More Than Money Can Buy

Honor culture and the cost of a curse (Defixiones, Part 6)

One of the many defixiones—lead curse tablets—found at the site of Aquae Sulis (modern Bath) is from one Docilianus. It’s famed for its Roman majuscules inscribed in a fine hand, though the text is fairly generic:¹

Docilianus Bruceri deae sanctissimae Suli devoveo eum qui caracellam meam involaverit si vir si femina si servus si liber ut […] dea Sulis maximo letum adigat nec ei somnum permittat nec natos nec nascentes donec caracallam ad templum sui numinis pertulerit.

Docilianus (son) of Brucerus to the most holy goddess Sulis. I curse him who has stolen my hooded cloak, whether man or woman, whether slave or free, that the goddess Sulis may afflict him with maximum death, and not allow him sleep or children now and in the future, until he has brought my hooded cloak to the temple of her divinity.

One does wonder what “maximum death” might refer to, but this and others from Roman Britain are of a different character from the more typical ones that seek to preventatively injure or constrain the target. By contrast, here:²

The overwhelming majority of curse tablets discovered […] were reactionary: an act or wrong had been done to the author and through the use of defixiones they sought to redress the matter.

Aquae Sulis in particular, is the find site for so many of this type of curse tablet; many theories were spawned about the possible reasons behind this:³

[T]he majority of thefts would have occurred at the baths, and at the hands of bathhouse thieves (fures balnearii), hence the large number of outer garments and coins lost. […] On the other hand, the loss may be due to careless and suspicious patrons of the baths misplacing such items, like rings, and instantly suspecting thieves.

The latter hypothesis is borne out by the fact that many incised gems found in the drain of the baths seem to have been lost when the adhesives with which they had been attached to rings were softened by the hot water, and appear among the “stolen” items complained of in the tablets.

Still, the fact that many of the defixiones from Aquae Sulis demand revenge for the thefts of only a few coins or inexpensive property stands in contrast to Uley, where:⁴

[T]he claims are of much greater value, with the greatest amount being of 100,000 denarii.

So of course the Aquae Sulis curse tablets raised speculation, such as that:⁵

[T]he majority of supplications were from individuals of a lower social standing, the victim not being able to afford a slave of his own, or to even pay one to mind his belongings while in the baths.

Or that:⁶

[A]fter the discovery of a theft while at the baths, the making of such incantations on curse tablets may have been a convenient method of exacting revenge at the height of the victim’s frustration.

But these are both unfounded. As we have seen, there is clearly a monetary component above and beyond that of the tablet itself. Rather, the cheapness of the defixio itself merely means there is a very low minimum threshold to that value. It is indeed the low value of the lead sheet itself that probably led to this element of the curse being discarded over time—as I’ve mentioned before, Aquae Sulis’ hoard of 12,000 coins attests this shift. Marina Pirinamonte cites the general decline in literacy as a reason for this, but as some defixiones are entirely pictorial, that’s clearly not the only factor at work.

An additional issue is that the settlement at Uley was rural and wealthy. Furthermore, the god worshiped there was Mercury Silvanus, syncretized with an unknown Celtic god, but perhaps similar to Moltinus, as the images of the deity here are notable for their horns. In any case, Mercury, as a god of commerce, cattle, and silver, would generally tend to have higher-class followers.

Sulis Minerva seems to have been less choosy—the baths were open to the public in a larger, urban setting. The goddess herself, with a Celtic name relating to the ideas of sight and light (cf. Old Irish súil, “eye”, Proto-Celtic *sūlos, “sun”), was perhaps a good choice to detect a thief, even of something small.

Returning to the curses, many describe the amount of money given to the god, typically some portion of the value of what was stolen, with one third being the lowest I’ve seen, for example in this quite businesslike message from Saturnina found at Uley, who may have been in the cloth trade:⁷

commonitorium deo
Mercurio Satur-
nina muliere de lintia-
mine quod amisit ut il-
le qui ho[c] circumvenit non
ante laxetur nissi quand[o]
res s(upra)dictas ad fanum s(upra)d[ic]
turn attul[e]rit si vir si [m]u-
lier si servus si liber
deo s(upra)dicto 
 [d]onat ita ut
exsigat istas res quae
s(upra)s(crip)ta sunt […].

A memorandum to the god Mercury from Saturnina, a woman, concerning the linen cloth which she has lost. (She asks) that he who has stolen it should not have rest until he brings the aforesaid property to the aforesaid temple, whether man or woman, whether slave or free. She gives a third part to the aforesaid god on condition that he exact this property which has been written above […].

The full value is also given sometimes, which seems strange, as, in effect, the curser is still losing that value.

In some cases, such as that of Basilia, which I presented in Part 4 or the one above, it seems the donation is to be made only if the property is recovered—in effect an ex voto. However, I think the evidence points in another direction: Just as Saturnina’s does, the word donat is nearly formulaically used in defixiones, which is, to be technical, the third-person singular present active indicative of dōnō, meaning “I give”. All of this means the best translation of donat is simply “(he/ she) gives”, leaving it unclear as to whether it is an action that has been or will be done.

Looking at another curse from Uley, it runs thus:⁸

Biccus dat M-
ercurio quidquid
pe(r)d(id)it si vir si m-
ascel ne meiat
ne cacet ne loqua-
tur ne dormiat
n[e] vigilet nec s[a]-
[l]utem nec sa-
nitatem ne-
ss[i] in templo
Mercurii per-
tulerit ne co(n)-
scientiam de
pederat ness[i]
me interceden-

Biccus gives Mercury whatever he has lost (that the thief), whether man or male (sic), may not urinate nor defecate nor speak nor sleep nor stay awake nor [have] well-being or health, unless he brings (it) in the temple of Mercury; nor gain consciousness (sic) of (it) unless with my intervention.

The only conditions made here seem to relate to the would-be victim rather than to the god or what is given to him. Furthermore, as we saw in Part 2, coins were placed within lamps, apparently as a substitute for defixiones, indicating the ritual and monetary offerings were commonly given at the same time. Looked at in this light, the conditions seem only to reflect what is being asked for in exchange for the value being given.

Taken together, this would mean after losing some property the supplicant would cast a curse and give the god they were entreating to intervene even more money, which, especially given some items were misplaced and not stolen at all, seems a clear case of throwing good money after bad: Even if their property was returned, which was far from certain, they might still be out the same amount, and as much as double if not. Nonetheless, it seems the injustice suffered was more the point than the monetary value lost.

In order to illustrate the concept at work here, we’ll have to examine another of the constellation of terms relating to value in the Graeco-Roman world, in this case timé (τιμή), which also means “honor”. In The Iliad, it was the timé Agamemnon took from him that sent Achilles to his tent, allowing the Trojans the upper hand in the war for a time.

Achilles argues with Agamemnon, largely as to the few riches he receives for “fighting himself weary” (ἐπεί κε κάμω πολεμίζων), with his final statement summing up the issue:⁹

νῦν δ᾽ εἶμι Φθίην δ᾽, ἐπεὶ ἦ πολὺ φέρτερόν ἐστιν
οἴκαδ᾽ ἴμεν σὺν νηυσὶ κορωνίσιν, οὐδέ σ᾽ ὀΐω
ἄτιμος ἐὼν ἄφενος καὶ πλοῦτον ἀφύξειν.

Now I will go back to Phthia, since it is far better to return home with my beaked ships, nor do I intend while I am here dishonored to pile up riches and wealth for you.

It’s very relatable 3000 years later; who hasn’t had a boss like that? But note Achilles’ use of the term ἄτιμος (atimos), indicating clearly it is his honor that has been taken, also forming a parallel to riches and wealth. I won’t gloss over the rather brutal fact that the timé being discussed is a human being: Briseis (Βρισηΐς), a Trojan princess whom the Greeks abducted and enslaved as a concubine.

Regardless, Achilles has what he feels is a legitimate grievance, and appeals to the gods for justice. Since his mother, Thetis (Θέτις), is a goddess, he doesn’t need to resort to the use of a defixio, but the language he uses is not dissimilar and again, the concept of timé is raised as central:¹⁰

μῆτερ ἐπεί μ᾽ ἔτεκές γε μινυνθάδιόν περ ἐόντα,
τιμήν πέρ μοι ὄφελλεν Ὀλύμπιος ἐγγυαλίξαι
Ζεὺς ὑψιβρεμέτης […].

Mother, since you bore me, though to so brief a span of life, honor surely ought the Olympian to have given into my hands, Zeus who thunders on high […].

Note there is a touch of dysphemia here: Achilles mentions his own disastrous fate, and lays the ultimate blame for Agamemnon’s failure to accord him honor at Zeus’ door. This is indeed part of the formula; Zeus owes him value, and so should act on his behalf. And the god does as he is asked: things turn quite badly against the Greeks, and even when Agamemnon eventually tries to coax Achilles back by meeting the demands he originally made, he refuses them. Clearly, the material value is less important than that of his injured timé. Only his rage when his cousin/ lover Patroclus (Πάτροκλος) is killed brings him back into the war.

While the Romano-Britons perhaps latched onto a particular aspect of the religio-magical tradition of defixiones, it seems clear despite the continuing worship of their local deities syncretized with or alongside those of Rome, the major elements of the practice remained very much intact.

This continuity extends from ancient Near East cursing praxes, which eventually came to be expressed as inscribed lead sheets. These, in turn, developed a distinct culture that spread right across the Graeco-Roman world, including a consistent set of analogies for sympathetic magic, rhetoric used to address the gods, and the exchange of value between gods and mortals. The fact that honor is set above pragmatic concerns seems to be yet another piece of this tradition that spans the whole region.

Read Subsequent Articles in This Series

Part 7: The Punic Curse Trail

Part 8: Hellenism Schmellenism

Read Previous Articles in This Series

Part 1: The Curses of Aquae Sulis

Part 2: Malefic Traditional

Part 3: Sympathy for Sauron

Part 4: Bargaining with the Gods

Part 5: Secundina’s Beef


  1. Tab. Sul. 10.
  2. Geoff W. Adams, “The Social and Cultural Implications of Curse Tablets [Defixiones] in Britain and on the Continent”, Studia Humaniora Tartuensia, 2006.
  3. Ibid, though the notion is attributed to Tomlin, Tabellae Sulis: Roman inscribed tablets of tin and lead from the sacred spring at Bath, 1988.
  4. Ibid.
  5. Ibid.
  6. Ibid.
  7. Uley 2, emphasis mine.
  8. Uley 4.
  9. The Iliad, Book 1, 168–71, A.T. Murray, trans., 1924, emphasis mine.
  10. Ibid, Book 1, 352–4, emphasis mine.

The Mysteries of Zūja-Go

The slang of Tokyo’s underbelly (Argots, Part 4/ Taishō, Part 4)

While researching the Asakusa (浅草) district during the Taishō period (大正時代, 1912–26), I ran across acharaka (アチャラカ), a term for a particular type of satirical comedy. Miriam Silverberg discusses it thus:¹

The derivation of the word acharaka [アチャラカ] has its own irony, for, like the Cockney and language of the down-and-out in Asakusa, it makes use of wordplay. Acharaka is a vernacular abbreviation of the phrase achira kara [あちら から], which means “from over there.” And the words “over there” referred to over across the ocean, from Euro-America.

What was this “vernacular”, I wondered—a cryptolect? Was there really a corollary to Cockney rhyming slang in Asakusa? This potential dovetailing of my interests was tantalizing. The research was difficult; there were many roadblocks, such as poorly informed and -written Wikipedia articles in foreign languages. But with luck, perseverance, and quite a bit of humility, I am able to present the story of zūja-go (ズージャ語), an argot from Taishō Japan.

Just as Silverberg did, Cockney argots are frequently referenced in connection with that of Taishō Tokyo. In particular, though lacking the global reach of rhyming slang, the spread and continued use of the Japanese cryptolect were otherwise similar. Junko Itō, et al. tell us:²

An argot […] known as zuuja-go, ‘jazz language, jazzese’, is widely used in Japanese jazz circles, from where it has spread to wider parts of the entertainment industry […].

That’s right, like Takurazuka, it has somehow survived from Taishō to now. The name zūja-go itself uses the argot term for “jazz”, as I’ll explain later, plus the Japanese for “language”, (語, go). Certainly terms for musical instruments and other elements of the jazz scene are central to the argot, but the inclusion in the lexicon of terms for things like karaoke and personal computer attest to its modern usage. The scholarly article continues, discussing how zūja-go works:³

The essence of the argot formation can be understood as analyzing words in two parts and switching their order […] The point of [zūja-go], the “fun of the game,” lies in a characteristic distortion of the input through reversal and further modifications.

The operative Cockney corollary here is one called back slang. As related by Silverberg, Wada Nobuyoshi made something of an ethnography of the down and out in Asakusa, in which:⁴

He points out that the hawker made use of linguistic reversal as one means of forging new, secret words. (One common example is the word enkō [エンコー], which is the reversal and abbreviation of the syllables comprising kōen [公園, “park”], the insider slang for Asakusa Park. The hawker language was thus not unlike Cockney, because of its consciousness of class base, and because of its use of the back-slang terms that reversed syllables in order to make a political point. It was, of course, also Japan-specific—a product of the modern years with a Japanese linguistic and social history.

The hawker is a figure at the margin of society. They are sellers of small goods, typically at least dodgy if not actually contraband. They need a patter, a shtick, to attract notice and move merchandise, often quickly before the authorities arrive. The Cockney type is well realized in Lock, Stock and Two Smoking Barrels character, Bacon (Jason Statham), selling jewelry “handmade in Italy, hand stolen in Stepney”, together with a shill, only to be interrupted by the “cozzers” (police). Here’s just part of his sales banter:⁵

It’s no good standing out there like one o’clock half-struck. […] These are not stolen, they just haven’t been paid for. And we can’t get them again—they’ve changed the bloody locks. […] It’s no good coming back later when I’ve sold out: “too late, too late!” will be the cry when the man with the bargains has passed you by. If you got no money on you now, you’ll be crying tears as big as October cabbages.

As for Cockney back slang, it essentially pronounces the phonological elements of a word in reverse. It’s similar to Pig Latin as far as the idea of reversal, but takes it further. The ludling yields results such as:

  • kayfabe: be fake
  • moniker: eke-name
  • pennif: from finnip, five-pound note, extended to any note
  • slop: from pols, police
  • yob: boy

I’ll note of these apart from moniker, which is etymologically dubious, yob, and kayfabe—an entry from American pro wrestling—none of these seems to have stuck. On the other hand, pennif and slop are noteworthy because, like some of the more impenetrable terms in rhyming slang, they layer back-slanging atop extant slang terms.

Zūja-go’s method of reversal sits somewhere between Pig Latin and back slang: it works with the syllabic nature of the language and flips those units. As with so many argots, there is a basic level that uses native Japanese words and performs a straightforward syllabic swap, as in terms like:

  • bukei (ブケイ): police inspector, from keibu (警部)
  • domoko (ドモコ): child, from kodomo (子供)
  • gaikichi (ガイキチ): crazy, from kichigai (気狂い)
  • suiya (スイヤ): cheap, from yasui (安い)
  • suriku (スリク): drugs, from kusuri (薬)

Then, of course, there are words borrowed from other languages. Foreign languages during the interwar period were quite prestigious, and especially so among devotees of the modern scene. “Ain’t that the Latest!”, a song from 1930, reflects the mixture of alienation and admiration foreign languages engendered at the time:⁶


The foreign talkie I don’t understand when I watch,
The bitterness of trying to look like I understand,
In front of the halfheartedly bobbed hair and Western clothing.
I imitate the foreigner sitting next to me a little,
The bitter laugh of one who fakes his way through.
Ain’t that the latest!

Certainly these qualities were emulated in the argot, which added a layer of both cool and obfuscation—and certainly the simple act of kanaization can make these words unfamiliar even to speakers of the languages they’re borrowed from—with terms like:

  • dammo (ダンモ): modern, modan (モダン)
  • hīkō (ヒーコー): coffee, kōhī (コーヒー)
  • kompaso (コンパソ): personal computer, pasokon (パソコン)
  • okekara (オケカラ): karaoke (カラオケ)
  • shītaku (シータク): taxi, takushī (タクシー)

Note some of these are already adding further layers because of the terms they derive from: karaoke mixes the Japanese kara (空) meaning “empty” with a shortening of English orchestra, while pasacon is an abbreviation of both of its English elements. Also in kompaso and dammo, we see a sound change occur as /n/ before a labial consonant (/b/, /m/, or /p/) shifts to /m/.

Other changes of this last type concern in vowel length, again fitting with Japanese’s typical patterns and making the argot words more wordlike, and hard consonants, a trickier one to explain: Orthographically, hard consonants are expressed by placing a character called a sokuon (促音) before the syllabic script element or kana (仮名) it affects. It is expressed as a small version of the kana tsu (つ/ ッ). As the sokuon and the subsequent kana are in different syllables, when argot terms are formed, there are essentially two options: applying it to a different consonant or simply pronouncing it as tsu. /n/ can also be a syllable on its own, and some reversals make this happen. Words evidencing each of these strategies are sampled below.

Change in vowel length:

  • zūja (ズージャ): jazz, jazu (ジャズ)
  • kūkya (クーキャ): audience, from kyaku (客)
  • rāko (ラーコ): cola, kōra (コーラ)
  • sharukoma (シャルコマ): commercial, komāsharu (コマーシャル)
  • mīno (ミーノ): to drink, from nomi (飲み)

Sokuon applied to a different consonant:

  • katte (カッテ): roll, from the sushi type tekka (-maki: 鉄火)
  • pakka (パッカ): water imp, kappa (河童)

Sokuon read as tsu:

  • kotsuya (コツヤ): guy, from yakko (奴)
  • kuribitsu (クリビツ): surprised, from bikkuri (びっくり)
  • patsura (パツラ): trumpet, from rappa (喇叭)
  • pīhatsu (ピーハツ): happy, happī (ハッピー)
  • totsuba (トツバ): bat, batto (バット)

Syllabic /n/:

  • mpa (ンパ): bread, from Portuguese pãopan (パン)
  • mpata (ンパタ): pattern, patān (パターン)
  • nto (ント): tone, tōn (トーン)

Then of course there are many that don’t fit the rules of the ludling, and therefore in effect better fitting the rules of an argot: sunite (スニテ, “tennis”), we might expect to be realized as either nisute or suteni, but the ni stays firmly in the middle. Yanopi (ヤノピ, “piano”) gains a /y/ present nowhere in English or Japanese. And bontoro (ボーントロ “trombone”) has clearly misplaced an /n/.

Shortenings, which indeed are the central matter of yet another Japanese argot, appear in terms like gishu (ギシュ, “socialist”) coming from the final syllables of shakaishugi (社会主義). Returning to acharaka, I’ve already given its derivation from achira kara, which Silverberg characterizes as an abbreviation, but we can see it’s not that simple: the argot flips the second word’s syllables, while in the first word, ir is dropped from the middle.

The next level of metamorphosis has to do with the Chinese characters—kanji—words are often made up of. When the positions of these are altered, so are their readings. Some examples include:

  • patsuichi (パツイチ): one shot, from ippatsu (一発)
  • kogaku (コガック): school, from gakko (学校)

If not for the kanji readings here, we might expect patsuitsu and kogatsu. But a still more extreme case is the term for the Ueno (上野) district, already represented by the simple argot term, Noue (ノウエ), but with its kanji reversed to 野上 changes its reading to Nogami.

Finally, we have terms such as inbenshon (インベンション), meaning “piss”, which layers a translingual pun over the ludling’s transformation: shonben, the Gunma dialect (群馬弁) pronunciation of 小便, to benshon (ベンション), which, with the addition of an initial in-, resembles the kanaized English invention. As we’ve seen repeatedly in argots, puns are commonly employed and the attractiveness of terms I’ve already mentioned, such as suriku and kompaso, comes from their resemblance to English words “slick” and “compass”.

The one I’ll end on is eburiuīku (エブリウイーク), the kanaization of the English phrase every week, which refers to shaomai (燒賣), the delicious dumpling. To get there, we take the Japanese name of the dim sum (點心) treat, shūmai (シューマイ) and flip it to maishū (マイシュー), a homophone of which is 毎週, translating into English as “every week”. Handily, the argot term also reminds you of how often you should eat shaomai.


Circling back to the Japanese Wikipedia Article for zūja-go recently, I found their documentation of the argot had expanded, including an entertaining way of expressing numbers using musical notes.⁷ The fact the argot was originally used in the demimonde of jazz musicians fits together with this delightfully. The pronunciations are based on the German note names and pronunciations of letters in the C major scale, thus:

  • tsē (ツェー): 1, from C
  • (デー): 2, from D
  • ē (エー): 3, from E
  • efu (エフ): 4, from F
  • (ゲー): 5, from G
  • ā (アー): 6 from A
  • (ハー): 7 from H
  • tābu (ターブ): 8, from octave, okutābu (オクターブ)
  • nain (ナイン): 9 from ninth, nainsu (ナインス)

Argot’s (and Taishō’s) thirst for the foreign can be seen here through the extra-obtuse use of both Western musical notation and German pronunciation. We can combine the set as normal Japanese numbers to get values like:

  • tābujūā (ターブ十アー): 86
  • dēsengēhyaku (デー千ゲー百): 2500
  • hājūēman (ハー十エー万): 730,000

Also included was the information that there has been a decline of the argot. Apparently, it saw frequent use by various performers in the ’80s. This meant broad audiences were exposed to the vocabulary. As I’ve noted, this tends to work against an argot’s ability to function as such, lacking the important connivance function. Now zūja-go terms fall under the heading of “Industry Jargon of the Bubble Era” (「バブル時代の業界用語」). The result of all this has been:⁸

[…] 隠語の意味合いが無くなってしまい […].

[…] the meaning behind the argot has disappeared […].

Certainly, in the US, there is a longstanding trend of commoditizing AAVE that mirrors this type of appropriation.The use of zūja-go in this context fits the classic definition of pastiche I quoted elsewhere:⁹

Pastiche is, like parody, the imitation of a particular or unique style, the wearing of a stylistic mask, speech in a dead language: but it is a neutral practice of such mimicry, without parody’s ulterior motive, without the satirical impulse, without laughter, without that still latent feeling that there exists something normal compared with which what is being imitated is rather comic. Pastiche is a blank parody, parody that has lost its sense of humor […].

It’s a fitting bookend to this series as I began it with a museum exhibition from that era looking back on the 1920s in Japan. As the exhibition’s catalog noted:¹⁰

The period was in some ways analogous to our own […].  “The 1920’s in Japan” tries to illuminate the age by looking at the art it produced.

As I noted, however, Japan in the time of the Bubble was marked by superficiality and consumerism. It seems retrospectively obvious cultural artifacts from an era with real depth would be revived, but used and understood on a skin-deep level. 

Read subsequent articles in the Taishō series

Part 5: How “Alice” Grew Big in Japan

Read previous articles in the Taishō series

Part 1: Japan’s Turbulent Taishō

Part 2A: Epochal Architecture

Part 2B: When Tokyo Moved West

Part 3A: Asakusa Movies

Part 3B: Asakusa Opera

Read previous articles in the Argots series

Part 1A: The Slang of Empyrea’s Automata

Part 1B: Canargy: a Cant How-To

Part 2A: Serious and Playful Cryptolects

Part 2B: Me Talk Pretty Ludling

Part 3: Rhyming and Stealing


  1. Miriam Silverberg, Erotic Grotesque Nonsense: The Mass Culture of Japanese Modern Times, 2007.
  2. Junko Itō, Yoshihisa Kitagawa and Armin Mester, “Prosodic Faithfulness and Correspondence: Evidence from a Japanese Argot”, Journal of East Asian Linguistics, 1996.
  3. Ibid.
  4. Wada Nobuyoshi (和田 信義) from Silverberg, 2007. I was unable to find the original work, so we’ll have to trust her.
  5. Guy Ritchie, Lock, Stock and Two Smoking Barrels, 1998.
  6. 松竹蒲田音楽部 (Shochiku Kamata Music Club) lyrics, 「尖端的だわね」 (“Ain’t that the Latest!”, Sentanteki Dawane), 1930.
  7. 「ズージャ語」 (“Zūja-Go”), 『ウィキペディア』 (Wikipedia), retrieved April 2023.
  8. Ibid.
  9. Fredric Jameson, “Postmodernism and Consumer Society”, The Cultural Turn: Selected Writings on the Postmodern, 1983–1998, 1983.
  10. Catalog, 「1920年代日本展」(“The 1920’s in Japan”), 東京都美術館 (Tōkyō Metropolitan Art Museum), 1988.