evident in our octopus, octothorpe, and, less appealingly, “Octo-mom” and her octuplets. Ernest Klein thinks it is a “dual form”, that is, meaning “twice four”.

How it came to be spelled “eight” as opposed to the somewhat more reasonable ehte has to do, according to Doug Harper, with a “scribal habit” of the Middle English period. It was common in Chaucer’s day to pronounce the consonants in words like fight, eight, and [k]night, so that “knight” became k’nikt. Sometimes, the gh group would be substitute by the yogh (ȝ), which represented the phoneme /x/ in this instance, though it was a workhorse symbol that served a lot of purposes for a lot of languages.
The Normans, Latinate as they were by way of French, began a program to replace all non-Latin letters with digraphs (two or more letters which replace a single one). Thus, niȝt became night and eȝt became eight, and as the spelling changed, so did the pronunciation, until finally these internal gh clusters became silent altogether.

Nine (9)

Nine comes from the Old English nigen, from the Proto-Germanic *niwun, from the PIE (e)newn. The neat and tidy etymology of this ones gives us the Latin novem (yes, from whence November), the Greek ennea, the German neun, and the Danish ni. Similarities to the German word for “no”, nein, are coincidental, since the latter comes from the PIE *ne-.

Zero (0)

Zero was a latecomer to the numeral game, as I mentioned long ago in the introduction. Though it makes perfect sense to modern readers, it is not quite the fundamental concept that, e.g. “one” is. Its etymology is just as nontraditional: from the Italian zero, it came from the Medieval Latin zephirum, itself from the Arabic صفر (sifr, our “cipher”), which is a rather poetic translation of the Sanskrit sunya-m, which means “empty place” or “desert”. In this way, it’s actually the neatest of the numerals etymologically, even if it’s something of a stepchild.

“Left” and “right” are such common words that we don’t often realize just how significant they are; but like all simple words, they tend to be venerable, storied, and much more interesting than it may first appear.

I was inspired to do these words because of an e-mail forward joking about “left” and “right” politics by quoting Ecclesiastes:

A wise man’s heart is at his right hand; but a fool’s heart at his left.

Left as we know it comes from Middle English—spelled various as left, lift, and luft—adapted from the Kentish adaptation of the Old English lyft, which meant “weak” or “foolish”. Circa 1200, when it began showing up in Middle English, it retained that same sense, but appropriated the sense being the opposite of right about a century later. It’s not a surprising event: the left side of the body has, back into antiquity, held connotations of evil, weakness, or general negativity.

The Latin word for left was sinister, and at some point gained the aforementioned connotation due to its speakers’ superstitions about the left side of things. Douglas Harper suggests this was due to Greek influence, since Romans generally regarded the left side as favorable. The connotation was strong enough that Old French adopted the word as sinistre, meaning “contrary” or “unfavorable” as well as its spatial definition. Today we no longer use sinister in its neutral form, but only as an adjective for evil. Left, meanwhile, though it began with largely the same connotations, only retains its connotations in certain phrases like “left-handed compliment”. The stigma of the left is sometimes retained in the language of other cultures in much the same way.

Most Romantic languages still use some form of the Latin sinister with the exception of Spanish, which does have the word siniestra in the English manner of something evil, but uses izquierda as the spatial left; this latter is one of a few Basque words which have found their way into mainstream Spanish.

By contrast, the Latin for right, dexter, gave us the adjective dexterous, which is full of positive connotations: in other words, right-handed persons were skilled, and the poor southpaws were considered weak at best and evil at worst. The modern “right” comes from the Old English riht, which meant “fitting”, “proper”, or “straight”. On the Latinate side, we can trace word back to the PIE root *dek[s]-, meaning “on the right hand”. Once again, there’s been a split between the spatial and connotative senses: in modern Spanish, for instance, diestro means right-handed or dexterous, but derecha means right in the sense of a la derecha, or “to the right”. In English we retain dexterous in the sense of skillful, and also “direct” from the same root.

These connotations explain many of the uses of these words—especially “right”—part from their spatial sense. But there are other connotations, too, such as the ones mentioned at the beginning of this article. The origin of political “Left” and “Right” is no secret, and gets bandied about during election cycles or by the annoying guy at parties. In fact, “Left” and “Right” are truncations of “Left-Wing” and “Right-Wing”, which refer to literal wings (as of a building) to the left and right of the president of Parliament during the French Revolution (circa 1789). The terms didn’t becoming commonplace in British or American political parlance for another century, however.

Codicil is known mostly as a legal term (for which see the official definition), but in practice is has come to refer figuratively to any addition or addendum, often with a quasi-scholarly connotation. Its use in English dates from the 15th century, when it came into the language from the French codicille and Latin codicillus , which referred to a short writing or small tablet (used for writing). It’s no surprise that the word’s origin is French/Latinate, since most of our legal terms come from that very source. Because French and Latin was, for a long time, the preferred language of the scholars and the judicial system after the Norman Conquest, our common words from that vocabulary Latinate almost to a one.

Codicil is a diminutive form of codex, which was Latin for both “tree trunk” and “book”, and which also gave rise to the more familiar code, initially in the form of a code of law or code of ethics, but which now refers to everything from the cheat code in Contra to the source code that I write at work.

Similar but distinct from a codicil is a corollary, which is not so much an addendum as it is a peripheral consequence of something. The official definition per Wiktionary is “Something which occurs a fortiori, as a result of another effort without significant additional effort.” From the Latin corōllārium, it originally referred to a gratuity, and was in fact ultimately from another diminutive form, this time of the Latin corona , or “crown”, and which is memorialized both in the terrible Mexican beer and the glowing rim of hot matter that surrounds stars. On the day this entry was written, in fact, the Earth was hit with a “coronal mass ejection”.

Though spelled similarly and with somewhat similar meanings, “ancillary” (“subordinate” or “auxiliary”) is not connected to corollary. It’s from the Latin ancillāris, from a word meaning “maid.” Ultimately, it derives from the well-known Proto-Indo-European root *kwol-o-, which means “move round, turn about, be much about”, and which is, appropriately, the origin of our word “cycle”.

When coming up with synonyms and related words for this week, it was difficult not to notice the surfeit of app- and add- roots. Consider appendix, which is English as well as Latin, and comes from appendō in the latter: ad (“on” or “against”) and pendō (“I hang”). Append is essentially the same word, except that it uses the suffix -ix, which is a common Latin suffix used in feminine agent nouns. It can also be found in words like cicatrix (“scar”, of unknown origin) and dominatrix, though generally speaking the feminine -ix suffix has been replaced with -[tr]ess suffix in English, and so words like actrix became “actress” (perhaps mediated by the French actrice).

But what of -add? Our word “add” (as in adding 2 + 2) is from the Latin addere, which means “to add”. Our addendum is a pure absorption of a Latin declension of that word, maintains roughly the same meaning: something that has or will be added [to something else].

As we’ve seen before in cases where it seemed to modern ears though words would be related, a combination of shared common structure, coincidence, and occasionally common ancestors lead to the pool of similar-ish words in today’s Word.

The modern English Color is now the same as the Latin from which it came, though the intervening steps are not: the Latin led to the Old French color, which led to the Anglo-Norman colur, which visited Middle English as colour. The Old Latin root is colos, which referred not to color in general but any sort of covering, which contributed to the earliest sense of the world, which referred to the color of the skin or complexion in particular. The Old Latin comes from the PIE *kel-, meaning to cover or conceal. Our modern definition is from the 14th century, from Middle English, at which point it had replaced the word previously employed, blee.

Blee was a perfectly lovely word, from the Old English blēo, and I’m sorry it left the language. It came from the Proto-Germanic *blījan (“light” or “happy”), itself from PIE *bhlē̆i-, which also meant “light” in color or complexion. Along an evolutionary fork, it gave us the Old English blīþe, from whence “blithe” (as in “blithely”), which meant a light of mood (“glad”) rather than of color.

Of course, as anyone who’s ever looked at a box of crayons can attest, there is a wide variety of color words in use, despite the relatively circumscribed nature of our word for the phenomenon in general.

    blue

Our modern “blue” carried forward from the Middle English blue, though it changed from blewe or blwe. The word we know is a deriviation from the Middle French bleu (think “chicken cordon bleu“, or literally “blue-ribbon chicken”). The Old French was blo, which had both the sense of “pale” and “light-colored” as well as referring specifically to the hue blue, and came from the Frankish blao and ultimately from the Proto-Germanic blæwaz, which is where most of the Germanic languages derive their words for it: Icelandic blár; Dutch blauw, German blau.

Interestingly, the etymology for “blue” is tied up with that of “yellow”. The Middle High German word for the latter was bla, for instance. This likely because the PIE root which, coincidentally, gave us blee (*bhlē̆i-) also referred to blue or yellow hues.

A short note on “indigo”: Those who remember their mnemonic for the colors of the spectrum may also note that both blue and “indigo” make an appearance. Indigo is officially a spectral color (as initially defined by Isaac Newton), but the human eye is generally unable to distinguish it from blue and violet, its neighboring colors. The word comes from the Greek indikon (ἰνδικόν), or “Indian dye”, referring specifically to the blue dye coming out of India at that time; it was first used to refer to a color in 1289. The Greek word for India came from the Old Persian hinduš (هند) and ultimately from the Sanskrit for river: síndhu

    yellow

The “yellow” we know and loves comes from a different path, however. The Middle English yelwe comes from the Old English geolwe, which itself is from the Proto-Germanic gelwaz, from whence also the modern German gelb and Swedish gul. The earliest known reference we have to the word is in Beowulf (XXXVI), which talks about a geolwe linde (yellow linden [tree]).

The Latin for yellow or blond is flāvus, which came from that same PIE root, *bʰlē̆i. But the Germanic branch comes from a different root, ĝʰel, which means “to shine”, and is the root not only for “yellow” but also “green” in some languages.

Many historical languages used their word for yellow to refer to both the hue and to the blond color of hair or complexion. Our modern word “blond[e]“, alas, has no etymological connection; it’s ultimately from the Frankish blund, which refers to a mix of yellow and brown specifically, but blended or mixed colors generally—and, you guessed it, “blond” is related to “blend” at the PIE level.

    red

Red is one of those simple, timeless words that seems like a cognate in just about every language. It’s from the PIE *h₁roudhós, which gave rise to everything from the Sanskrit rudhirá (रुधिर) to the Greek erythrós (ἐρυθρός) to the Latin rūfus to the German rot. Our modern “red” comes by way of the Proto-Germanic rauđaz and the Old English rēad. It’s also the source of “ruby” (from the Latin rubeo, or “I am red”) and ruddy (from the Old English rudiġ, from the same PIE root).

Red is unique because, though one might think that our basic colors would have simple and definite roots, it is the only color for which there is a single, definitive root in Proto-Indo-European.

    green

One might expect the etymological origins of a word to relate to the concepts our ancient forebears word have described with them. Blue for the ocean, perhaps, or red for blood (in the case of the Sanskrit rudhirá, this is actually the case, but it didn’t appear to have that meaning in PIE). So far this isn’t so (although “rust” is ultimately derived from the PIE red).

That changes with green, which has (I think) the neatest etymology of all the colors. We get it from the Old English grēne, via the Proto-Germanic *ʒrōniaz, and ultimately from the PIE root *gʰrōni-, which means “to grow”; the corollary to this is that when most plants are healthy and grow, they also turn or stay “green” as well. In this case, it appears likely that the word which described a phenomenon of healthy plants also came to describe their color, as well. The word had split off by the time of Old English, so that “grow” is growan.

    orange

Orange, like purple, has no other word in the English language that rhymes with it. It is also semi-famous for being both a color and a fruit; the fruit existed before the color. When the word first arrived in a form which would look familiar to modern speakers, it was orenge, from Old French by virtue of Anglo-Norman, which likely borrowed it from some construct of Italian. What we know for certain is that the ultimate origin is Sanskrit (नारङ्ग, or nāraṅgaḥ) by way of one or more Dravidian languages such as Tamil. The Spanish naranja borrows directly from this lineage.

Because the geographic distribution of oranges excluded much of the Western, English-speaking world for a long time, orange the color was referred to as geoluhread—which, if you’ve been paying attention to the Old English words so far, is a compound noun meaning “yellow-read”, which is not only straightforward but accurate as well.

    purple/violet

Our modern “purple” comes with only minor orthographic changes from the Old English purpul (first recorded in 10th century Northumbrian text), which got it from the Latin purpura, which got it from the Greek porphyra (πορφύρα), which was used to refer specifically to a color of snail-derived dye (called “Tyrian purple”) used extensively by the Phoenicians. It was expensive, and used largely by royalty, which is why deep purple is still associated today with royalty. The Hebrews called it argaman (ארגמן), though their word means “crimson” and not purple.

Though we most commonly use “purple” to refer to the color, the correct spectral term is “violet”; purple occurs somewhere between violet and red. That is to say, “violet” is a wavelength of light; “purple” is a combination of blue and red light. As with the orange, the term “violet” to refer to a color arose from the flower known as “violet” and not the other way around.

    white

White is not technically a color, but rather the effect produces by unsplit spectral light. Regardless, it’s often thought of as a color, certainly in terms of cultural or linguistic connotation. We English speakers get it from the Old English hwīt, from Proto-Germanic *khwitaz, and you can still see it in Icelandic as hvítt and slightly less so in the German weiß. Ultimately, it’s from the PIE kwintos, or “bright”, and which makes sense, because a white object is reflecting all visible spectra of light back at the observer’s eyes.

    brown

Brown is not a spectral color; it’s not even a particularly well-defined color. In fact, the word comes from the Old English brún, which can refer to just about any dark or fuscuous color; the meaning as we know dates from Middle English in the 14th century. The ancestor of this word was the Proto-Germanic brûnoz, which referred to dark colors with a shine (think polished wood), and which also gave rise to the word “burnish”.

The ultimate origin of “brown” is the PIE *bher-, which is related to *bheros, from whence words like “beaver” and “bear” (in the sense of a “dark animal”).

    grey

Grey first arrived in English in about 700 C.E. as grǣġ from the Proto-Germanic *grēwjaz and the PIE ghreghwos, which, strangely, means “to shine or glow”. Ironically, IE had a number of roots which refer to grayness (pel-, sal-, smel-, k̑as-, and k̑ei-, the latter of which had a meaning closer to the obsolete meaning of “brown” in the sense of any dark or dull color).

Most Americans spell it “gray”, but for most of the word’s life as we know it, “grey” is the accepted spelling. In fact, the Americanized “gray” didn’t appear until the 19th century, along with the other changes that distanced American English from British English (such as color → colour, organise → organize, and theatre → theater).

    black

Black is everything or nothing: in substractive color, such as paints, it is created by combining all pigments together. In spectral light, it is created by the total absence of visible light spectra. We get it from the recognizable Middle English blak, from the Old English blæc, from the Proto-Germanic *blakaz (which denoted something burnt), and ultimately from the PIE *bhleg- (to burn, or to shine).

There is an interesting double-meaning inherent here: we can understand that the end result of a conflagration is blackened fuel, and in that regard the root makes perfect sense. But technically speaking it refers to burning or shining, which seems the opposite of our notion of blackness. Other words for burning derive from the similar *bʰel- (i.e. the Latin flagrare, from whence “conflagration”). Compare the Old English blæc (“black”) to its companion word, blac (“bright” or “shining”). Previously, Old English used sweart for black, from whence “swarthy” and “sordid”.

Cultural Connotations (updated on 25 June 2010)

When I originally did this entry, I had thought about including a section on the cultural connotations of the various covered colours, but the entry was already one of the longer Wednesday’s Words that I’d done, and the process seemed a bit too daunting. I just today happened across this handy chart that summarizes it. It’s hosted locally (click for a larger version), but comes from the creative folks at Information is Beautiful.

Ketchup seems as American as apple pie (which itself is English, not American), but just like the pizza we know and love originated in Greece, so the tomato ketchup we use today has a history very different from Heinz 57.

The origins of the word come from a Chinese dialect: 鮭汁, or kê-chiap (“brine of fish”), which was taken into Malay as kicap (pronounced “kichap” but also spelled as kecap and ketjap). Our early Anglicization was catchup (c. 1690), which transmuted into catsup (first used by Jonathan Swift, by all appearances, in 1730), which is the still-used alternative to “ketchup”. Our modern firm first appeared in 1711 in An Account of the Trade in India by Charles Locklear.

Though the lineage here seems straightforward enough, some have advanced the notion that our ketchup is a cognate of the French escavèche (“food in sauce”) but also more importantly the Spanish/Portuguese escabeche, which refers both to a style of food and the brine-like sauce used to marinate it. The word has been traced back to al-sikbaj, of which Karen Hess’ proposed iskebey may be a poor transliteration. Regardless, the etymology refers to a pickled dish: in the former’s case, it comes from the Persian sik (“vinegar”) and ba (“food”).

If you look carefully at a bottle of ketchup, it will likely refer to itself as “tomato ketchup”; this may seem redundant until you realize that a tomato-based version was a fairly recent change in the evolution of the condiment (specifically, the very early 19th century). Though the original may have been fish, its earliest forms were mushroom, walnut, and other things that don’t sound nearly as appetizing (in fact, as late as the early 18th century, tomatoes were considered poisonous).

It has also been suggested that the early origins of ketchup—that is, as kê-chiap—eventually led to the modern condiment known as soy sauce, more popular in inland regions of China. In countries whose cuisine more prominently features seafood, fish sauce is still alive and well: in Vietnam, it is known as nước mắm; in Korea it is aek jeot; &tc. How much it resembles the earliest forms of ketchup, however, is anyone’s guess.