Monthly Archives: December 2022

Butterflies of the mind

Picture a butterfly. What does it look like? What is its dominant colour?

Did you say blue? You might not have, but when I look at renditions of butterflies in art and crafts, they seem to have quite a lot of blue. For example, a business near where I live calls itself Monarch Dentistry, and its sign features a blue butterfly. 

As you may know, monarchs are not blue – they’re orangeish and black.

I’ve flown across the continent to visit my parents; I’m sitting at their table writing this. They happen to have a few art and craft butterflies around the house, so I’ve taken pictures of them (and mixed in one from my own place). Have a look at them. All but one has at least some blue (or at least bluish-purple) in it.

Do you reckon they’re representative of butterflies in the real world, or do they skew in a certain direction? If they skew, is it because we think of butterflies as pretty, and of certain colours and features as prettier than others? Is there something essentially butterflyfish about blue, or vice versa? I won’t say there is, but I certainly wouldn’t say there is something un-butterflyish about it. A language might or might not distinguish ‘butterfly’ and ‘moth’, but you wouldn’t expect a language to have one word for ‘butterfly’ and a different word for ‘blue butterfly’.

This brings us again to the butterfly effect. No one has observed the butterfly effect; it’s not observable. It’s an imagining produced by observation, inference, and desire. Likewise, we see butterflies, we infer things about what butterflies are like – from that and from other things that seem relevant, such as that pretty things have pretty features – and we decide what is so on the basis of what we want to be so. Let’s call that the blue butterfly effect. It may be why blue seems to be a butterfly colour. Or it may just be why I think blue seems to be a butterfly colour. Perhaps my data is skewed. Perhaps it just makes for something catchy to write about.

But we definitely expect pretty things to have pretty features, and we definitely think of butterflies as pretty – pretty, decorative, insubstantial, unserious. A “social butterfly” is someone who flits around from person to person and has no depth. It is, as I said in part 11, the curse of the charming and decorative. I went to the Art Gallery of Ontario last week and looked at all the art in all the galleries to see what butterflies I could find. In total, in all the artworks in all the galleries, I saw four butterflies, and all of them were in the same painting (The Wisdom of the Universe, by Christi Belcourt). Then I went into the gift shop, where I immediately saw several times as many butterflies just on one item (a jigsaw puzzle box) – and more elsewhere around the store. 

Butterflies, in our minds, are decoration, not art; they are for gifts, not presence. Butterflies are not deep and moody. Well, perhaps they are, but we don’t think of them as such. No one ponders the inner life of a butterfly. In our culture, we have projected on them an insubstantial, carefree quality, because that’s how they seem and that’s what we want them to be, and we can neither confirm nor disconfirm it; it’s a blue butterfly effect.

So, now, what about words for ‘butterfly’? Imagine you’re creating a language – a whole new language, with new grammar and all new words. What will be your word for ‘butterfly’?

I’m genuinely curious at what word comes to mind right away – and whether you would change it to something else after a bit more reflection. Will you choose a pretty word? Pretty in what way? Or will you deliberately choose something unpretty – hengzkog, say, or, um, boterschijte? As we’ve seen, the words for ‘butterfly’ in the languages of our planet are diverse. But if I were to present you a few made-up words – say, gayok, henjitum, dwesp, fenufon, and pilapila, which one would you say is most suitable for a butterfly?

I have no statistics, and it would be difficult to produce ones that didn’t have confounding factors skewing them, but my impression, from the words we have seen over the course of my articles, is that there are three things that seem more common than chance in words for ‘butterfly’: a liquid (/r/ or /l/), probably in the middle of the word somewhere; a labial or labiodental (/p/ or /f/ or sometimes /b/ or /m/) at or near the start of the word; and reduplication. Of these, reduplication is probably the least common, but it still stands out, as we have seen. And, on the other hand, velar consonants (/g/ and /k/ in particular) seem relatively uncommon, as do back vowels (/o/ and /u/). In general, butterfly words are at or near the front of the mouth – they are not, physically, deep.

But is this a real tendency, or just my impression of a tendency? How much is observation, how much is inference, and how much is desire? And if it is a real tendency, is it an effect of something – such as sound symbolism (the same thing that causes people to tend to assume that a word like kiki is better suited to a pointy shape and a word like bouba is better suited to a round shape) or ideas of what sounds are prettier? As with any butterfly effect, we have no real way of knowing for sure. Still, we might get a better idea from words that people have made up for ‘butterfly’ when they’ve made up their own languages.

That’s just might, though. Among constructed languages (“conlangs”), some were based (a little or a lot) on existing languages, and others were made up on the basis of particular features, ideas, and desires. Some were made up as part of a fictional world, while others were made for use in the real world. The ones that were made up for fiction will tend to have features that the author thinks of as appropriate for the kinds of characters who speak them, which might lead to different choices than if the author were making up a word for more general use. Ones that are made for real-world use tend to lean towards learnability and usability – and they often have a European bias.

Among the fictional languages, the two great early examples are the ones J.R.R. Tolkien invented for his Lord of the Rings trilogy: the elvish languages Quenya and Sindarin. He modeled Quenya generally on Finnish and Sindarin generally on Welsh, but that’s not to say there’s a discernible resemblance between his languages and the ones that inspired them. What we do see is a resemblance between the words in the two languages, and there’s no reason to call it mere coincidence: the Quenya word is wilarin and the Sindarin word is gwilwileth or gwiwileth. The w and land th are very characteristic of Tolkien’s elvish languages, and consequently now of ideas of mystical magical languages, and this has connected back to Welsh and ideas of Welsh. (Tolkien also created the rudiments of an evil language of Mordor, which he called the Black Speech, but it’s no surprise that it seems not have a word for ‘butterfly’.)

Perhaps the most popular fictional language, however, is Klingon, from the Star Trek TV and movie series. At first, there was no actual developed and schematized language, but in 1985 Marc Okrand was commissioned to create a well-developed and usable Klingon language for the movie Star Trek III: The Search for Spock, and his results – which were intentionally “alien” – were published in the wildly popular Klingon Dictionary. Why so popular? Klingons represent something that fans enjoy: a very un-butterflyish culture, a culture in which you can wish a person “May all your meals be served live.” But does their language have a word for ‘butterfly’? It does, and not a pretty one: Duj (D stands for /ɖ/, which is a retroflex d).

A very similar group are the Dothraki, from the TV series Game of Thrones, based on books by George R.R. Martin. That language, too, has a word for ‘butterfly’: zhalia. In Valyrian, a more high-toned and noble language from the same fictional world, the word is sōvion.

There are other fictional languages, too, of course, if not as well known. Suzette Haden Elgin, in her novel Native Tongue, set in a dystopian future in which women have had many of their rights taken away, features Láadan, a language created by and for women. Its word for ‘butterfly’, the etymology of which is described as a “visual/aural analog,” is áalaá. And the Indian epic adventure film Baahubali: The Beginning uses the language KiLiKi, created by Madhan Karky; its word for ‘butterfly’ is susutiri.

And what do all these fanciful words suggest to us about what the writers have observed, inferred, and desired about languages and their speakers, and the particular kinds of speakers they have in mind? And how much of what it’s suggesting to me or to you is just our own blue butterfly effects?

Then there are the conlangs intended for actual use, generally as auxiliary languages – easy-to-learn neutral languages that people of different countries can use to talk to each other, rather than learning (not necessarily well) each other’s languages. As they are meant to be easy to learn, their vocabularies are often intended to be as familiar as possible to the intended users. The three best known of these are Volapük, Esperanto, and Interlingua, and you can immediately see what community of speakers they particularly had in mind: their words for ‘butterfly’ are, respectively, pab, papilio, and papilion. Another one, Glosa, also uses papilio. And Sambahsa, which is based on Proto-Indo-European, has pelpel.

But there are others, too, and some of them are more pointedly neutral. Staren Fetcey, of Canada, developed Kotava specifically on the principle of cultural neutrality, with a vocabulary that does not give an advantage to any language or family of languages. And what is the Kotava word for ‘butterfly’? It’s bord.

And there is Mirad (also known as Unilingua), developed by Noubar Agopoff, of France. It is a very carefully constructed language; every letter maps to a semantic value. And, unsurprisingly, its vocabulary is not based on any existing language. Its word for ‘butterfly’, I learn, is gopelat. But its dictionary also gives another word, gipelat: it means ‘blue butterfly’.

Pinning down South American butterflies

South America has probably the greatest diversity of butterflies in the world. There are thousands of species, dozens of genera (that’s the plural of genus), and at least a half-dozen distinct families. Their diversity is aided by the same thing that challenges those who wish to study them: the geography of South America, which includes quite a lot of dense jungle and quite a lot of mountains. Isolation and lack of mobility can foster considerable differences over time. 

And yet, butterflies of entirely different families sometimes develop considerable similarities when they are in long-term close contact in the same environment – another sort of butterfly effect: two genetically unrelated butterflies might look nearly identical, while two genetically related ones might look quite different, just because of who’s been around whom in what environment.

South America also has – and, much more so, had – one of the greatest diversities of languages in the world. While the Niger-Congo region of Africa is hard to beat for sheer density of different languages, South America is a contender, and in some ways it’s even more diverse: while Africa has, depending on whom you ask, four or maybe five or six families of languages (though each has considerable internal diversity), and perhaps a few isolates (languages that don’t seem to be related to any other language), South America has, as far as linguists have been able to sort out, about 40 language families (including a dozen major ones) and more than 80 isolates. When Europeans first came to South America, there were at least 600 languages. Many of those are extinct now, and many others are endangered, but there are still hundreds of languages spoken there. 

And the diversity of South American languages is aided by the same thing that challenges those who wish to study them: the geography of South America, which includes quite a lot of dense jungle and quite a lot of mountains. Isolation and lack of mobility can foster considerable differences over time. But related languages don’t stay all in one place; over time the speakers move around – a map of language families of South America shows a pattern of language families as variegated as a butterfly wing.

And while in a real way language is culture, you can, in a given area, have people with similar culture but unrelated languages, and you can also have people who speak related languages but have quite distinct cultures, just because of who’s been around whom in what environment. And languages of entirely different families sometimes develop considerable similarities when they are in long-term close contact in the same environment.

But butterflies and languages – and language speakers – have some important differences. If you catch and kill some butterflies, you can study them after they’re dead and pinned down on a board, and you’re not going to be killing off the rest – at least not on purpose. By contrast, you can’t study a language if the speakers are dead – and when Europeans came and interacted with the indigenous peoples of South America, there was a lot of killing. And even if the speakers are not killed off, their languages may not survive; in fact, the same kinds of contact that bring people who study the languages (with an aim of recording and even preserving them) also bring commercial and cultural forces that are more than happy to dominate and suffocate languages that hinder them. 

Certainly, we can record the words, but dictionaries are the butterfly boards of languages. You see the specimens, inert, isolated, arranged neatly; you don’t see them in motion and transformation. Just as a butterfly on a board is clearly not a bat, but you don’t know whether it flies any different from a bat until you see it in motion, you really don’t know how words are used until you see and hear them in use. You may not even know something as simple as how the word changes when there’s more than one of the thing – one butterfly, a million butterflies.

On the other hand, the cultural contacts that can endanger languages can also create new languages through the contact and interaction of languages, as we’ve seen with Tok Pisin in Papua New Guinea. The area of the Caribbean has some notable creoles: for example, Sranan Tongo is a main language of Suriname; Papiamento is very common in Aruba, Bonaire, and Curaçao; Kreyòl ayisyen (Haitian Creole) is the main language of Haiti; and although English is officially the language of Jamaica, Patwah (Patois) is spoken by the majority of Jamaicans.

One fact coming out of all of this, however, is that while South America is as rich for linguists as it is for lepidopterists, although the armchair entomologist can learn about thousands of pretty bugs from the comfort of home, the armchair etymologist is, frankly, pretty bugged: online resources are limited (and in fact published resources in general are not as abundant as you might think), and those that have lists of words do not, in general, have any information about the origins and development of those words. 

Anyway, the truth is that if you study South American linguistics, you are more likely to be studying varieties of Spanish and Portuguese. And while there’s a lot of variation in Spanish varieties between different countries of Latin America, it’s similar to the variation in kinds of English – una mariposa es una mariposa. And while Brazilian Portuguese has various local varieties, and all of it can immediately be distinguished by ear from European Portuguese, no entanto, uma borboleta é uma borboleta.

But at least I can pin down a board of a few words for ‘butterfly’ in South American and Caribbean indigenous languages and creoles. It’s not much, but the languages – and their speakers – live on.

There is panambí, from Guarani, which is spoken by more than 6 million people in Paraguay and nearby areas – and is the most spoken language in Paraguay. Panambí is also the name of a municipality in Argentina (and just across the river from Paraguay) and a different municipality in southern Brazil.

There is pillpintu, from Quechua, which has up to 10 million speakers – mostly in the northern Andes countries – and which was the language of the Inca empire. (The ll is said similar to Italian gl or some versions of Spanish ll). 

There is pilpintu, from Aymara, which has nearly 2 million speakers in Bolivia and the adjoining areas of Peru, Argentina, and Chile. Seeing these Quechua and Aymara words, by now you should not be surprised when I tell you that the two languages are… probably not related in their historical origins (it’s “disputed”). Oh, the two words are related – one language got it from the other – but that’s a thing that can happen when languages are in long-term close contact with each other in the same environment. 

There is jampvzkeñ, from Mapudungun, which is spoken by about 200,000 people in Chile and Argentina. (The v stands for the vowel also written ü; the z is like English th; the j is like the Quechua ll.)

There is julirü, from Wayuu (also known as Guajiro), which is spoken by more than 400,000 people in Venezuela and Colombia. (The j is said as in Spanish, like English h.)

There is palanpalan, from the Galibi variety of Carib spoken by a few thousand people in the countries of the northern coastal area of South America. I can’t tell you whether this is also the word that was spoken in the Island Carib of the Caribbean Islands; that stopped being spoken a century ago. The Caribbean, like many places heavily settled by people of European extraction (including many suburban subdivisions), is named after something (in this case a people and a language) that was displaced by what it’s known for now (sugar plantations, resorts, et cetera).

I tried, but was not able, to find out the word for ‘butterfly’ in any of a number of other indigenous languages of South America and the Caribbean. But I found words in the creoles I mentioned above.

There is kaperka, from Sranan Tongo, spoken in Suriname by a half million people, including more than 100,000 for whom it’s their first language (Sranan Tongo means ‘Suriname Tongue’). Though Sranan Tongo has a substrate of English, a lot of its vocabulary comes from Dutch, and there are also words from Spanish, Portuguese, and some West African languages. I don’t know for sure where kaperka comes from, but it seems relevant that, while kapel is Dutch for ‘chapel’, in some varieties of Dutch it also means ‘butterfly’.

There is barbulet, from Papiamento, spoken by more than 300,000 people in Aruba, Bonaire, and Curaçao; although it is spoken in the Dutch Caribbean, Papiamento is a Portuguese-based creole (as you may have guessed from the resemblance to borboleta).

There is papiyon, from Haitian Creole, which is spoken by more than 12 million people in Haiti. (And yes, the source of the word is French papillon.)

And, finally, there is bat, from Jamaican Patois, spoken by more than 3 million people in Jamaica. Yes, according to my source, the word for ‘butterfly’ is bat. The word for ‘bat’, on the other hand, is rat-bat.

Next: Central and North America, fluttering from the tropics to the Arctic

Oceania: wandering butterflies, wandering people

The world’s second largest migration of butterflies takes place every year in Taiwan.

Given what I’ve said about the long distances that some butterflies travel, you may wonder “between Taiwan and where?” But I don’t mean the second longest distance – this migration covers only about 400 kilometres. I just mean the second largest number (after the monarch butterflies): millions of purple crow butterflies fly from the north of Taiwan to the south, and it can stop traffic – partly because they’re so beautiful and there are so many of them, and partly because some of them can end up on windshields. I suppose occasionally, on the other hand, some of them might fly into a car or truck and get a free lift. Perhaps they will wander from their flight path. But not all those who wander are lost.

Words for ‘butterfly’ also wander, with the languages that carry them. And it happens that one of the world’s longest-distance wanderings of languages started in Taiwan millennia ago: the Austronesian languages.

You may think of Taiwan as a Chinese-speaking country, and that’s true, but it’s true in about the same way as Australia is an English-speaking country. Chinese speakers arrived later and generally crowded out the indigenous languages – of which there were many. But by that time, the spread of languages across the islands to the south, west, and east had been in progress for a long time. By 6,000 years ago, the Austronesian languages had started spreading south to the Philippines. From there they moved, with the people who spoke them, to New Guinea, Borneo, and all of what is now Indonesia. And they kept going: west all the way across the Indian Ocean to Madagascar; east into Melanesia, Micronesia, and Polynesia in the Pacific Ocean. Now their spread goes halfway around the world – Madagascar is almost exactly on the opposite side of the globe from Hawai‘i.

They have subdivisions, of course. The deepest divisions are between the languages of Taiwan itself, which is one way that historical linguists know it all started there – the farther from origin languages of a family are, the less concentrated the diversity. Alas, the indigenous languages of Taiwan have few speakers and a long history of suppression, and consequently few resources, so I have been able to find a word for ‘butterfly’ in only one Taiwanese indigenous language: in Paiwan, it is kalidungudungul.

Does that have a familiar characteristic? Yes: you see the dungul twice. Reduplication, as we have seen, shows up in many other places with words for ‘butterfly’ – but in Austronesian languages, it shows up very frequently throughout the language. It is one of the signal characteristics of many Austronesian languages. 

Another characteristic that is not universal among them but very common is one the Paiwan word does not show: very limiting phonological rules. Austronesian languages typically have relatively few phonemes – Hawai‘ian famously uses only twelve letters (well, thirteen, counting the ‘okina, which is ‘ ) – and in many languages their syllables can have only one consonant followed by one vowel (or in some cases a diphthong) or just a vowel by itself.

You’ll see what I mean when we fly across the waters from island to island seeing their words for ‘butterfly’. In Filipino (Tagalog) it’s paruparo. In Balinese, Indonesian (aka Bahasa Indonesia), and Malay – which are all very closely related – it’s kupu-kupu, and in Javanese (from Java, where Indonesia’s capital, Jakarta, is) it’s kupu. In Malagasy – from Madagascar, much closer to Africa than to Indonesia, and yet – it’s lolo. 

It’s not that they all have these characteristics. When we look to Micronesian languages, we see Nauruan iyewaeoe and Palauan bangikoi. Meanwhile, Melanesian languages, which include the languages of Papua New Guinea, have the world’s greatest density of languages per square kilometre of land, but Melanesian languages are a geographic grouping, and the Papuan languages in particular include quite a few languages that don’t appear to be related to each other and aren’t part of the Austronesian language family. And I don’t have the resources to tell you what ‘butterfly’ is in any of the Papuan languages, even though some of them have more than a hundred thousand speakers.

But one well-known member of the Melanesian languages is definitely Austronesian, and is also culturally and geographically close to Polynesia: Fijian, in which ‘butterfly’ is bebe. And when we move into the Polynesian languages, which have the simplest syllables and the most reduplication, we see that in Samoan, Tuvaluan, Tahitian, and – I suspect – several others, ‘butterfly’ is pepe; in Māori, of New Zealand, it’s pēpepe, though Google gives me pūrerehua (and I have also gotten pūrehua as an alternative in Tahitian); and in Hawai‘ian, it’s pulelehua, though Google Translate gives me lelelele. (But I’ll trust my actual printed-and-bound dictionaries of Māori and Hawai‘ian over Google Translate.)

That is not much variety over quite a dispersed area, spread originally by boat across long distances with no land in sight for long stretches. And it took time – Samoa has had humans on it for perhaps 5,000 years; Tonga, 3,000 or so years; Tahiti, at least 2,500 years; Easter Island (Rapa Nui) and Hawai‘i, both about 1,000 years; New Zealand (Aotearoa), about 700 years. And over most of Polynesia, the first humans arrived with their Polynesian language.

And then, mostly in the last half millennium or less, other humans arrived with Indo-European languages. It’s like if a butterfly hopped into a car – or onto a boat, as the monarch butterfly did. We know that monarch butterflies are from the Americas, but there are lots of them in Australia, and have been since… the late 1800s. At first there weren’t so many, but then Europeans planted the kinds of plants that monarch butterflies like, and now there are lots, wandering (but not lost) all over the continent. Oh, by the way, monarch butterflies are called wanderer butterflies in Australia.

It’s not that Australia didn’t have butterflies before then. Of course it did! Australia, famous for its odd animals and big bugs, has the largest kind of butterfly in the world: the birdwings, which can have a wingspan of up to almost a foot (although the very biggest of them are actually found in New Guinea, not Australia). And the wanderers haven’t crowded out the indigenous butterflies in Australia. But the same can’t be said for people and languages. There were more than 250 Aboriginal languages spoken in Australia at the time Europeans first arrived. Now there are not much more than 100, and all but about 10 of them have very few remaining speakers. And I can’t tell you too much about them, alas, but I can tell you words for ‘butterfly’ in three of them, according to Glosbe.com: in Dhuwal, it’s buurnba; in Murrinh-Patha, it’s manman; in Warlpiri, it’s cinci-maɭu-maɭu.

And in most of Australia now, the word for ‘butterfly’ is… butterfly, of course, since English speakers are dominant throughout the country now. Indo-European languages have wandered and left their mark throughout the region. Consider the place names, so many of which are formed from our usual Latinized Greek linguistic Lego kit: Oceania, obviously related to ocean; Austronesian, from roots meaning ‘south island’; Micronesia, ‘small islands’; Melanesia, ‘black islands’ (as in dark-skinned people); Polynesia, ‘many islands’; Australia, ‘south place’; and then there are names like New Guinea, New Zealand, Philippines (named after Philip II of Spain), Marshall Islands, Gilbert Islands…

Wait, though: the Gilbert Islands are now called Kiribati. That’s good, right? It’s a Polynesian word? Well… it’s really the name Gilbert as adapted by the local language: gi becomes ki; l becomes ri; ber becomes ba; t becomes ti (actually pronounced “s” but that’s just the way t is said before i in that language, sort of like how ng in English is not the usual “n” sound plus “g”).

So, yes, as we have seen before, languages can take words that have wandered in and make them local. But the process can happen in a much more involved way, too. Local people and international traders develop business languages using reduced versions of the grammar and phonology of the local language along with modified words from the traders’ language, to create what is called a pidgin – in fact, the word pidgin comes from a modified version of the word business in one such. Pidgins are simplified languages, but once a pidgin becomes well developed and situated in a community, children start to grow up speaking it and it becomes a fully elaborated language, known as a creole. There are several of these throughout Oceania; the most widely known one is Tok Pisin (Tok Pisin is from the words talk and pidgin), a language of Papua New Guinea, with about a million people who speak it as their main language and a few million more who are conversant in it; as it spreads it’s increasingly dominating other languages of Papua New Guinea. 

And what is the Tok Pisin word for ‘butterfly’? It’s bataplai.

Next: panambi, pillpintu, and more.

East and Southeast Asia: the lepidopteran effect

The thing about butterfly effects is that they’re chaotic – the butterfly effect is an idea from chaos theory, after all. What that means is that sometimes it carries through to where you’d think it wouldn’t, and sometimes it doesn’t carry through to where you’d think it would. And, because it doesn’t generally give pictures of where it came from, sometimes – most of the time, in reality – you’re not entirely sure. You can have similar things happening at great distance and have no reason to think they’re connected – right? And you can have very different things happening close together that are connected …maybe.

We can make our best guesses, of course, and rate things as more or less likely. So just because, for instance, butterflies are Lepidoptera and that has given us words such as Croatian leptir, it does not mean that some butterfly word halfway around the world that also has lep- is related. As linguistics like to say, etymology by sound is not sound etymology! Yes, of course similar-sounding words may be related, and similarity of sound can be a lead to follow, but you can’t conclude anything on the basis of sound resemblance alone – for instance, isle and island are unrelated, and the s in island got there by mistake. And on the other hand, as we have already seen, words can sound quite different and yet still be related.

So when we look at the Proto-Sino-Tibetan *lep root, which meant ‘butterfly’, we have no good reason to think it’s related to lepidoptera, nor, for that matter, to flap or any other similar-sounding word from the farther west world. There are various theories on exactly where it comes from. Could be a root meaning ‘thin, flat’; could be one meaning ‘flash; lightning; glitter’; could be ‘flapping’. If you’re interested in more details, Wiktionary is a place to start. But we do know that it descended into Burmese လိပ်ပြာ (/leɪʔpjà/, often rendered as lippra) and to Tibetan ཕྱེ་མ་ལེབ (phye ma leb), with some added bits. 

And we know that it added a prefix in Old Chinese to be *ɡaː l’eːb, which became Middle Chinese *ɦuo dep, which became modern Cantonese wu dip, Hakka fu tiab, Wu hhu diq, words in other varieties of Chinese that look more or less like one or more of the preceding, and modern Mandarin húdié. And all of the modern Chinese versions are written the same: 蝴蝶.

Let’s take quick look at those characters, because who doesn’t like to take apart Chinese characters to understand how they work? After all, they literally give you pictures of where they came from. If you’re wondering what the meaning decomposes to, since it’s two characters, the answer is that both of them mean ‘butterfly’ – a redundancy similar to the word pussycat, say, but they’re always used together. And the first thing you can see is that both of them, 蝴 and 蝶, have the same bit on the left: 虫. It’s the semantic component of each, and the right side is the phonetic component (Chinese compound characters are normally composed of semantic and phonetic parts – so they indicate ‘thing with this kind of meaning and this kind of sound’). Want to guess what it means? The answer is ‘insect’ – or ‘worm, snake, or other creepy crawly thing’. The original character looked more like a slithering snake, but it’s changed much over the ages.

As to the phonetic components, the hú, on the left, has 胡, which when used by itself has a lot of meanings, including ‘barbarian’ and ‘recklessly’, and – you may have noticed – it is itself made from two bits: the one on the left, 古, which by itself means ‘old, ancient’, is made further from the character for ‘ten’ (十, but it might originally have been something else) on top of the character for ‘mouth’ (口), and the one on the right looks like the character for moon (月) but is actually a simplified form of the character 肉, which is from something like a side of beef (hmm… maybe a tauriņš? Nah). Even the evidence before your eyes can mislead!

Meanwhile, the dié, on the right, has 枼, which by itself means ‘table, slat, slip, leaf’ and is also made of two parts: 世, made from three of the character for ‘ten’, and meaning ‘thirty’ as in years and thus ‘generation’, and 木, meaning ‘wood’ and looking originally – and still – like a tree. Each year’s worth of leaves are a generation for a tree, you see. Not that that has any reference to the generations of butterflies it takes to complete a migration.

That second half of the Chinese butterfly, 蝶, is also the character used in Japanese for ‘butterfly’. That’s not surprising, since Japanese uses a lot of characters of Chinese origin – although it also uses a lot of characters you won’t see in Chinese. And the Japanese language is historically unrelated to Chinese; grammatically and phonetically it is strikingly different, and, though it has many Chinese influences, its core of vocabulary is also altogether different, even when it uses characters from Chinese to represent the words. 

Thus, in Japanese you can see a Chinese character with a pronunciation (sometimes more than one available pronunciation for the same thing) entirely different from any Chinese pronunciation. The way 蝶 is pronounced in modern Japanese (and rendered in Roman), for example is chō. That traces, historically, to teu, which came from tewu, from tefu, from tepu, from… Middle Chinese dep. Yes, Mandarin dié and Japanese chō not only use the same character, they come from the same original pronunciation. Don’t ask me why Japanese took its word for ‘butterfly’ from Chinese – it took quite a few things from Chinese, but far from everything!

We can’t talk about Chinese and Japanese without talking about Korean, though it’s not related to either of them (unless so far back that it’s all arrant speculation). The Korean word for ‘butterfly’ is 나비 (nabi), which might – might – originally have meant ‘flutterer’.

We also can’t move away from East Asia without looking at the other language families spoken in the area. One of them is the Austroasiatic family, the best-known members of which are Khmer (Cambodian) and Vietnamese, though it has a large number of other members spoken as far south as Malaysia and as far west as India. The Khmer word for ‘butterfly’ is មេអំបៅ (mei ʼɑmbaw). In Vietnamese, the word for ‘butterfly’ is bướm – but sometimes they call it bướm bướm, and some related languages also reduplicate the word: Wiktionary lists as cognates Thổ pəmpɨam, Hoà Bình Muong pɨəm pɨəm, Sơn La Muong pɨəm, Thanh Hoá Muong bɨəm, and Muong Bi pươm pưởm. All of those seem to mean just ‘butterfly’ and not come from anything else – though of course other things may be named subsequently after butterflies, and if you’re curious about what kinds of other things, you can always look at, for instance, the Wiktionary entry.

But when we turn to the Kra-Dai family of languages, which include Lao, Thai, and a number of other languages, including some spoken in inland southern China, we find a more poetic approach. In both Lao ຜີເສື້ອ (phi seu) and Thai ผีเสื้อ (pǐi-sʉ̂ʉa), the name of the insect means literally ‘tutelary spirit’. Tutelary? That means ‘guardian’ – as in your guardian angel, or, in this case, perhaps an ancestor who is there to guide you. You may remember Persian parvâne, from a root meaning ‘guide’. Of course it’s not related – for one thing, in Lao and Thai the word order is ‘spirit guide’ and so you can forget about connecting the p’s. But why wouldn’t people in different parts of the world look at the same airborne miracles and get similar ideas about them?

Mind you, I’m sure that they had different ideas of who the guide might be. But, as I’m thinking about guides and butterfly coincidences, I must also mention the Mongolic languages, which include Buryat, Kalmyk, Santa (yes, really – but it’s also called Dongxiang), and of course Mongolian. Their words – as far as I can see – are all pretty similar: Buryat эрбээхэй, Kalmyk эрвәкә, Mongolian эрвээхэй (apologies for not including the Mongolian-alphabet versions, but I can’t render them in the proper vertical orientation here). In case you don’t read Cyrillic letters, they’re pronounced erbeehey, ervəkə, and erveehey (the ee is like “ehh” not “eee”). And they all trace to Proto-Mongolic *herbekei. Which, I assure you, has an only purely coincidental resemblance to Harbeck, though I am serving as your butterfly guide here.

Next: from island to island across the oceans.

Middle Eastern and African butterflies: keep on flapping

As we flew eastward into Asia, we bypassed Africa and much of the Middle East, so we’re going to double back now. I wanted to treat all the language families of Africa together, and one of the language families – Afro-Asiatic – is also dominant in the Middle East – particularly two of its members: Arabic and Hebrew.

You may not think of Arabic and Hebrew as African languages, but it’s not just that Arabic is spoken (in several varieties) across northern Africa; it’s that all of the other languages of the Afro-Asiatic family are found in Africa: Afar, Amharic, Berber, Hausa, Oromo, Somali, Tigrinya, and more. The languages are not all closely related, but you’ve already seen how much diversity there can be in a language family (a.k.a. phylum, such as Indo-European). And, as you might expect, the words for ‘butterfly’ in the different Afro-Asiatic languages are diverse: Amharic ቢራቢሮ (bīrabīro); Berber aferṭeṭṭu; Hausa malan boɗe ido; Oromo billaacha; Somali balanbaalis; Tigrinya ጽንብላሊዕ (ṣəmbəlaliʿ); Afar laahinti (which means ‘cow’s eye’).

For that matter, Arabic itself is diverse; it has quite a few varieties, just as English has. But the standard Arabic word for ‘butterfly’ is فراشة (farāša). I don’t know its etymology, alas, but I do notice that the word in Hebrew – which is closely enough related to Arabic that you can often notice similarities between the two – is פַּרְפַּר or, without the vowel marks, פרפר (parpar), which has par where Arabic has far.

And what is that par? Well, after we’ve seen papilio, farfalla, pilipala, pilipilinpauxa, peperutka, perhonen, and parvâne, it seems just about par for the course – both its form and its reduplication. I am told by a regular reader of Sesquiotica, Israel (Izzy) Cohen, that in Hebrew you can use parpar as a verb – specifically, l’parper – to mean ‘be indecisive’, i.e., ‘flounder, flip-flop’. And he says that the butterfly is called parpar in Hebrew “because of its spasmodic flight-pattern.” So it is, I think, imitative – like some other words for ‘butterfly’ we have seen.

Imitative? Well, you see those flat wings flapping up and down, and you can picture in your head how they must sound when you can hear them (which you seldom can). And of course it’s not just one flap; it keeps on flapping – ergo the reduplication.

I keep mentioning reduplication (repetition of part or all of a word, either exactly the same or in modified form), and that’s because it keeps showing up. Worldwide, languages use reduplication in varying amounts for an assortment of reasons, as we sometimes do – not always so formally – in English. Some use it for pluralization (“Cheeseburger! cheeseburger! cheeseburger!”). Some use it for emphasis or clarification (“a drink drink”). Some use it for an endearing diminutive or to express fond familiarity (“Night-night, choo-choo”). Some use it to indicate repetition (“tap-tap-tap”; “pitter-patter”). 

This last kind is quite common in many African languages – specifically, it is used as a frequentative form, which means it means something keeps happening. For example, in Swahili, the reduplicating verb root -pepea means ‘wave back and forth’. And when you add the right noun class prefix and the right suffix to it, you get kipepeo, which means a thing that waves back and forth… specifically, a butterfly.

Which is fun because, though kipepeo seems like it would be related to parpar, Swahili doesn’t even belong to the same language family as Hebrew. But of course languages in the same area can influence each other: for example, the Arabic word for ‘book’ is كتاب (kitab, pronunciation varying between kinds of Arabic) and the Swahili word for ‘book’ is kitabu – note that same ki- noun class prefix as on kipepeo: the Arabic word was reanalyzed as root -tabu plus noun class prefix ki-.

What language family does Swahili belong to? There are ongoing terminological discussions, but the established name (and the one I learned when studying African languages in university) is Niger-Congo, because they all seem to trace from the area around the Niger and Congo rivers, at the angle of the African west coast. They have spread widely from there, throughout sub-Saharan Africa all the way to the east coast and the southern tip, and there are a lot of them. Not only do I not have the resources to look up ‘butterfly’ in all of them, I don’t even have the resources to name all of them (nor the space here either). But I do have more than a dozen I have at least been able to find in online resources – though I can’t always vouch for the reliability of the resources (mutters something about the uneven distribution of resources and attention between colonizers and colonized).

The first language I’ll name, Bambara, spoken in Mali, may not even be Niger-Congo; the Mandé languages of the northwest of Africa are quite different from the rest of the Niger-Congo languages. But I cannot not tell you that the Bambara word for ‘butterfly’, according to Google Translate, is nfirinfirinin, which seems to me like a whole tree full of butterflies all taking to wing at once. On the other hand, Glosbe.com gives me two words: dimago and npɛrɛnpɛrɛ. The first seems dull, but the second is glorious to behold. Fula, a not-too-closely related language spoken throughout west Africa, uses bedelallah, and I would love to know if there’s an Arabic influence, but I don’t. Wolof, related to Fula and spoken in Senegal, Gambia, and Mauretania, uses lëp-lëp, and I have a guess that it might be formed in about the same way as the Yoruba word.

Yoruba is definitely a Niger-Congo language, spoken in Togo, Benin, and Nigeria, and it uses a feature common in languages of the family (and many others around the world): ideophones. The Yoruba word for ‘butterfly’ is labalaba, from an ideophone indicating ‘light and floppy’. What’s an ideophone? A sort of demonstrative, expressive interjection; we use them some in English – “Bam, the lights come on, and whoosh, it’s like someone sucked all the air out of the room.” They’re like onomatopoeia but don’t necessarily imitate actual sounds. Think of the kinds of standardized interjections you often accompany with hand gestures. 

For instance, in Ewe, spoken in Ghana, Togo, and Benin, lilililili on a high tone means “nice good sweet smell” and on a low tone means “very bad smell.” I don’t know whether ideophones affected the Ewe word Google Translate gives me for ‘butterfly’, akpakpaluʋui, or either of the words Glosbe.com gives me, bakboloowhen and dyekpakpa. But I can see kpakpa in two of those words, and I can tell you that kpakpaxe means ‘duck’. (By the way, the syllables are kpa-kpa; you say the “kp” with the back of the tongue and the lips at the same time. These coarticulated stops are a common feature in Niger-Congo languages.)

I see something resemblant (though without the coarticulations) in lipekapeka, a word for ‘butterfly’ from Lingala, which is spoken in the region of the Congo river and nearby countries, but I don’t know whether it’s related. In Igbo, which is spoken in Nigeria, I find the word is urukurubụba. On the other hand, Twi, spoken in Ghana, has a soft and floppy word: afafrantɔ (by the way, that ɔ represents the vowel sound English speakers make in “court”).

As we follow the sweep of Niger-Congo languages across Africa to the east coast and then southward, we find the word in Kikuyu (spoken in Kenya) is kĩĩhuruta and the word in Luganda (spoken in Uganda) is ekiwojjolo. But in others we come back to the common theme of republication: in Kinyarwanda (spoken in Rwanda), it’s ikinyugunyugu; in Chichewa (also called Nyanja, spoken in Malawi, Mozambique, and Zambia), it’s gulugufe; in Shona (spoken in Zimbabwe), it’s shavishavi; in Sesotho (spoken in Lesotho), it’s serurubele.

The Niger-Congo languages – and their speakers – made it all the way to South Africa, arriving within about the last five hundred years; the two most spoken languages in South Africa, Zulu and Xhosa, are both Niger-Congo languages… but with quite a lot of influence from languages that were already spoken in the area when they arrived. (The third most spoken is Afrikaans, a creolized language based on Dutch; its word for ‘butterfly’ is skoenlapper, which means ‘shoe-patcher’.) Zulu and Xhosa so famous for their “click” sounds that it may surprise you to know they’re not actually part of the famously click-using Khoisan family of languages, spoken by mainly desert-dwelling people of the region from time immemorial. And it should surprise you; it’s unusual for a language to pick up a striking and difficult sound from an unrelated language. But this is a thing that can happen when speakers of different languages get together and meld families and customs; indeed, research suggests that the clicks started appearing in Zulu and Xhosa specifically as a way of getting around taboos against saying names of certain relatives and animals (especially ones you’re afraid of).

Naturally, as I’ve said so much about clicks in these languages, you can only be expecting that their words for ‘butterfly’… have no clicks in them at all. Sorry. The Xhosa word is ibhabhathane; the Zulu word is uvemvane (though there is also ijubajubane – which is apparently related to ijuba, ‘pigeon’ – and itwabitwabi). Well, heck, butterflies don’t click, do they? And no one’s afraid of them. But the words do reduplicate!

But what about the Khoisan languages, which have a much higher density of click sounds? Alas, they have a much lower density of lexical resources available. Not Khoekhoe, not !Kung, certainly not any of the ones with very few speakers, and for that matter not Sandawe – spoken farther north, in Tanzania – not one of them is well enough covered in online resources that I can find its word for ‘butterfly’, though I’m sure they all have one, because butterflies live where they live.

But I have not quite covered all of the language families of Africa. There is one more, the Nilo-Saharan, a family of languages of the central-northeast part of Africa, including Dinka and Maasai; more than a dozen of the languages have over a million speakers. To say they are not all closely related is an understatement, however; in fact, many scholars view Nilo-Saharan not as an actual family but simply as a catch-all for languages not part of the other three families. 

But they’re still in Africa, and at least some of them also use reduplication. The Kanuri word for ‘butterfly’ it’s fátáfátámà (why all the accents, by the way? Kanuri, like many languages in Africa, including most Niger-Congo languages is a tone language – in fact, the majority of the world’s tone languages are in Africa, but their use of tone is somewhat different than the use of tone in East Asian languages). In Maasai, it’s ɔsámpúrimpúri. In Lugbara, it’s alapapa. And in Dinka? In Dinka, at least when referring to a small butterfly, I’m told it’s dap.

Next: lepidoptera of East Asia.