languagehat.com

John Hooper’s Guardian obituary of the late-blooming writer Andrea Camilleri discusses the very interesting linguistic elements of his popular novels featuring the Sicilian detective Salvo Montalbano (or, per Sicilian Wikipedia, Salvu Muntalbanu):

In one sense, the Montalbano novels were not at all innovative: Camilleri named his hero after the Spanish author Manuel Vázquez Montalbán, and admitted he had given him some of the traits of Montalbán’s gourmet investigator, Pepe Carvalho. Moreover, Camilleri churned out the exploits of his most popular character in a way that was decidedly more industrial than creative. “All the Montalbano novels are made up of 180 pages, tallied on my computer [and] divided into 18 chapters of 10 pages each,” he once told an interviewer.

But in an important respect, the Montalbano stories were utterly original. What is not apparent to readers of the stories in translation or to the many non-Italian fans of the television series that sprang from them is that they are written in a language of the author’s creation: a blend of standard Italian with Sicilian dialect.

In La Lingua Batte Dove il Dente Duole (2013, literally Where the Tongue Touches the Toothache), a book-length interview with the linguist Tullio De Mauro, Camilleri explained that the idea arose from the circumstances of his father’s death in the late 1970s and inspired him to try out the technique, unsuccessfully, long before the first Montalbano book appeared.

“One day, to distract him, I said: ‘You know, Dad. I’ve thought of a story,’ and I told him the plot of my first novel … My father goes: ‘Why don’t you write it?’” Camilleri replied that he found it difficult to write in Italian, to which his father replied: “And why do you have to write it in Italian?”

To publishers, Camilleri’s linguistic mish-mash, which even non-Sicilian Italians have difficulty in understanding at first, must have seemed like a refined form of literary suicide. The author was no stranger to rejection slips. But over the course of his much-delayed career Camilleri sold more than 10m books. They were translated into more than 30 languages and adapted for a hugely successful television series that has been sold to more than 20 countries. It was Montalbano’s success on screen that turned Porto Empedocle, the model for his beat, Vigata, into a holiday destination for his many fans. So proud was the town of its most famous son’s literary creation that from 2003 to 2009 it called itself Porto Empedocle Vigata.

I have to correct Hooper’s translation of La Lingua Batte Dove il Dente Duole; it’s The Tongue Hits Where the Tooth Hurts, not “Where the Tongue Touches the Toothache.” Thanks, Trevor!

You may well have seen a story about the recent discovery in London of a Roman stylus with a long and amusing inscription, but if so you probably wondered, as I did, what the Latin said — it’s hard to make out from the published images. Fortunately, I was able to find a MOLA story that provides it:

A unique Roman stylus, with the most elaborate and expressive inscription of its kind is set to go on display for the first time in a new exhibition at the Ashmolean: Last Supper in Pompeii. It was discovered by MOLA archaeologists during excavations for financial technology and information company Bloomberg’s European headquarters in London, on the bank of the river Walbrook – a now lost tributary of the Thames. The iron stylus – used to write on wax-filled wooden writing tablets – dates to around AD 70, just a few decades after Roman London was founded. […]

The inscription has been painstakingly examined and translated by classicist and epigrapher Dr Roger Tomlin. It reads:

‘ab urbe v[e]n[i] munus tibi gratum adf(e)ro
acul[eat]um ut habe[a]s memor[ia]m nostra(m)
rogo si fortuna dar[e]t quo possem
largius ut longa via ceu sacculus est (v)acuus’

‘I have come from the City. I bring you a welcome gift
with a sharp point that you may remember me.
I ask, if fortune allowed, that I might be able (to give)
as generously as the way is long (and) as my purse is empty.’

In other words: the stylus is a gift to remind the recipient of its sender; the sender acknowledges that it is a cheap gift and wishes that they could have given more. Its tongue-in-cheek sentiment is reminiscent of the kinds of novelty souvenirs we still give today. It is the Roman equivalent of ‘I went to Rome and all I got you was this pen’, providing a touching personal insight into the humour of someone who lived nearly 2000 years ago.

So there you have it. And while I have your attention, I have a problem with a phrase my wife asked me about in the David Magarshack translation of Anna Karenina she’s reading. It’s in Part 5, chapter 21 (emphasis added): “He [Karenin] proposed bed and lavished on his fiancee and his wife all the feeling of which he was capable.” She said it seemed unlikely that the strait-laced Karenin would have directly proposed sex, and I agreed; I dashed off to see what Tolstoy actually wrote, which was “Он сделал предложение и отдал невесте и жене все то чувство, на которое был способен”: ‘He made an offer and gave his fiancee and wife all the feeling of which he was capable.’ In this context, “made an offer” is equivalent to “proposed marriage”; what I am wondering is whether there is an archaic idiom “propose bed” in this sense (the OED doesn’t mention it), or whether Magarshack utterly misunderstood the Russian (which seems unlikely).

It occurred to me to wonder where the word domino was from, and what was the earliest sense; per the OED (entry from 1897), the first meaning was “A kind of loose cloak, apparently of Venetian origin, chiefly worn at masquerades, with a small mask covering the upper part of the face, by persons not personating a character,” attested from 1719 (Free-thinker No. 138. 2 Thersites..instead of covering Himself with a Domine, dresses..in the Habit of a Running-Foot-man), “Sometimes applied to the half-mask itself” (1860 R. W. Emerson Illusions in Conduct of Life 276 The masquerade is at its height. Nobody drops his domino). Sense 2, “A person wearing a domino,” is attested from 1749 (H. Fielding Tom Jones V. xiii. vii. 56 Jones..applied to the Domino, begging and intreating her to shew him the Lady), and the modern sense 3, “One of a number of rectangular pieces (usually 28) of ivory, bone, or wood, having the under side black, and the upper equally divided by a cross line into two squares, each either blank or marked with pips, so as to present all the possible combinations from double blank to double six” or “A game played with these pieces,” only from 1801 (J. Strutt Glig-gamena Angel-ðeod iv. ii. §18 Domino..a very childish sport, imported from France a few years back). As for the etymology, the antiquated OED one is:

< French domino (16th cent. in Hatzfeld & Darmesteter) ‘a kind of hood, or habit for the head, worne by Cannons; (and hence) also, a fashion of vaile vsed by some women that mourne’ (Cotgrave): compare Spanish domino a masquerade garment.
Du Cange cites domino in Latin context, in the sense of a covering of the head and shoulders worn by priests in winter: ‘utantur..caputio vulgariter ung Domino’, ‘caputium seu Domino panni nigri’. Derived in some way from Latin dominus; Darmesteter suggests from some Latin phrase, such as benedicamus Domino. According to Littré, sense 3 came from the supposed resemblance of the black back of each of the pieces to the masquerade garment.

The Online Etymology Dictionary has:

1801, “one of the pieces with which the game of dominoes is played,” from French domino (1771), perhaps (on comparison of the black tiles of the game) from the meaning “hood with a cloak worn by canons or priests over other vestments in cold weather” (1690s), from Latin dominus “lord, master” (from domus “house,” from PIE root *dem- “house, household”), but the connection is not clear.

Klein thinks the game name might be directly from dominus, “because he who has first disposed his pieces becomes ‘the master.’ ”

The Klein suggestion is just silly, but other than that I have nothing to add to this murky subject. If you’re curious about “J. Strutt Glig-gamena Angel-ðeod,” see this 2016 post.

Fintan O’Toole’s wonderfully titled “The Ham of Fate,” in the new NYRB, is about a failed newspaperman who has been forced to earn his crust of bread in other ways; at one point in his checkered career, he wrote a novel, from which this is quoted:

There was something prurient about the way he wanted to read about his own destruction, just as there was something weird about the way he had been impelled down the course he had followed. Maybe he wasn’t a genuine akratic. Maybe it would be more accurate to say he had a thanatos urge.

O’Toole says:

The Greek terms stand out. In part, they function as signifiers of social class within a long-established code of linguistic manners: a sprinkling of classical phrases marks one out as a product of an elite private school […] and therefore a proper toff. […] The choice of thanatos is interesting […]. But it is akratic that intrigues. […] Akrasia, which is discussed in depth by Socrates, Plato, and especially Aristotle in the Nicomachean Ethics, […] means literally “not being in command of oneself” and is translated variously as “weakness of will,” “incontinence,” and “loss of self-control.” To Aristotle, an akratic is a person who knows the right thing to do but can’t help doing the opposite.

I looked it up in the OED, whose entry (from December 2011) reads:

akratic, adj.2 and n.
[…]
Etymology: < ancient Greek ἀκρατής powerless (see akrasia n.) + -ic suffix.
Chiefly Philosophy.
A. adj.2
Exhibiting or characterized by lack of restraint or weakness of will. Also: characterized by the tendency to act against one’s better judgement.

1896 Free Rev. Apr. 86 It does not concern us..to know whether there be antagonism between the socialisation of the means of production and the acratic form of society [i.e. anarchy].
1913 H. Scheffauer tr. R. Mayreder Surv. Woman Probl. 262 Carried to its extremes, this acratic tendency produces licentious domineering masculinity and weak, insignificant and passive, or else crafty, false and ludicrous femininity.
[…]
1980 A. O. Rorty in Social Sci. Information 19 908 If philosophers who deny acrasia are self-deceptive and akratic, so are those who deny the integrative functions of the varieties of rational strategies.
2009 P. Poellner in K. Gemes & S. May Nietszche on Freedom & Autonomy viii. 156 An addict or acratic person—a slave of momentary affect and desire.

B. n.
An akratic person.

1913 H. Scheffauer tr. R. Mayreder Surv. Woman Probl. 262 The commonest type is the acratic, the partially developed being of unmitigated sexuality.
[…]
2010 D. Charles in J. Cottingham & P. Hacker Mind, Method, & Mortality iii. 51 There are, it seems, two different types of impetuous acratics under consideration in these passages.

A useful word for a common human phenomenon; too bad it’s so ostentatiously posh. It should be given a sparrowgrass-style makeover as, say, craddock. “Why am I doing this dumb thing? It’s the craddock in me!”

David Robson reported recently for BBC WorkLife about some interesting studies:

Cecilia Pemberton at the University of South Australia studied the voices of two groups of Australian women aged 18–25 years. The researchers compared archival recordings of women talking in 1945 with more recent recordings taken in the early 1990s. The team found that the “fundamental frequency” had dropped by 23 Hz over five decades – from an average of 229 Hz (roughly an A# below middle C) to 206 Hz (roughly a G#). That’s a significant, audible difference.

The researchers had carefully selected their samples to control for any potential demographic factors: the women were all university students and none of them smoked. The team also considered the fact that members of the more recent group from the 1990s were using the contraceptive pill, which could have led to hormonal changes that could have altered the vocal chords. Yet the drop in pitch remained even when the team excluded those women from their sample. Instead, the researchers speculated that the transformation reflects the rise of women to more prominent roles in society, leading them to adopt a deeper tone to project authority and dominance in the workplace. […]

In one experiment, Joey Cheng of the University of Illinois at Urbana-Champaign asked groups of four to seven participants to perform an unusual decision-making task that involved ranking the items that an astronaut would need to survive a disaster on the moon. And at the end, she also asked each member to (privately) describe the pecking order of the group and to rank each member’s dominance. Recording the participants’ discussions throughout the task, she found that most people quickly shifted the pitch of their voice within the first few minutes of the conversation, changes that predicted their later ranking within the group.

For both men and women, the people who had lowered their pitch ended up with a higher social rank, and were considered to be more dominant in the group, while the people who had raised their pitch were considered to be more submissive and had a lower social rank. “You were able to predict what happened to the group, in terms of the hierarchies, just from these initial moments,” Cheng says.

Suggestive (if unsurprising) results, though of course one would want more confirmation. Thanks, Kobi!

The other day I heard a news story about the garment called in Spanish faja (more or less ‘corset’), and of course I wondered about the etymology. It turned out to be from Latin fascia ‘band, bandage, swathe, strip, ribbon,’ but not directly — it’s borrowed from Aragonese faxa. The inherited form is haza ‘small field, plot of arable land’; there’s also a scientific term fascia ‘layer of loose tissue’ borrowed directly from Latin, making three doublets. And if you go to the Wiktionary fascia link and scroll down, you’ll see a whole list of descendants, from Aromanian fashi ‘bandage, dressing; swaddling clothes’ to Greek φασκιά ‘swaddling clothes’; a particularly interesting entry is Old French faisse, fece, which gives French fasce (re-Latinized), borrowed into English as fess(e) “An ordinary formed by two horizontal lines drawn across the middle of the field, and usually containing between them one third of the escutcheon” (per the OED). It would have thrilled me as a lad getting interested in language to have instantly available such comprehensive displays of historically related forms, and I don’t suppose I’ll ever be blasé about it.

Lucas Petersen’s Irish Times article “José Salas Subirat, the eccentric first translator of Joyce’s Ulysses into Spanish” tells a remarkable story:

Although both James Joyce and his editor Sylvia Beach included Spain from the first moment in their international promotion strategy for Ulysses, and the book had acquired considerable fame throughout the literary world in the West, Spanish-language readers had to wait 23 years, until 1945, to read Joyce’s magnum opus in their own language.

They had to wait for a humble insurance salesman with an erratic literary career, with practically no background in translation, and with a knowledge of English that was centainly below what could be expected for such a task: José Salas Subirat faced this titanic challenge all alone and out of love for the task itself.

The translator’s exploits became one of the most talked-about stories among Latin American Joyceans and, of course, one of the landmarks in Spanish language translation history — even Gabriel García Márquez remembered the day he met Salas Subirat and wrote an almost preposterous portrait of him. […] Negatively criticized when it first appeared, as years went by Salas Subirat’s translation won more and more fans. Among them, some contemporary writers who now have their place in the Argentine literary canon, such as Juan José Saer and Ricardo Piglia. At the beginning of this century, after two new translations in Spain, Salas Subirat’s work was brillantly defined by Carlos Gamerro, who taught courses about Ulysses for many years, as the one that has more mistakes but also the one that has more good choices. […]

“The original Ulysses,” wrote Gamerro, “is not written in one language or dialect, but in the tension between a discredited variant (Irish English) and a dominant one (imperial British English): a relation that could be compared, even if it is not equivalent, with the one that exists between the Spanish of Spain and the Spanish of the other Spanish-speaking countries.” For Gamerro, Salas Subirat’s Ulysses “reproduces, with all of its imperfections, that strain that’s in the original work. Hesitant, multilingual, scrambled: that’s the friction that fires Ulysses’ English, and makes our creole Ulysses have a similar vitality.” […]

Who was Salas Subirat? How did he become one of Argentina’s most famous translators? These are the questions that I will try to answer here.

I’ll send you to the link for the answers; here I’ll just mention that Salas Subirat, the son of Catalan immigrants, among his many other jobs worked as a translator for Yuzhamtorg (the Soviet company that traded with South American countries in the 1920s). And when Borges said of his translation “It is really bad,” someone responded: “It might be, but if it is, Mr Salas Subirat is the greatest writer in the Spanish language.” Thanks, Trevor!

I recently heard a discussion of stents on the radio and of course wondered about the etymology; it sounded Latin, but I couldn’t go beyond that. So I looked it up and discovered it’s (in the words of the OED, in a 1916 entry) “< the name of Charles T. Stent (1807–85), English dentist.” How about that!

And yet it’s not that simple. Wikipedia says:

As Ariel Roguin describes in his paper “Stent: The Man and Word Behind the Coronary Metal Prosthesis”, the current acceptable origin of the word stent is that it derives from the name of a dentist, Charles Thomas Stent, notable for his advances in the field of denture-making. He was born in Brighton, England, on October 17, 1807, was a dentist in London, and is most famous for improving and modifying the denture base of the gutta-percha, creating the Stent’s compounding that made it practical as a material for dental impressions.

The verb form “stenting” was used for centuries to describe the process of stiffening garments (a usage long obsolete, per the Oxford English Dictionary) and some believe this to be the origin. According to the Merriam Webster Third New International Dictionary, the noun evolved from the Middle English verb stenten, shortened from extenten, meaning to stretch, which in turn came from Latin extentus, past participle of extendere, to stretch out.

So maybe it’s Latin after all? In any case, the OED’s first citation for the original sense (“a substance invented by [Stent] for taking dental impressions; (also) an impression or cast of a part or body cavity made of this or a similar substance, and used to maintain pressure on it so as to promote healing”) is:

1878 C. Hunter Mech. Dentistry i. 2 Wax as an impression material is now seldom used, composition (Godiva, or Stent) or plaster of Paris being now almost invariably employed.

For the newer (“A tube implanted temporarily in a vessel or part”), it’s:

1964 Jrnl. Prosthetic Dentistry 14 1168 All stents must be removed daily and cleaned. A pipestem cleaner is effective in cleaning the tube.

I was reading Marina Warner’s NYRB review (subscriber-only) [archived] of Stephen Greenblatt’s The Rise and Fall of Adam and Eve when I was startled by an implausible-looking equation. But before I get to that, I’ll quote the amusing opening of the review:

In 1872, when the brilliant young Assyriologist George Smith found a cuneiform tablet in the British Museum inscribed with part of the story of the Flood, he became so excited that he began undressing, though the comparative literature scholar David Damrosch thinks that he might have been merely loosening his collar, Stephen Greenblatt tells us—still sign enough to alarm Smith’s Victorian confreres into fearing that he was overborne with passion.

OK, so later on Warner parenthetically notes that “The Orientalist Stephanie Dalley has argued that the name of the hero Buluqiya in a long quest tale in the Arabian Nights derives from Gilgamesh.” That made me sit up and take notice, since the two names seem very different; fortunately, JSTOR provided me with Dalley’s article “Gilgamesh in the Arabian Nights” (Journal of the Royal Asiatic Society Third Series, 1.1 [Apr. 1991]: 1-17), where we find the following explanation:

The two personal names Buluqiya and al-Khidhr can be connected with extreme antiquity. Buluqiya is not an Arabic name, nor is it a name for a king of Israel, even though Tha’labi’s version calls him son of Josiah. The name can be explained as a hypocoristic of Gilgamesh’s name in a pronunciation attested both in Sumerian and in Hurrian: Bilgamesh. In the element bilga the third consonant exhibits a standard change, from voiced G to unvoiced K, the Akkadian hypocoristic ending –ya is added, and the second element mesh is omitted. The name Gilgamesh is presumed to be Sumerian, although it does not conform to any clear type of name in that language. The affix –ya is typical of Akkadian names, and it corresponds very closely to the Sumerian hypocoristic affix –mu. The ending –ya is, however, capable of an alternative interpretation; as a short, theophoric element standing for Yahweh. This analysis would give credence to the secondary use of the name for a supposedly Israelite king, even though no such king is named in the Bible. If this is the correct explanation of the name, it would imply that the pronunciation Bilgamesh continued alongside Gilgamesh during the first millennium B.C. Vowel changes in abbreviated Akkadian names are regularly found, such as Šūzubu from Mušēzib-Marduk. This analysis of Buluqiya as a form of Gilgamesh goes hand in hand with the choice of Mesopotamia’s most famous hero for showing that the coming of Muhammad was pre-ordained. Pseudo-prophecies such as this are always put into the mouths of famous men of old, to give them the stamp of authority. […]

Why has it taken so long to discover Gilgamesh in the Arabian Nights? Since the Epic was first discovered in Akkadian, written on clay tablets found in the ruins of Nineveh, progress in its decipherment and piecing together fragments has been slow. For a start, the name of the hero himself was wrongly read as Izdubar until late in the nineteenth century, when it was first correctly read as Gilgamesh. Not until about 1960 was the Sumerian and Hurrian pronunciation as Bilgamesh appreciated. As for the story itself, three episodes came to light quite early, and were the hallmarks by which the epic was recognised, namely Enkidu’s seduction by the harlot; the main heroic episode of Gilgamesh and Humbaba, and the Flood. None of these episodes is found in any version of the Buluqiya story.

I confess I have no idea how much credence to give to this. I’m automatically suspicious of any argument that depends on general similarity of content and ad hoc explaining-away of phonetic dissimilarity (“the consonants count for very little and the vowels for nothing at all”), but obviously this is a topic that has been much discussed by experts in a field where I am only a distant onlooker, so I turn to the assembled Hattery: does this Gilgamesh/Buluqiya thing seem plausible?

Update. The answer is apparently “No”; see ulr’s comment below, quoting A. R. George:

there is no evidence that the old pronunciation Bilgames (as opposed to the spelling ᵈbil.ga.mes) survived into the first millennium. All the evidence from cuneiform and alphabetic sources is that by that time the name was always pronounced with initial /g/. […] As a name, Buluqiya is not a version of GIlgameš.[…] the ‘overall story line’ of Buluqiya is not very similar to the plot of written Epic of GIlgameš. […] the tale of Buluqiya is so far removed from the period of cuneiform writing that speaking of the influence on it of compositions of the cuneiform scribal tradition is so speculative as to almost meaningless. How much else there was that stood in between!

I’ve posted about the spread of Persian as a lingua franca before (2013, 2018), and Amanda Lanzillo has a very interesting essay at Ajam Media Collective about an aspect of its history in India I wasn’t aware of:

In the standard narrative of the decline of Persian in India, as the Mughal Empire and its successor states waned and the British East India Company consolidated power on the subcontinent, Persian was displaced as a literary, intellectual, and administrative language. In this narrative, a loss of patronage, the slowing of migration from Iran and Central Asia, and elite use of “vernacular” Indian languages like Urdu all sped the downfall of Indian Persian. This narrative captures several processes by which Indian Persian declined between the eighteenth and twentieth centuries. However, it also obscures the dynamic language politics of colonial India, in which users of Persian negotiated the place of the language with the colonial state. The narrative of a linear displacement of Persian by Indian vernacular languages and English was a colonial ideal concealing a messier reality.

Persian was a major language of literary and intellectual production among North Indian Muslim elites from the twelfth century. By the sixteenth century, through Mughal patronage, it crystallized as the language of empire and the most prominent language of North Indian written discourse. Written Indo-Persian provided a shared idiom for the polyglot empire. Strong knowledge of Persian became a requisite for employment in many professional positions, including those traditionally held by Hindus; both Hindus and Muslims also sought Mughal literary patronage through mastery of Persian. Deccani dynasties likewise patronized Persian, and in both North Indian and Deccani contexts Iranian and Central Asians migrants contributed to the language’s prestige.

The British East India Company initially maintained Persian’s official position, relying on it to communicate with local power-brokers. However, following the administrative switch to English in the 1830s, Persian was increasingly marginalized in Indian society, to the degree that it largely disappeared from the public sphere by Indian independence. […] For colonial administrators, Persian had little claim to “Indianness” because it lacked inherent religious relevance or a vernacular constituency. By the mid-nineteenth century the regime encouraged vernacular education in languages like Urdu. Due to both Indian patronage and colonial encouragement, Urdu — a Persianized register of Hindustani — emerged as both a language for popular Islamic discourse and a shared secular idiom for discussing law, politics, and literature. […]

In mid nineteenth-century North India, many students were educated not by the colonial government, but in schools organized by neighborhood or religious leaders. Among these, Persian-medium schools were often the most prestigious, because Persian had previously been linked to the prospect of government work. Colonial reports admitted that even after the removal of Persian as a language of administration, Persian schools were attended by students of diverse religious and caste backgrounds “who can afford… this luxury,”because they were seen as imparting economically viable skills. […]

The desire of colonial education officers to draw the sons of prestigious Indians to government schools meant that they were willing to consider some Indian perspectives on what constituted a well-rounded education. In what one colonial administrator termed “a concession to popular opinion,” in the Northwestern Provinces the administration began to reintroduce Persian courses to government-run schools by the mid-1860s. As they did so they developed an educational style markedly different from the Persian education that dominated local schools.

In locally-run schools Persian was taught through exposure to classical texts. Poetry used in Persian education included ʿAbdul Rahmān Jami’s Yūsuf o Zulaykhā, the Būstān of Saʿadī Shīrāzī, and Qaṣāʼid of Urfī Shīrāzī. Prose texts included the Bahār-i dānish of ʿInāyat Allāh Kambūh or Mīnā Bāzār, an eighteenth-century work by an unknown author. To learn writing style, Indian students studied epistolary collections known as inshā’. These texts connected Indian Persian learners to a trans-regional literary sphere, while also offering access to localized intellectual heritage. Works like Mīnā Bāzār and Bahār-i dānish and many of the inshā’ texts, were specifically Indo-Persian, referencing local geographies and practices. […]

The importance of Persian education to families living in late-nineteenth century North India is often overlooked, perhaps because colonial rhetoric in the period treated Persian as irrelevant and emphasized the English-vernacular debate in education. Nonetheless, for many Indian elites, Persian remained a vital part of a well-rounded education. Persian literacy offered access to an extra-colonial identity marker and extra-colonial forms of employment and patronage.

I love this kind of excavation of forgotten elements of history, and I’m very fond of the Persian language, which I studied for a while a couple of decades ago — it’s easy to learn, fun to speak, and has one of the world’s great poetic corpuses. (Thanks, Trevor!)