The Eight Trigrams and Their Changes

MatthiasHayek

An Inquiry into Japanese Early Modern Divination

In thisarticle I willstudy a peculiardivination method involving the eight trigramsknown as hakkethat, I willattempt to show,was amongthe most populartechniques used in Japanfrom the end of the sixteenth to theend of theeighteenth century. My goal hereis to showhow this mantic knowledge was passedon througha specifickind of manual while undergoing several transformations.These changes, far from being coincidental, may be linked to theinner evolutions of Japanese society and cultureduring the Edo period. Thereforesuch an inquiryshould help us to gaina betterunderstanding of thereciprocal informing relationship between mantic knowledge (correlative thinking)and peoplesgeneral expectations and/or mentality. I will first pres- entthe nature of hakke-uranai, before tracking down how and by whom it was used.Finally, I will detail how its inner structure relates to theway the clients ofthe diviners were viewing fate, time, daily life, and the world.

keywords:divination - hakke-uranai - printed manuals - correlativethinking - wayof yin and yang- popular knowledge

MatthiasHayek is an associate professor inthe Department ofOriental Languages and Civilizations(lcao) andmember ofthe crcao, umr 8155, Paris Diderot University.

329

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions ( uranai bokusenЬй) in Japanhas yetto receiveatten- tionfrom scholars equivalent to thatwhich specialists of China have Divinationgiven to continentalmanteia. To be moreprecise, although seminal studiessuch as Blacker s (1975) havebeen conductedon "non-systemized"(or "shamanistic")divination (that is, divination that relies on theinner capacity of thediviner to directlycommunicate and transmitinformation from a superhu- man source),the same cannotbe said about "systemized"divination. Pioneer Westernscholars, first Severini (1874),but also Aston (1908) and others,did showsome curiositytoward the subject, and abouthalf a centurylater, French researcherBernard Frank (1998)produced a detailedstudy on a peculiaraspect of ancientJapanese hemerology1 that is stillregarded as an essentialcontribu- tionto thecomprehension of Heian cultureand lore. Evenin Japan,however, academic interest in whatI shalldefine as a corpus- based technicalknowledge used to deciphermundane events through a process of encodingand decodingreality in analogical/symbolicalterms rarely leaves theboundaries of ancient Japan. Although valuable studies about the social sta- tusof "religious specialists" whose activities included divination have provided us witha moredetailed understanding (Hayashi 2006), thecontents and the sourcesof these divination practices are still opaque. Eventhough Japanese systemized divination was closelyconnected to Chi- nese manticknowledge, considering how important(and numerous)the diviners seem to have been in premodernurban and ruralJapan, it deservesto be examinedfor its own sake. Given the triangular relation of information between diviners,their source(s) of knowledge, and theirclients (Zeitlyn 2001),it can be assumedthat by studying the nature of Japanese mantic practices, one could unveilthe specificities of the way people apprehend the surrounding world, fate, and everydayevents at thattime.

1.The various terms used in this article to refer to a precisetype of divination generally fol- lowsthe typology established bysinologists inKalinowski 2003. Hence, "cleromancy" refers toany divination involving todraw/throw/toss/flip anitem, in order to randomly obtain one ormore figure from a preset list. Classical Yi jing- based divination (achilleomancy) qualifies perfectlyfor this labeling as itconsists of one or two hexagrams, line by line or trigram by trigram,byseparating atrandom a pack fifty sticks, and subtracting sticks several times until a meaningfulnumber isreached. As for "hemerology," itis any kind of divination taking tem- poralparameters, such as year,month, day, hour, or any combination ofthese as itsprimary variableto determine either ones faith or the auspicious character ofa givendate regarding variousactivities.

330

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 33I

Thisis preciselywhat this article attempts, by paying particular attention to a peculiardivination method involving divinatory figures commonly known as theeight trigrams (Jp. hakka or hakke, Ch. baguaAíh). First,I willdescribe this technique,pointing out itsimportance in earlymodern Japan, before giving a briefoverview of its origins, characteristics, and mediaof diffusion. Then I will considerevidence of its use fromthe sixteenth to thetwentieth century, portray- ing itsusers by quotingtestimonies from contemporaneous sources. Finally, I shallpresent what the mechanics and structureof this method reveal about the preoccupationsand representationsofthe clients of the diviners.

Prologue: What is a trigram? Beforegoing into the detailsof the Edo-periodusage of the eighttrigrams, I shalltry to clarifya fewpoints regarding the nature and originof these figures, as wellas theirrelationship to divinatorytechniques. The firstdifficulty faced when trying to understandthe hakke-uranai is that theeight trigrams are commonlyassociated with a specifickind of divination thatdiffers greatly from the one I willtry to discusshere. Indeed,when looking at a dictionarylike theNihon kokugo daijiten E№H we can see thatthe word hakke refers implicitly to thefamous Chinese classic,the Book of Changes, Yijing (Jp. Ekikyõ ЛИ), also knownas theChanges ofthe Zhou , Zhouyi(Jp. Shüeki Л): In thecontext of the Changes , the eight forms are composed of three divina- tionrods figuring yin (broken) and yang (plain) lines. Thisbook, assumed to havebeen firstput together around 700 все duringthe Zhou dynasty(Suzuki 1963,15), at itscore servesas a base and a referentfor a divinatorytechnique involving a randomdrawing of yarrow stalks in orderto obtainnumerical values (cleromancy).These values are used to form,step by step,a divinatoryfigure ultimately composed of two sets of three lines. Both the three-linefigure, or trigram,and thesix-line figure, or hexagram,are calledgua (Jp.ka , ke). The linescan be eitherplain (uneven,or yang)or broken(even, or yin),their status being determined by thenumbers which were drawn. More- over,depending on thevalues obtained,the capacityof each line to "change" fromyin to yangor fromyang to yinrespectively is also determined.Therefore, on a practicallevel, a gua is nothingless thana figurativeportent, to be inter- pretedby looking at the parts of the Book of Changes dedicated to eachhexagram. However,beside this practical aspect, the Yijing itself,and the bagua along withit, are also a base formetaphysical speculations. This particular side has been extensivelydeveloped since the beginning of Confucianism (the commen- tariescalled "tenwings" are attributedto Confuciushimself), and is a partof

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 332 I JapaneseJournal of Religious Studies 38/2(2011)

SupremeUltimate

yang ' шшв

greatyang //^/sman yin smallyang greatyin

ш я ш ш m ж & w Ken Da Ri Shin Son Kan Gon Kon

diagram1. Formation ofthe trigrams. thecommon culture of the various intellectual and religioustraditions of China, includingof course Buddhism and Daoism. Thus,the formation of the trigrams is oftendepicted as a generativeprocess resultingfrom a successionof combinations starting from the emergence of the two poles,yin and yang(monads), fromthe SupremeUltimate (Taiji, Jp.tai- kyokuicffi). As such,this process can be consideredas a symbolicalexpression ofthe creation of the cosmos itself. Thiscosmological aspect is notabsent from the divinatory use ofthe Book of Changes: the trigram and hexagramrandomly obtained work as a symbolization ofthe state of the macrocosm at thetime of the consultation, and the"changes" whichmight occur to thelines express the cosmic dynamics the diviner should catchto foreseeupcoming developments. However,along withthe systematizationof the variouscosmological ele- mentsled by Han (202 BCE-220 ce) Confucianscholars like JingFang ЖМ (77-37все), thetrigrams were integrated in a broadercorrelative framework, as wellas thefive phases, the twelve branches, the ten stems, and othermarkers of space and time(Suzuki 1963).From these times onward, they were to be found in otherdivinatory techniques than the originalcleromancy, most notably in hemerologicalpractices. Contrary to cleromancy,which uses randomvariables (numberof rods/coins)unrelated to the subjects*individual data to obtaina portent,hemerology is based on calendricalelements and uses fixedvariables organicallylinked to thesubject (for example, the birth year of the client, the day and timean eventoccurred, and so on) as a basis forits prognostics. As such,

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 333

figure1. Hakke-bon, undated (Shidõ figure2. Hakket 1708 edition by Seishõ bunko,Keiõ Gijuku University). (Collègede France, ihej). whilecleromancy emphasizes the intervention of a divinedesign in thedrawing, thisset of techniques oifers a visibleand organicrelationship between the omen and theclients individual situation. The hakke-uranaithat was in use in Japan preciselypertains to thisnew category of techniques, as we willsee below.

Lookingfor the Trigrams in Early Modern Japan Gettingback to the definitionof hakkein the Nihonkokugo daijiten , we are presentedwith a second meaning:"divination (by themeans of the Changes). Hakke-mi,lit. one wholooks at theeight trigrams. Diviner." Weare then led to believe that there were, during the Edo period,diviners called hakke-miAihMi using the yarrow stalks (cleromancy) as theirmain technique. Evenif it might very well have been true by the end of the early-modern era, when Yijing-based cleromancymanuals were flourishing (see, forexample, below), we cannotmake the assumption that the technique used bythese hakke-mi has alwaysbeen thesame. The key to solvingthis problem lies in themeaning of the wordhakke in the early-modern context. The Nippo jisho Bffiiîlr (Õtsuka 1998),

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 334 I JapaneseJournal of Religious Studies 38/2 (2011)

a Nippo-Portugesedictionary published by the Jesuits of Nagasaki in 1603,gives a ratherdifferent definition of the word:

Facqe: A calendaror repertoryused by astrologers ( astròlogo ). Facqeo miru: To lookat thisbook or repertoryin orderto knowthe destiny and fateof the people. (Otsuka 1998,150) Fromthis contemporaneous definition, we can saythat around 1600 theword hakkedid notspecifically referred to thetrigrams of the Yijing , butto a kindof book used bydiviners. Moreover, the comparison made withcalendars, which are usuallyfolded books ( orihon), givesus a hintregarding the material form of said books. However,this is merelythe tip ofthe iceberg. Browsing through the Union Catalogueof Early Japanese Books , we findthat at leasta hundredbooks bearing hakkein theirtitle were produced between 1611 and theend ofthe Edo period. We are in factfacing a whole genre,which might collectively be classifiedas booksof trigrams, hakke-bon Ai№.

BOOKSOF TRIGRAMS:TYPE, CONTENT, AND EVOLUTION

To date,only one academicarticle has triedto shedlight on thebooks of trigrams: thisis Masuko Masarus(2006) briefpresentation of their global structure, which alsogives valuable bibliographical information. Still, Masuko s studyremains incom- plete,and does notclearly explain the origins and transformationsofthese books. Therefore,having defined hakke-bon , I shallillustrate several hypotheses regard- ingtheir origins, and givean overviewof their developments and achievements. Hakkematerials can be dividedinto three different types. First, we have a groupof foldedbooks, orihonï/r^, and theirreprints, whose "commercial" publishingstarted as earlyas 1611.According to theirform, we can assumethat theseare thebooks mentioned by the Jesuit witnesses, which means - giventhe - publicationdate of the dictionary thatthey were already to be seen,probably as manuscripts,before 1611. The contentsof theseearly books are essentially technical,providing mainly tables and diagrams,with almost no detailsabout theprocedures. Second, from c.1660, we findbound books thatgive detailed instructionsabout how to use divinationtechniques, and explanationsof the meaningof the rather esoteric indications presented in theearlier folded books. Thissecond category,quite different in naturefrom the earlierone, marksan importantstep in theevolution of the hakke material. Finally, at theend ofthe seventeenthcentury, "special" editions start to appear,which greatly exceed thetwo previous ones in contentby including several other types of divinatory knowledge,such as physiognomy( ninsõ Affi) and thelike, almost unrelated to theoriginal technique. The mostexciting observable feature, when looking at

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 335 thetransition, or ratherevolution, from the original folded books to thelatest compilations,is how itparallels developments seen in othertypes of early mod- ernpublications. As such,it echoes the changes occurring in broaderintellectual tendencies,mainly Neo-Confucianism, but more broadly the emergence of the kõshõgaku,which I translateas "philology"in variousfields of knowledge such as classicalstudies, historical studies, linguistics, and so on. Throughthe content and structureof these manuals, we willsee how divinatorytechniques evolved fromcomplex, esoteric knowledge reserved for rather educated elites, to a more practical,simplified, and yetdiverse knowledge, made availableto a broader audienceby self-asserting compilers and authors.

EARLYFOLDED BOOKS AND THE CORETECHNIQUE

As a whole,beside their titles, trigram books can be definedby their lowest com- mondenominator: the divinatory scheme they all provide.This consists of eight square-shapeddiagrams, each dividedinto nine squaresections, with one trigramin thecenter and eightothers disposed around. Each centraltrigram has thename of a Buddha or a bodhisattvaassociated with it, and theeight other trigramshave different positions with specific names attached, as wellas other elements,numbers, directions, agents, and so on. Thesenames, zettai (collapsingbody), zetsumei (collapsingdestiny), kagai ШШ (disaster),seike (birthhouse), fukutoku ÎSfê (fortuneand virtue),у йпеп Ш ^ (annualtransfer), уйкоп (soul transfer),and teni 5ÇS (heavenlydoctor) represent what we maycall "mantic functions": they lead to theresults of the divination, depending on thevariations of the parameters and manticvariables. Thus, the diagrams can be seen as specificconfigurations of the parameters.The followingschematic representationsbear all thefeatures I just described,and can be consideredas archetypicalof what these diagrams are. The core techniqueinvolving the diagramsis a formof hemerology/horo- scopybased on calendricalvalues. Most ofthe books beginby presenting two keyvariables and waysto obtainthem. First comes the subperiodof thebirth of an individual,gen té, whichdesignates a span of sixtyyears (one complete cyclein thehexadecimal calendar), included in a super-cycleof one hundred and eightyyears, so thereare threesuccessive gen , a superior,a middle,and an inferior.Second, we usuallyfind a tabledetailing the "induced sound," natchin Йн, associatedto each ofthe sixty combinations of stems and branches.These soundsare in facta developedform of the five phases (wood, fire, earth, metal, and water)and thereforerepresent the peculiar agent of an individualaccording to his yearof birth.According to theseparameters, and the sex of the person concerned,the attributed (lit. "hit upon") trigram,tõke S ÌK and subsequently the correspondingdiagram, can be determined.From there, the divinercan

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions Fortune Annualtransfer Soul transfer 5 wind 3 fire 8 earth

dragon3 snake4 5 horse sheep6 monkey7

Birthhouse Seishi(Ri) Heavenlydoctor 4 wood fire 2 metal 2 rabbit 8 rooster

Disaster Collapsingbody Collapsingdestiny 7 mountain 6 water 1 heaven

tiger1 ox 12 rat 11 boar 10 dog 9 DIAGRAM2

Heavenlydoctor Soul transfer Annualtransfer 5 wind 3 fire 8 earth dragon3 snake4 5 horse sheep6 monkey7

Disaster Dainichi( Kon) Fortune 4 wood earth 2 metal 2 rabbit 8 rooster

Birthhouse Collapsingdestiny Collapsingbody 7 mountain 6 water 1 heaven tiger1 ox 12 rat 11 boar 10 dog 9

DIAGRAM3

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions Soul transfer Heavenlydoctor Fortune 5 wind 3 fire 8 earth

dragon3 snake4 5 horse sheep6 monkey7

Collapsingdestiny Fudõ (Da) Annualtransfer 4 wood metal 2 metal

2 rabbit 8 rooster

Collapsingbody Disaster Birthhouse 7 mountain 6 water 1 heaven

tigeri ox 12 rat и boar 10 dog 9

DIAGRAM4

Disaster Collapsingdestiny Collapsingbody 5 wind 3 fire 8 earth

dragon3 snake4 5 horse sheep6 monkey7

Heavenlydoctor Amida(Ken) Birthhouse 4 wood heaven 2 metal

2 rabbit 8 rooster

Fortune Soul transfer Annualtransfer 7 mountain 6 water 1 heaven tiger1 ox 12 rat 11 boar 10 dog 9

DIAGRAM5

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions Birthhouse Collapsingbody Collapsingfate 5 wood 3 fire 8 earth dragon3 snake4 5 horse sheep6 monkey7

Fortune SenjuKannon (Kan) Disaster 4 wood water 2 metal 2 rabbit 8 rooster

Heavenlydoctor Annualtransfer Soul transfer 7 mountain 6 water 1 heaven

tiger1 ox 12 rat 11 boar 10 dog 9 DIAGRAM6

Collapsingdestiny Disaster Birthhouse 5 wind 3 fire 8 earth

dragon3 snake4 5 horse sheep6 monkey7

Soul transfer Kokúzõ( Gon) Collapsingbody 4 wood mountain 2 metal 2 rabbit 8 rooster

Annualtransfer Heavenlydoctor Fortune 7 mountain 6 water 1 heaven tiger1 ox 12 rat 11 boar 10 dog 9

DIAGRAM7

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions Collapsingbody Birthhouse Disaster 5 wind 3 fire 8 earth dragon3 snake4 5 horse sheep6 monkey7

Annualtransfer Monju (Shin) Collapsingdestiny 4 wood wood 2 metal 2 rabbit 8 rooster

Soul transfer Fortune Heavenlydoctor 7 mountain 6 water 1 heaven tiger1 ox 12 rat 11 boar 10 dog 9 DIAGRAM8

transfer Fortune Annual Heave"doctor 5 wind 3 fire

dragon3 snake4 5 horse sheep6 monkey7

Collapsingbody Fugen(So«) Soul transfer 4 wood wind 2 metal 2 rabbit 8 rooster

Collapsingdestiny Birthhouse Disaster 7 mountain 6 water 1 heaven 11 boar 10 9 tiger1 0x12 rat dog DIAGRAM9

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 340 I JapaneseJournal of Religious Studies 38/2(2011)

proceedon to two differentkinds of operations.The firstand mostobvious is to use thediagram as a guideto annual(month by month) predictions and rec- ommendationsfor a givenindividual. Depending on theposition of the mantic functions,auspicious or inauspiciousmonths, days or directionscan be inferred, and theresults refined through the relation of each trigramand theattributed one and/ortheir respective agents. In fact,this is the onlymethod the folded booksexpose, though manuscript annotations point toward other possible uses. Theother method, which became visible to thepublic eye with second-generation books,is ratherdifferent, as it assignsnumbers to various"objects" of divination (seen things,excepted things, awaited person, and so on) and introducesexter- nal variablessuch as time.Ultimately, the procedure consists of a calculation in orderto obtainone ofthe numbers attributed to each manticfunction. This techniquerepresents a significantevolution compared to those foundin the first-generationbooks, and wentthrough several transformations over time. I shallexamine it more precisely in thelast part of this article. In additionto this core, hakke-bontypically present at least threeother hemerologicalmethods. Two ofthem are ratherpervasive through the differ- enttypes of manuals, while the last one progressivelyappears less and less fre- quentlyas we approachthe later publications. We thusfind a systemreferred to as the twelveconducts, jüni-un t-1, whichestablishes five cycles of twelve stagesfor each agent.These cycles govern the fortune of an individualaccording to thestem of his birthyear. The secondsystem involves nine "luminous stars," kuyõshõЛВЩМ, that is, the two luminaries, the five planets (Mars, Venus, Mer- cury,Jupiter, and Saturn),and twoextra "pseudo" planets, Rago ШШ and Keito Iti R.2It is used in a fashionsimilar to theattributed trigram, but only gives one typeof result for each planet.The last essential hemerological scheme, compara- tivelyshort-lived in the Edo period,consists in a fulltable listing the 28 (27) lunarlodges for every day of the year.3

2. Thesetwo additional planets originate inIndian astrological views that were passed to Japanthrough Buddhist scriptures such as theSukuyõ-gyõ ШВ1Ш (Treatise ofthe lodges and luminaries), complied under Amoghavajras (Jp.Fukù ^ 5?)direction. They can be spotted in variousiconographie material depicting Buddhist astral deities such as starsmandalas, often bearinga dreadful appearance. Due to the close association between Buddhist astrology and courtonmyõdõ sincethe beginning ofthe Heian period, they were soon incorporated intoonmyõjis hemerological practices, eventually becoming associated with other deities of Chineseorigin (though such an association might well have been already established inChina, beforecoming to Japan). Linked to the eclipses and the comets, they are known in Western astrologyas the Dragon head and tail, Caput and Coda Draconis. On Buddhist astrology see Yano1986. 3.The 28 (27) lunar lodges (Sk. naksatra) which should not be mistaken for the 28 mansions ofChinese astronomy (though they bear the same Chinese characters), area productof Indian astrology.Basically, they form a lunar zodiac (division ofthe ecliptic) of 28 signs or lodges. The

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 34I

figure3. Hakke>1708. Diagramof the borrowed path(left) and diagram of thenine luminous stars (right).(National Diet Library)

All thoseelements, as wellas othercontent specific to each book,did notpop outof nowhere at thebeginning of Edo. However,due to thelack of detail in the firsteditions (they basically give the bare bones diagramsand tables,without eventrying to explainthe actual methods, much less quotingsources), it might provequite difficult to tracethe origins of thesemethods. Fortunately, besides theearly manuals, I havebeen able to retrievea manuscript,probably dating to themiddle of the sixteenth century. While its content bears great similarity to thefirst printed books, the specifics of this example shed some light on thepro- cess thatled to thedevelopment of the hakke-bon genre. Itopens with a tableshowing correspondences between the twelve hexagrams and thetwelve earthly branches, symbolizing the months of the year, immedi- atelyfollowed by an introductorypart, which is almostsystematically found in thefirst printed books. This introduction reads as follows: Themethod of yin-yang and theeight trigrams: it is said,in theNine Palaces Treatisebrought by Kibi no SaneyasuAson,4 that when Heaven and Earth partedaway, a Kinoe-ne[yang wood-rat] year, the first subperiod started. It is also saidthis was 61,672years before year 1, yang wood-rat, of Jinki in Japan. listis usually limited to 27, as themoon needs approximately oneday to go through a lodge, but thereis a intercalarylodge, Abhijit, orthe Ox lodge in Sino- Japanese denomination (gyüsuku 41 ît),held for "hidden" (hisuku ШИ), and inauspicious. 4.Although there islittle evidence asto whom this name is supposed to refer, itfeels reason- ableto assume this is Kibi no Makibi InHÄDi (695-775), who allegedly brought yin-yang texts andknowledge from China.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 342 I JapaneseJournal of Religious Studies 38/2(2011)

Thethree periods complete a cycleevery 180 years. Anticipating that many untutoredpeople will experience difficulties in understanding, this text tries to calculatethe periods. The inferior period ended in Jian3, yin water-boar [1023],then, in Manju1, yang wood-rat [1024], we entereda superiorperiod. In Kennin3, yin water-pig [1203], an inferiorperiod started. According to the sametreatise, the superior period rose in the first palace5 in yang wood- rat, the middleperiod rose in thefourth palace in yangwood- rat, the inferior period rosein the seventh palace in yang wood-rat, and so on.In Kakitsu3 [1443],the superiorperiod ended, and in Bunan1 [1444],the middle period started. It endedin Bunki3 [1503],and in Eishö1, yang wood-rat, we enteredan inferior period. In otherwords, it offersan exampleof a calculationof the subperiodsvital to thismethod of divination. We noticea referenceto a treatise,which I shallintro- ducelater. The manuscript then continues on withthe kind of content described above,adding some elementsto thediagrams like, for example, figurations of thenumbers by countingrods, or internalorgans. Most noticeably,the writer feltthe need to includeseveral mantras written in Sanskrit.Lastly, results correspondingto thenine luminous stars (or planets),kuyõshõ , the twelve conducts, jüni-иПу6and the twenty-sevenlodges are prominentlydescribed, compared to thefolded books, and mostof all, the main objects of divination seem to be eitherillness or warfare.Moreover, the results concerning the former include- aside fromthe eight positions of the trigrams - anotherfunction, great misfortune,daiyaku whichis neverto be seenin theorihon. Fromall theseclues, we can alreadymake a feweducated guesses about the originof the described methods. First, looking at theassociation between hexa- gramsand months,the symbolist theories that are characteristic of Han divina- tionimmediately come to mind.The inclusion of the Yijings figures to a broader analogicalreseau is alreadyclearly established in thethought of Jing Fang (Shin 2002,102-17). Furthermore, this peculiar combination, which in factrepresents the increaseand decreaseof the yin and yangduring the year,can be found in Xiao Jilo's Wuxingtaiyi (Jp. Gogyõ taigi) Sfr (Compendium of

5.The notion of the three monads or subperiods, sangen isclosely related to the one of thenine palaces, kyügü. See Kalinowski 1985, 774-811. 6. Thesetwelve conducts represent the cycle of the five phases according to themonths. Theyare supposed to influence the destiny of one individual, depending on the stem of his birthyear, and thus follow the course of human life. We have, in order, the embryo, tai In, nutrition,yõ birthand growth, chõ(sei) purification,moku(yoku) r^(fër), maturity, kan(tai)Щ^г), takingof position, rin(kan) ËÛCb), reign, tei(õ) íír(BÍ), deliquescence, sui illness,byõ ШУ death, shi tomb,bo Ш , andexpiration/formation ofthe "breath," zetsu Ш (betweenthe parentheses, I have filled in the signs usually found in hakke books according to theWuxing taiyi).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 343 the fivephases) (Kalinowski i99i> 93»236-37). This workof the Tang period is well known forhaving been introducedin Japanalong with the Chinese bureaucracy,and was highly regardedby the specialists of the Divination Bureau.7The con- nectionsbetween the Wuxing taiyiand the hakke-bondo not end there:Xiao Jťsbook also containsvarious elements that we can relate to the manuals and thatwere, moreover, allegedly the source of divinatory practices during the Heian Period. The method of the attributedtrigram, for instance, greatlyresembles the calculation made by courtdiviners to determinethe annual "trigram forbiddance"for an individual, hakkeno imi (Frank 1998, 88, 118). In return,this practiceis clearlybased on the Wuxingtaiyi. What is more, this book directlyquotes the figure4. Tableof the 12 birds and beasts in the the treatise Kyügü-kyö earlyversion of the Hakke-bon (International cited in the introductorypart ResearchCenter for Japanese Studies [ircjs], top) of the trigrambooks. This lost andthe 1708 Seishõ version (bottom). textseems to have presenteda systemprecisely involving magic squares in relationto thecalendrical calculationof a "manticdeity," the Grand One, Taiyi (Kalinowski 1985).Yet, sev- eraldifferences preclude the assumption that there was a directrelation between thetwo methods. The subperiods,most noticeably, although they are somewhat describedin theWuxing taiyi , are not involved in thedetermination ofthe forbid- den trigrams.Moreover, even if the Gogyõ taigi evokes mantic functions, itonly

7.According tothe Code (the Taihõryõ)y theonmyõryõ wasan administrative organof thestate dedicated tocalendar production, astronomy (astromancy), anddivination proper (Frank 1998).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions figure5. Shinsen iriyõ hakke narabini shô , 1667, table of contents (Collège de France, ihej).

figure6. Shinsen iriyõ hakke narabini shõy 1667, list of results.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions figure7. Shingon himitsu hakke kudeny Higashi 1693. Manu- script,diagram ofthe borrowed path (Collège de France, ihej).

figure8. Daikõyaku shinsen hakke-shõ genkai, 1718, Ri diagram and explanations (ircjs).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 346 I JapaneseJournal of Religious Studies 38/2 (2011) refersto twoor three,not to eight.However, various Tang manuscripts found in theDunhuang caves do includedivination techniques very similar to whatcan be seen in thehakke books, most noticeably one called"annual transfer on the eighttrigrams," bagua younian AíhJSÈ^ (with subperiods and manticfunctions), butalso thenine luminous stars, lunar lodges, or twelveconducts. In particular, a treatiseof medical hemerology bears many common points with the Tenmon manuscript,notably the calculation of the great misfortune (Kalinowski 2003, 502). Therefore,one maysuppose that such compilationswere transmitted to Japanat somepoint, and servedas a base to whatbecame hakke-bon. The foldedbooks themselvesalso givesome insightsregarding the origin of theircontents, especially on whatthey do not sharein commonwith the man- uscriptof the National Observatory. There are morethan ten different books of thistype that have made their way to us, nowpreserved in variousJapanese and Europeanlibraries. The oldest, Hakke zue AíMUâ" (1611),presents an unmatched specificity:it doubles each diagramwith an illustratedmagic square. Though the exactmeaning of the pictures has yetto be unveiled,it is possibleto findsome redundanciespointing to a logicalassociation between the depictions and each manticfunction.8 Among the other books of this type, usually entitled hakke bon , or hakkesho , onlya fewcan be preciselydated, but most of them are very similar to theOnmyõ hakke no hõ ШШAítóžř (1628),which is presumedto be theoldest nextto theHakke zue. However,later editions show several différences, including newdiagrams and a lackof some minor features.9 Concerning their sources, upon examinationwe can see that they heavily reuse whole passages of the Sangoku sõden onmyõkankatsu hoki naiden kinu gyokuto-shü ЯШ (see Nakamura 2000), usuallyabridged as Hokinaiden , an esotericcompi- lationof hemerological knowledge, including elements from both the curial and themonastic (Buddhist) mantic tradition (mostly in theparts devoted to astral deitieslike Konjin jfe# or Daishõgun;MÇj|l).10 Furthermore, they innovate by

8.For example, pictures ofpeople lying on the ground are associated with the zetsumei function,and horses to the kagai function. 9.Interestingly, thelast known version (1708) met with an important downscaling interms of size(almost reduced to half what it used to be) and quality (most of the pictures are absent and replacedby simple names). Such a transformationmightbe interpreted as a loss of influence of thistype of book at the beginning ofthe eighteenth century. 10.Regarding theHoki naiden , there are several theories about when and by who it was com- piled.Usually presented as Abe no Seimei s apocryphal work, it seems to be closely related to theYasaka shrine in Kyoto, aswell as toEsoteric Buddhist schools. As a whole,the Hoki naiden canbe characterized asthe greatest example of how divinatory knowledge has been transmitted throughthe medieval period in a mythologizedform: most of the first part of this text describes variousmantical elements embedded ina mythicalnarrative about Gozu Tennõ the deityof the Yasaka shrine, and his struggle against the great king Kotan gJL, his arch enemy See Nakamura2000, 237-330; SAITÕ 2007, 140-97.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 347 includingelements from a Song almanac,Yanqin doushu sanshixiang-shu (Jp. Enkintosu sanzesõ-shó) like a horoscopicsystem involving 11 twelvebirds (Masuko 2006, 52-55). Therefore,the early hakke-bon as a whole appearedas a firstattempt to exotericizea peculiarkind of manticknowledge thatpassed from China to Japanbefore falling under secret transmission during theMiddle Ages. They reveal a panel oftechniques whose main use was seem- inglyto determinethe outcomes of illnesses and militarycampaigns, while add- ingmore "popular" devices focused on individualfate. Though early manuals are usuallydevoid of elaboration, they do notgive any kind of results associated with thediagrams, nor explanations about the aims of the method, and someof them includeannotations prefiguring the listed results of later books. In fact,given their formatand contents,they can be characterizedas "tools"and it is thoughtthat moredetailed guides explaining divinatory methods circulated among the users in manuscriptform.12

THE FIRSTMANUALS: UNVEILING THE TECHNIQUE

A second generationof hakkematerial appeared as earlyas the middleof the seventeenthcentury, around the Kanbun era. Contrarily to thefolded books, the firstbound books bearingthe word hakke in theirtitle contain only a fewdia- grams.Rather, they offer detailed presentations of the method. As such,we can assumethey were supposed to be used alongwith a foldedbook, providing the userwith explanations while the orihon served as a basis fordivining. The Shin- sen iríyõhakke narabini shõ publishedaround 1667, is a per- fectexample of this early type of handbook: it gives precise instructions on how to do hakkedivination, as wellas listsof results.By the end ofthe seventeenth century,publishers thought of mixing the content of the folded books withthe explanations,producing top-annotated books (tõchubonSlíÍÝ). The original contentof the folded books is reproducedon thelower part of the page, where

11.According toKalinowski (2003, 231), this "zodiacal" system was already to be seen in Dunhuangmanuscripts, though they never became widely popular. 12.A manuscriptentitled Тогуй hakke kikigaki hiden-shõ froma collec- tionat Hikone Museum (Kindõ bunko isa goodexample. Another one, supposedly datingfrom 1668 (Kanbun 8), can be found inside a broader shugen ШШ manual used by yama- bushiin Echigo province. See Miyake Hitoshi 2007, 163-91. What is more, a manuscript manual in twovolumes from 1693, Shingon himitsu hakke kuden writtenbya Buddhist- relateddiviner named Higashi Rintõ (1693), is now conserved atthe Bibliothèque duCollège de France(Institut des hautes études japonaises). For the most part, it duplicates the contents ofthe earlyprinted manuals, like the Shinsen iriyõ hakke narabini shõ. Judging from that book, we can see thatthe same kind of knowledge circulated inmanuscript andprinted form at the same time, ulti- matelyleading to uniformize thetechnique. The same can be said about a 1708 version ofthe folded bookconserved atthe same place: it is heavily annotated with excerpts from the printed guides.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 348 I JapaneseJournal of Religious Studies 38/2(2011) theupper part is dedicatedto variousannotations explaining the meaning of the text,giving hints, results, or alternativemethods. Puremanuals or hybrid,these books feature lists of results, while progressively cuttingout otherelements, mostly the astrological ones (likethe twenty-seven lunarlodges). Though the bird zodiac is neverto be seenin thesebooks, they are stilllinked in some fashionto Chinesealmanacs. To be moreprecise, while it is difficulttojudge which genre influenced the other, bound hakke books share some similaritieswith the Japanese version of Sansesõ. Thus, the buddhas and bodhisat- tvasassociated which each annualtrigram are sometimes graphically depicted, a featurethat can also be foundin Sansesõ.Some manuals even add Japanesedeities correspondingto thesebuddhas, with variations from edition to edition,such as in theDaikõyaku shinsen hakke-shõ genkai (1718).As opposed tothe folded books, most of these items seem to be targetedat an audienceof "new- comers,"that is, people who do notalready know the technique nor possess a com- plementaryguidebook. Still, early annotated manuals basically retain the essential structureof the first hakke-bon , giving only practical directives and listsof results. In thatsense, although they represent a step further in theexotericization process ofmantic knowledge, they neither put their contents in perspective nor clearly dis- close theirsources. Designed to be used in combinationwith the folded books, theycan be saidto continueto diffusea "medieval" view of divination.

COMPENDIAOF THE EIGHTTRIGRAMS: TRANSMISSIONOF KNOWLEDGE,AUTHORSHIP, AND CRITICAL THINKING

Thirdgeneration books showrather interesting changes, both in termsof form and content.Their evolution can be summedup by pointingto threemain features:a refinementof the core method,a broaderspectrum of techniques, and a tendencyto show more and more of the presenceof the author/compiler.Such an evolutionfollows what can be observedin therealms of Confu- cian scholarshipand literaturearound the same period,with the emergence ofthinkers and authorslike Kaibara Ekken (1630-1714)or Ihara Sai- kaku #)ЖШ®(1642-1693). Along with the explosion of commercial printing, such authorsbegan assertingtheir presence in theirbooks, so thatpublishers, knowingthey would be able to increasetheir sales on theirname alone,were askingthem for new works(Kornicki 1998,227-30). Regardingthe divina- tionmanuals, the first two new featurescontribute to a significantaugmenta- tionof volume, and thethird is accompaniedby an increasein criticalattitude. Startingwith the Kokon hakke taizen (1671),these new manuals usuallytake more liberties with the original content as theypresent themselves as selectionsor compilations.Thus, not only do theydrop the traditional introduc- tionin favor of more detailed, if not always accurate, depictions of the transmission

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 349 ofdivinatory knowledge in Japan,but they also claimto referto preciselyidenti- fiedsources. The complicationof the main method relies on a (re)integration of the"original" properties of the trigrams, to be combinedin orderto formYi Jings sixty-fourhexagrams. The resultsassociated with each manticfunction for each diagramare thusdrawn from the hexagrams obtained by combining the central trigramto the peripheral ones. These new procedures result in a dramaticaugmen- tationin size,since in a fewextreme cases, such as theHakke ( mokuroku) ketteishü Aíh (Chizõin 1697),each diagramrequires a wholevolume. Aside fromthis expansion, which in the end does not seem to have been retainedin laterbooks, probably because it made themanuals quite unwieldy to use and handle,not to sayfar more expensive, these compilations also include variousexplications of othertechniques that may or maynot be relatedto the techniquesdescribed in earliercompilations. Although the Kokon hakke taizen presentsmore divination methods than before, the Wakkan hakke shoshõ taisei ÍÜAiHf (Okamura 1695)and theHakke ketteishü greatly outweigh their predecessorin thatmatter. In thelatter two, we can findarticles about kasõ ЖШ (topomancyapplied to thehouses), which I havechosen to translateas "domognomy,"sigillomancy, han hanji (divinationof personal seals), and even glancesat an arithmomanticsystem attributed to Song-thinkerShao Yong SP Ш (1012-1077).Thus, they give us an accurateoverview of what kind of mantic knowledgewas circulatingat thetime of their compilation. In thisregard, the Wakanhakke is themost precious item of its kind, since it offersa listof thirty- " fivereference works, including hakkebooks producedin our country,"Hoki naiden, alongwith Chinese texts of various horizons, like Song-period buddhist monkMayi s physiognomytreatise, commentaries by the Cheng Ш broth- ers(Confucian thinkers of the twelfth century), or evenrhyming dictionaries.13 Byexposing their sources, the compilers of these books clearly seek a newkind oflegitimization, substituting referenced authority for secret transmission. More oftenthan not, they make no mysteryof their goals. For instance,the compiler ofthe Kokon hakke taizen , presumablysomeone of Buddhist obedience, judging bythe numerous esoteric references he putinto his commentaries,explains that "havingstudied the eight trigrams of the Shingon ÄH and Tendai schools, butalso ofthe astronomers tradition," he "gatheredthe secrets of various traditions,correcting what was bad and selectivelyabridging what was too verbose, whilegiving detailed explanations." Similarly, the compiler of Hakke ketteishü ,

13.The compiler, Okamura Kotõken (1695), gives to his readers no less than forty different references,sometimes inthe form of generic titles (for example, Honchõ ruidai hakke shoshõ Ф №[Books of trigrams ofour country]). Among them we can recognize texts like the onesimply entitled "Mayi s physiognomy," Maininsõ-hõ whichwill play an essential rolein the development ofnew mantic techniques during the Edo period. See Ogawa 1996.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 350 I JapaneseJournal of Religious Studies 38/2(2011) anotherBuddhist literatus of the Tendai school, basically tells us thathe tookthe good and discardedthe bad fromdivinatory knowledge, in orderto presentit to beginners.Likewise the author of the Wakanhakke states in his introductory remarksas wellas in variousparts of the book thathe has correctederroneous contentof other traditions and proposedan introductionto authorizedChinese textsin simplifiedlanguage (using the kana syllabary)to helpbeginners. Criti- cismsof the past, correction of mistakes, and establishmentof authoritative refer- encedknowledge for the uneducated are thus the three main goals the compilers ofthese new manuals assigned to themselves.This tendency comes to itszenith withBaba NobutakeЩЩШ Ä (?-i7i5),a lateseventeenth- century physician and literatus(Hayek 2008). Duringhis life,he publishedsome forty books, most of themdivination manuals, but also popularizationsof Chinese military tales, and evenan introductionto newastronomical theories. Two ofBaba s workscan be regardedas partof the hakke-bongenre, the Shüeki hakke zõshõ shüsei Л§ШЛ (1698) and the Tsühenhakke shinan-sho ## (1703),the secondbeing an expandedversion of the first. In theseas in otherwritings, Baba clearlystates that he intendsto rectifyimproper interpretations and calculation errors,while giving useful explanations and tipsfor beginners, hatsugaku Thuswe can see thatthrough the evolution of trigram books, the efforts made to exotericizedivinatory knowledge go withan aim to homogenizeand ratio- nalize thatknowledge, to purgeit of its"medieval" obscurantism, in orderto modernizeit in all sensesof the word. In thisregard, divination manuals reflect a globaltendency of the eighteenth-century mind. For instance, it is possibleto findquite similar views among literary critics of the same period.The seventh lesson of the Shikashichiron ШШ-кШ, professedin 1703by Ando Tameakiï: (1659-1716),precisely stresses the necessity to "correcterroneous traditions,"as does Ozawa Roan 'ЬуЯШШ(1723-1801) one centurylater in his Furu no nakamichi (Groupe Koten 2009, 26-27,64). All in all, thewill to offera correct(ed)knowledge to guidethe ignorant on theright path is congruentwith the Confucian ideal ofself-cultivation and conformityto thePrin- ciplewhich was broughtto theforeground by early-modern Japanese thinkers.

EarlyModern Diviners Having shownhow trigrambooks werea media of one of the main divinatorymethods of Edo Japan,I shallnow examineto whatextent these manuals, or moreprecisely the kind of knowledge they bear, were in use in premodern Japan.Such a taskmight prove problematic, since it is obviouslynot possible to directlyquestion long-dead practitioners about their activities. However, liter- aryand historicalsources reveal indirectly that these books were indeed widely used amongprofessional diviners, or at leastthey were so perceived.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 351

A KYÕGENSOOTHSAYER

The oldesttrace I havebeen able to isolatepresenting evidence of hakke- related knowledgeis a Kyõgen$£m farce entitled Igui ЯгШ/ШЬi. It featuresa diviner, referredto as sariokiЖШ, who givesa lengthyexample of divination during the courseof the play. I willnot assert that the content of this play should be taken as historicalreality, but simply that the way the fortune-teller is depicted, comic reliefaside, draws on actualobservation by the original creator. Thescript comes in three versions, one for each of the surviving Kyõgen schools, Õkura and IzumifPÄ, and one forthe extinct Sagi Ж school.14The Okura versionis said to be older (end of the sixteenth-beginningof the seventeenth century)than the Izumi, which in returnseems older than the Sagi version.All in all,the three versions present only minor variations, and theplot is identical. To give a briefsummary of the plot: the main character,Igui, has gained the interestof a protector(referred to as master,teishu However,he is ratherupset by the masters attitude toward him. Every time they meet, his pro- tectorfeels it necessaryto slap him on the head. Afteraddressing a prayerto theKannon Ilia worshipedat Kiyomizudera,Igui is granteda hood withthe powerto renderhim invisible. He thenreturns to his mastersdwelling, and just beforethe expected slap, he uses thehood and disappears.Quite conveniently, a fortune-tellerhappens to pass bythe mansion of the puzzled patron at thatvery moment.

D: Fortune-tellingand calculus,fortune-telling and calculus,and rathergood atit... P: Ah ! Herecomes a diviner!For these are such strange events, I might as well giveit a go. D: Indeedvery good at it. . . P: Diviner! D: Areyou addressing me? P: Indeed,I am.Please come inside, I wouldlike to consultyou on a littlemat- ter. D: So,what is thematter? P: SomethingI lost. D: Andwhen was that? P: Justnow. D: Todaywe arethe X month,X day,X year.Tan , ChöKen , Ro, Gin, Nan,Ba , Háku,Dõ , ShitsUyShi , Kõ, En. Place,Reign, Assistance, Death, Detention, Aging.Isn't it something alive? P: Well,as a matterof fact, it is. Howgood you are!

14.See Õtsuka 2006, 436-43; Yoshida 1959, 3-27; and the Sagi version, inYoshikawa and Nonomura1956, 142-51.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 352 I JapaneseJournal of Religious Studies 38/2(2011)

[Thediviner performs divination by throwing sticks] D: Sortingthis out, it gives: one-virtue, six-damage, water, two-righteousness, seven-yang,fire, three-birth, eight-trouble, wood. Metal generates water, thus firstthere is a generativerelation. But because there is also prevalence,since metalprevails over wood, we cannotsee it.However, it is somethingthat has notleft this residence. (Otsuka 2006,438-39)

Thisextract is fromthe Õkura version, and does notgive enough detail to con- firmthe sanoki is indeedportrayed using hakke divination. Nevertheless, I shall pointout thatthe divinationprocess starts with a precisedate (performersof theplay use the currentdate), and seems to involvevarious elements related to thefive phases. The terms"reign," "assistance," and so on referto a cycleof the phases of the year,which can be seen in the Gogyõtaigi as well as other Chinesesources, and werereportedly used duringthe Heian periodas bipar- titeforbiddances, õsõ Biffi.15Essentially, these terms somewhat resemble the jianchu system,16insofar as theyare only depending on theseasons and not of a personsbirth year. The phrasespreceding the enunciation of these phases bythe sanoki can be partiallyfound in theHoki naiden , and are also referenced in TerashimaRyõans famous encyclopedia Wakan sansai zue (the threepowers of Japanand China in picturesand diagrams).According to the latter,these terms pertain to a systemused to determineones inducedsound usinga singlehand. Fifteen"sounds" are distributedamong the phalanges of the hand,and the ten stemsto the fingertips, allowing the divinerto "calcu- late"with his hand (Terashima C1713?,vol. 5, 12thfolio r°). The secondpart of thedivination, called "throwncalculus" in theÕkura version, involves a suite of numbersassociated with a wordand an agent.They can also be foundin theHoki naiden, in an articlededicated to the"nine diagrams," and beforethis in Dunhuangmanuscripts closely related to themethod of theeight trigrams

15.Õsõ here refers to a kindof cycle of auspicious (or inauspicious) elements according to theseasons. There were already several traditions regarding the õsõ during the Heian period, mainlybased on the Wuxing taiyi. Court onmyõdõ seem to have preferred a simplified cycle with onlytwo elements, õ (reign) and sõ (assistance)over the full system, counting eight elements andassociated with the eight trigrams (Frank, 1998, 216-41). However, what can be found in premodernmanuals isanother õsõ system, with five elements connected tothe five phases. This lastone, namely reign õ (BI),aging rõ (3š), detention, qiu (Jp. shu) И, death, si (Jp.shi ) ?E,and assistance,xiang (Jp. sõ) ffi, is strongly linked to divinations concerning illness, and, moreover, militaryprospects. 16.This refers tothe so-called jianchu (Jp. kenjo) system, or,as it is usually to be seen in Japa- nesebooks, the twelve direct relations, jünichoku ~hllü¡L. Itis also a kindof monthly cycle, based onthe rotation ofthe Big Dipper. It starts each month when the direction pointed by the tail of thedipper corresponds tothe branch associated with a givenday. The following days are then consideredauspicious ornot for several types of activities (rites, plantations, andso on).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 353

(Kalinowski 2003,235). This suggests a connectionbetween the sanoki' s action and our books.Most interestingly, the Izumi and Sagiversion gives a ratherdif- ferentlist of manticterms: уйпеп , уйкоп, zettai, and so on, thatis, the mantic functionsdescribed in thehakke books. Moreover,in the same Izumi text, variousstage directions refer to an objectin thediviners possession, an itemhe places in frontof himself, unfolds, and pointsto whileperforming his divination.Unsurprisingly, this object is calledhakke , eight trigrams, and is quiteobvi- ouslya foldedbook.17 Finally, the Sagi versiondraws an even strongerlink by depictinga calculatingsarioki who obtainsas a resultthe trigram da Ã, referred to as Dajõdan й_ЬШг(interrupted in theupper part): this way of calling the tri- gramsis thesame as whatcan be foundin hakkebooks. Therefore,we can assumethat, although the play may have gone through dif- ferentversions before it was writtendown in a finalshape, and whilethe depictionof the divination process might not be entirelyaccurate, the sanoki in this farceis intendedto representan actualuser of the eight trigrams method. I havealready studied the features and characteristicsof these late medieval divinersin anotherarticle, and the conclusions can be summedup as follows:1. the originof the sanoki can be tracedback to divinersof the Kamakura period, spe- cializedin the "way of calculation" ( sandõ ); 2. theywere popular itinerant diviners, althoughsome of them may have had properstands in bigcities, like Kyoto; and 3. theirmain divinatory method method was hakkeuranai , usingfolded "calendar- like"trigram books, along with manuals and guides(Hayek 2010a,20-22). I shall add here thatthe sanokiin thisplay calls himselfTenguzaemon in someversions, where in othershe declareshis methodto be "throwncalculus sticksof the Tengu."Reference to tenguclearly identifies him withthe usual depictionof mountain ascetics, yamabushi liltt, or practitionersseeking magi- cal powersthrough asceticism. These religious specialists are closelyrelated to Buddhisttemples, and theirbeliefs and systemof referenceare highlysyncre- tistic.I have alreadystressed that manuscripts prefiguring trigram books, as wellas thesemanuals themselves, retain a strongBuddhist coloration (Sanskrit scriptures,associated buddhas, and so on). Evenif they draw on textsrelated to onmyõdõ, such as theHoki naiden , thesesources were already tainted with Bud- dhistconcepts. Therefore, there is littledoubt that the sanoki were bearers of a

17.An illustration confirming this hypothesis can be foundin a picturescroll of the Muramochiperiod, the Kumano honji emaki conservedatthe Kumata jinja shrineof Osaka. In this scroll, which tells the story of the divinity ofKumano à la Lifeof Shaka, a peculiarscene shows a diviner perform a divination topredict the future ofa childto be born(who will eventually become the divinity ofKumano). Although the diviner isdescribed in thetext as a "seer,"sõsha orsõnin +BÀ, he is depicted inthe scroll using calculus sticks. At hisside lies a foldedbook, which is quite obviously the same kind of item as the one described in Iguis scenicindications, that is, a hakkebook.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 354 I JapaneseJournal of Religious Studies 38/2 (2011)

kindof knowledge closer to medievalyamabushi than to Abe no Seimei ВД(92i?-ioo5), the famous diviner of the Heian period.

MONKSOR PRIESTS?THE ELUSIVESTATUS OF GENROKUDIVINERS

Tracingthe occurrences of sarioki leads to new"evidence" that early Edo diviners wereindeed using trigram books. The Jinrin kinmõ zui AfrafllüHÄ,an opus of theend ofthe seventeenth century, presents the term as a synonymfor diviner, uranai-shiй№. Thedescription, though very informative about the state of divinatory knowledge duringthe Genrokuera, is not whatwill retain our attentionhere, and a pictureaccompanying the Jinrin kinmõ zui s presentationgives more valuable cluesabout what I am tryingto establish.In thisportrait, a diviner,dressed like a monkwith a shavedhead, sits under a tree.Two clientsare facing him while he pointsto thesticks scattered on thefloor, reminding us ofthe diviner from Igui. In frontof him is an open book witha circulardiagram on one ofits pages (Jin- rinkinmõ zui 1990,118). This schema is verylikely to be a roundversion of the rokujü-zu7' + Bl,"diagram of the sixty," used in hakkedivination to determine someonesinner agent, or theshakuto-hõ no zu faüíí ¿Я, "diagramof the borrowedpath" which is requiredto calculatethe personal trigram of an individual. Thisillustration confirms at leasttwo points. First, Genroku diviners were usu- allyseen as close to Buddhistmonks. Second, they were using books while they performeddivination - probablyhakke books. BabaNobutake provides an interestingstatement regarding the identity of hakke diviners.After attempting to amendthe eight trigrams divination, in Shosetsuben- dan p#ШШШ (hiszuihitsu ffiiř) he criticizesthose who make indiscriminate use ofthis mantic technique. Having stated that "Priestesses ( miko ) and priests(kan- nushi)and so on use theeight trigrams, the nine luminous stars, and thetwelve conductsto tellthe luckiness or the unluckiness of people," he pointsout their lack ofknowledge of the "principles of the Changes" (Baba 1978,48). Baba clearlyidentifies hakke users as "Shintopriests and priestesses,"whose main concernis to determinecauses of diseasesby means of the hakke,nine luminousstars, and twelveconducts. Thus, judging by the evidence, from the sanokiin Iguito thisstatement by Baba, it can be said thathakke divination was used by a wide rangeof specialistswho belongedto one or anotherreligious group,usually with strong connections to Buddhistor Buddhist-relatedfactions. However,judging by the rest of the description of the Jinrin kinmõ zui, they also had otherskills such as sigillomancy,and "domognomy,"which were not coveredby the earliest manuals. Therefore, though we can assumeeight trigram divinationwas stilldominant at thattime, this statement might not standtrue forthe whole Edo period.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 355

WhenKitamura Nobuyo (1783-1856),in hiszuihitsu entitled Kiyü shoran commentson Jinrinkunmõ zui's picture,he statesthat such a portraitis old-fashioned(Kitamura 1979,350-51), which implies that by his timethis kind of diviner, and maybethe divination method itself, was no longer seenas reliable,at leastin Edo city. Did the hakkeuranai fade awayafter the 1750s?It probablydid, in urban contexts,where it was progressivelyreplaced by "newer looking" shineki (a numericalYi Jing-based methodattributed to Shao Yong),darieki ШШз (a hexagram-basedmethod emphasizing more on thecorrelative properties rather than on Yijingitself), and later by new yarrow-stalk techniques (re)created by Japanese authors,such as HirazawaZuitei TSPPtf Ж (1697-1780)and Arai Hakuga fr#â№ (1715-1792).This evolution is congruentwith what can be observedin manuals.As Masukohas alreadyshown, hakke books become more and moreformulaic after themiddle of theeighteenth century (Masuko 2006). Moreover,with the (re) introductionof the nine stars ( kyüsei%Ш) techniquevia "domognomy"books basedon theQing compilationXieji bianfangshu (Jp. Kyõki benhõ-sho) ШУhakke divination was drivento theverge of extinction, or assimilation(Hayek 2010b).The ninestars technique is veryclose to annualtransfer on theeight trigrams,and bothcan be seencoexisting in Dunhuangmanuscripts (Kalinowski 2003,269-81). However, Qing orthodoxy seemingly forgot about this ancient relation,as itdeclared the hakke divination to be a recentdegradation of the nine stars methodand triedto mergethe two techniques. This judgment was notwithout consequenceon thesituation in Japan. Nevertheless,judging from some transcriptsof quarrelsopposing onmyöji (divinersfranchised by the Tsuchimikado family) and divinersof other obedience (Hayashi 1987;1994), some practitionersin theEdo area,lay or religious, werestill using hakke-bon as referencesin 1770,and one couldassume that eight trigramsdivination did survivein thecountryside, even if not in big cities.Not - unlikewhat Yanagita Kunio assertedabout dialects thatsome wordsthat are used at thecultural center of the time of their creation slowly but surely prog- ressto the periphery,where they remain even afterhaving disappeared from thecenter (Yanagita 1930)- it seemsthat trigram divination was stillin use in morerural areas up to thebeginning of the twentieth century. I was able to find thismethod, with the lists, in a manualpartially copied in 1928which appears to havebeen circulatingin a familyof folk religion practitioners of the Izanagi- гуй-Í in theKochi area, along with Meiji reprintsof Edo "domognomy" classics.18

18.1 wish to express my gratitude here to Professor Komatsu Kazuhiko (ircjs) for showing methe manuals he found during his own field work in the area.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 356 I JapaneseJournal of Religious Studies 38/2(2011)

TheEarly-Modern World View as seenthrough Divination Havingpresented the evolution of hakke divination and establishedthe identity ofits users, I shallnow discusswhat can be learnedof Edo mentalityby analysis ofthe particulars of mainstream divination. As I havealready pointed out in the firstpart of this article, the main addition made to trigramdivination during the Edo periodis indisputablythe "list calculation" procedure. At itscore, list calculation ( mokurokusanSii#) is a techniqueby which a fullhexagram can virtuallybe derivedby means of a doubleor triplenumerical conversion.Each trigramin theeight diagrams has an assignednumber, and so does each object.Parameters such as direction,day, and hourare also reduced to primarynumbers. First, the diviner needs to determinethe attributed trigram ofhis client,depending on hisbirth date. Then, by a successionof additions and subtractionsof numbers, he obtainsa numericalresult giving the position of the manticfunction on thediagram corresponding to theattributed trigram. Subse- quentprocedure basically comes in twofashions, a complexone, which tends to disappearfrom books during the eighteenth century, and a simpleone. The for- merconsiders separately the sixty-four possible combinations of trigrams- that is,full hexagrams. The latter yields results that are indexed to manticfunctions. The firstfolded books presentedsome elements,often manuscripts, indicat- inga formof correspondence between mantic functions, objects, and resultswas alreadyestablished prior to theappearance of the second generation of books. Earlyexamples of these lists to someextent mirror what can be foundin Chi- nese texts-they associate eight "purposes" or queriesto each manticfunction. Thiscan be seenin "big"hakke books of the Genroku period that give a complete overviewof each combinationof trigram. Thus, the Hakke ketteishü provides the followingeight categories: rank and remuneration( kanroku ИШ), diseasesand epidemics(shippeil shitsubyõ Ш*М), husband and wife(fusai ^#), domestic- ity( kenzoku#Д), habitation(jüsho f±0f), warehouse ( kozöHÄ), enemyand resentment( onteki ?&Ш), and finallylongevity (jumyõ #¿p). Besidesthese Chi- nese wordswe also findphrases in vernacularJapanese. These categories varied frombook to book,but usually they can be brokendown intosome tendiffer- entqueries: seen things( mimonoЖ%), heardthings ( kikigotoMW), obtained things( emonoШШ), awaitedperson ( machibitoШК, withsometimes a distinc- tion betweenawaited person proper, and intrauterineawaited person tainai machibitolê l*JÍ#À),strange things (keji 11#),lost things (usemono often worksalso forfugitive, hashirimono Ž%), hoped-forthings ( negaigotoЩ#), travels( kadoideP^ffi), transactions ( baibai tcJÏ), trials and judgments( kujizata ^УЙ'ЙС), and dreams(yume W). Thoughmost of thesecategories existed in divinatorymethods long before the Edo period,they became considerably more numerous,detailed, and centeredon individualfate during premodern times.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 357

When embeddedin lists,trigram divination primarily consists in numerical calculation.Upon askingthe clientwhat he wantsto have looked at, the divinerneeds to obtaina numericalparameter. Conveniently, the twelve earthly branches,as a keyelement of the Chinese correlative system, were used to rep- resenttime as wellas space,that is, not only hours, days, months, and yearsbut also directionswere expressed by the branches.In the divinatorymethod in question,these twelve items are yet again associated with numbers from one to eightas follows:rat=6, ox-tiger=7, rabbit=4, dragon-snake=5, horse=3, sheep- monkey=8,rooster=2, dog-boar=i. Anothernumber is attachedto each divinationobject, hence the word nimokurokuZlBiS, "doubleentry list."19 Having obtained the two numbers, the divineradds themand thensubtracts eight from the resultif it is greater thaneight. The finalnumber indicates which mantic function should be used to producethe result of the divination. The soothsayerthen refers to thelist correspondingto thisfunction. Of course,various other numerical operations or factors,based on thepersonal agent or theclient, for example, might be invoked to refinethe judgment. To makethe process easier to understand,I shalltry to simulatea divinationby list. Lets assumethat we arein Genroku11 (1698), older earth-tiger. A thirty-two- year-oldman (in the traditionalcounting fashion, kazoedoshi îfcx.^),20 asks aboutsomething he saw at thehour of the dragon (between eight and tenin the morning).First, we needto determinehis attributedtrigram, toke. Being thirty- twonow, he was bornin Kanbun7 (1667),yin fire-sheep. He thereforebelongs to thesuperior monad and to a groupcalled to.We thencheck the diagram of the borrowedpath. It indicatesthat we should startfrom the Ri Ш trigram, and countcounterclockwise while skipping or countingtwice several trigrams accordingto the rules.21In the end, we obtainRi as the tõkeof thatperson. Besides,according to thegraphic of the sixty (induced sounds), the agent of that

19.Another popular method, often added near the end of second or third generation hakke books,involved four different items/numbers, andwas hence called "quadruple list," shimokuroku И SÜ. 20.In the traditional system, one is already one year old at birth, and gains one year every newyear, without regard for the actual birth date. 21.The skipping and jumping ofthe trigram, orkoyuru-odoru implies that starting from thetrigram corresponding toa given subperiod and sex, we should skip (tobikoyuru ) the first 8thtrigram, jump (count twice, odoru ) on the 41st, skip the 48th, jump on the 81st, skip the 88th, andjump on the 101st. These rules are usually given with several ways to make the calculation easier.For example, early manuals suggest tocount the years past twenty ten by ten or to abridge thefirst steps by counting 10on the second trigram, and so on. For instance, starting from the Ri $1trigram, one should skip Son Ü, count10 on Shin M, skip Gon M, count 20 on Kan ifc, and so on,while "jumping" when necessary.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 358 I JapaneseJournal of Religious Studies 38/2(2011)

man is fire(over the sky). Next, we haveto convertthe parameters (hour of the day) intonumbers as follows: In ourcase, we retain5, the number corresponding to thehour of the dragon. The "seenthings" category is associatedwith the number 3.22 By adding the two values,we obtain8. Since thereis no need forsubtraction, we thencheck the properdiagram in orderto knowthe applicable mantic function. If Ri is thecen- traltrigram, the eighth position is heldby the trigram Kon #, withthe function yukon, soul transfer.Here are the results given by Baba (1698):statue of Buddha (honzonФ#), somethingadamant, nut, bone, quadruped, glittering thing from thesky, something unstable on itsfeet, wife. To knowif the vision was a good or a bad omen,we haveto matchthe agent of the consultant (fire) to thatof the function(earth in thiscase). Sincefire generates earth, we concludethe encoun- terwas auspicious.Even if these results are sometimesobscure, they are seldom redundant(for a givencategory) or contradictory.On thecontrary, they often proveto be surprisinglyprecise, especially when they give directions or places. As one can see throughthis brief introduction to themethod, its core nature can be summedup by a processof numericalencoding of theworld through a correlativesystem in orderto put the differentfactors in a simpleequation, followedby a decodingof the results in thesame way.Such a structureimplies thatdiviners not onlyhad to assimilatethe correlative system written down in books,but they also had to be skilledin basic arithmetic.They had to be above averagein termsof knowledge and education.Furthermore, given the relation betweena divinerand his client,involving a doubletransaction- monetary on the one hand,and cognitiveon the other- the encodingstructure must have been explainedin layman'sterms to theclients, informing their way to see the worldand themselves.Moreover, the very categories of hakke divination are to be takenas an expressionof the clients expectations and concerns. As faras listcalculation is concerned,it seemsprobable that the preoccupa- tionsof Edo commonersmainly revolved around three matters: the outcomes of ongoingactions or events(trials, requests, illness, and so forth),the conse- quencesof things a clienthad experiencedor wouldexperience (strange events, encounters,verbal exchanges, travels, and so forth),and the whereaboutsof

22.The fact that the number 3 is associated with the item "things seen" is highly relevant to howa mantictechnique from China was transformed tofit in with Japans own correlative thinking.The Japanese reading for 3 ismi(tsu), and so is the one for "to see," mi(ru), hence the connec- tion.There is no way such a correlationcould have preexisted inthe Japanese adaptation ofthe hakkemethod. This kind of word play is crucial to understanding howJapanese diviners, aswell astheir clients, were able to link together phenomena which would seem totally unrelated. Such analogicalthinking, deeply related to the very nature of the Japanese language, serves as a basis forother folk beliefs aside divination. Homophonie taboos, for instance, widely rely on similar associations(the most well known being the relation between the number 4,shi, and death, shi).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 359 thingsand personsthat were beyond the range of the clients vision or understanding(lost things, robbers, expected person, supernatural influence, and so forth).All ofthese have in commonthat they target something that has yetto actuallyhappen. Therefore, it appearsthat trigram divination in itslist calcula- tionmode primarilyconcerns phenomena which are stillin therealm of poten- tiality.One mightobject that this is trueof any kind of divination, but this would be omittingthe explanatory role divination can play:for example, unveiling the supernaturalcause of an illnessin orderto findthe most appropriate remedy used to be a crucialrole of court diviners in theHeian period(Hayek 2005). Whatis more,it must be notedthat most of the items in thelists are centered on individualfate, and theoutcomes of the processes are depictedfrom the client'spoint of view Thisis particularlytrue for the expected person category, an itemthat was alreadyin Heian manualsand stillappears today on paperomens (omikuji&ШШ). As opposedto theChinese word for this object, usually xingren ÍTÀ (personwho is going),the Japanese machibito , focuses on theindividual who stayedbehind, a Penelopeexpecting the arrivalof herUlysses who went away,where "away" appears as an utterlydifferent and unknowablespace.23 Nonetheless,results of divinationregarding these objects strike the viewer by theirlevel of precision(directions, places), thanksto the expansivenessof thecorrelative system and thelow redundancyfor any given category. Such a specificityshould be consideredin associationwith two other properties of the technique:first, the rather linear character of the process; and second,the fix- ityof variables (time, objects, and so on). Thesevariables are transformedin numericaldata integratedvia an invariantformula (addition then subtraction), so thatwith the same parametersone willalways obtain the same results,and thoseresults do notdepend on an interpretationbut rather are to be foundon a preexistinglist. All ofthis points out the utterly predetermined structure of this manticmethod. Thisis notto saythat Edo commonerswere living under the pressure of an unavoidable,preset fate, since there are indeedexceptions and limitsto prede- terminedresults. Illness, for instance, is clearlythe most ambiguous object of divination.Some manuals simplyavoid includingdiseases among the items oflisted results, while others give up to six parametersto a listcalculation for

23.This perception ofthe outside of a givensocial group/space as "another world" echoes KomatsuKazuhiko's views on the status of the "stranger" ijinЛ À inJapanese folklore. The ijin, a kindof absolute "other," when arriving ina grouphe does not pertain to, coming from "another world"might be taken as a scapegoatand murdered for no other reasons then his "otherness." Themachibito here is in some way the reflection ofthe ijin : to the group where the ijin arrives, he isa strangercoming from the "outside" and therefore carrying various "negative," inauspicious omenpotentially harmful tothe community. However, for the people he left behind, he is some- onewho went to the unfriendly "outside" and who they anxiously long for (Komatsu 1995)-

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions Збо I JapaneseJournal of Religious Studies 38/2(2011)

illness,or clearlywarn the readerto make the appropriateverifications (that is, to double-checkthe resultvia anothermethod) before stating the resultof divination.Therefore, though there is a strongtendency for both parties to the divinatorytransaction (diviner/client) to prefer predetermined results, there are severalarrangements made to circumvent the intolerability ofa totallypreset fate. All in all,I wouldsay that predetermined results themselves express a formof systemizedthinking, insofar as theyrelate to a categorizationof the phenomenal world,and also a formof rational thinking: even if the results are predetermined, therelation between them and their causes, for example original data transformed and encodedthrough mantic parameters and functions,follows a wellorganized formula.In thatregard, we can tryto findsome sortof symmetrybetween the systematiccharacter of these relationships and thesocial organization of the Edo period.In thehighly hierarchized and relativelystatic Edo society,with a central power(the bakufu) trying to categorizepeople in a rigidand easilycontrollable classification,the clients of the diviners were seeking systemized relations of con- sequences,with the same causes inevitably producing the same effects.

Conclusion: Divination , Beliefs , Superstitions , and TheirSociocultural Context Divination,as a culturalfeature bound by social and historicalborders, tells us a greatdeal aboutits users. In thisbrief overview, I have tried to shedlight on a largelyunknown, though essential, type of mantic technique. Hakke divination, as I observed,began to fadeaway at theend ofthe Edo period(the first half of thenineteenth century). It was eithersupplanted by newly- formed techniques, likethe abridged yarrow stalks method of Hirazawa Zuitei and AraiHakuga, or else absorbed,as was itsChinese avatar, in a newform of "nine stars" divination. Overthe more than two and halfcenturies of Tokugawa rule, it had thetime to be widelypopularized, not onlyby the diviners,but also by householdency- clopedia ( setsuyõ-shuШШШ, chõhõkiÄSSfi) and almanacs ( õzassho^cît#) directedtoward readers of commoner status (Hayashi and Koike 2002,193-94). Althoughit declined in thelast decades of the Edo period,its influence is still palpablein variousterms remaining in contemporaryJapanese. The expression in fourcharacters used to designan inextricablesituation, zettai zetsumei WlW lÈíp, forinstance, though mistakenly linked to "ninestars" divination in most dictionaries,comes directly from the hakke-bon. This is also truefor the expression honkegaeriÝíbSW, whichmarks the fact that someone has turnedsixty yearsold and is sometimesused in place ofa morecommon term, kanreki (literally,complete calendar loop). Honkehere designs the main trigram, that is, theattributed trigram afferent to thebirth year. However,the exact meaning of these expressions is almost,if not totally, lost to commonunderstanding. Correlative schemes that were once used routinelyby

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 361

Japanese(and Chinese)to definetheir place in theuniverse and to obtainsome insightregarding their personal interactions have gone out of use. Thoughthe practiceof divination based on theanalogico -symbolical network is stillalive, the intricaterelation between that system and a givenindividual cannot compare to whatit used to be. As I haveshown in thisarticle, one ofthe key features sustaining most mantic proceduresof Chinese originsinvolving fixed variables is the capacityto integrateany phenomenon into a mantic"formula" via a correlativeencoding. Trigramdivination is no exceptionto thisfeature. Directions and datesshare thesame symbols,and so do humanbeings, mainly through their birth dates. More oftenthan not, the main purposeof such a divinatorysystem is to provide,in additionto a glanceat thefuture, simple categories whose interactions can be predictedthrough a correlativepreset. Ultimately, the clef de voûteof the integrationof individuals into the system is thesexagesimal calendar, hence the predominanceof the subperiodcalculation in hakkebooks. The suppression ofthis calendar interface at thebeginning of the Meiji era and itsreplacement bya "new"(Western) calendar, lacking such correlative properties (at leastto a Japaneseaudience), inevitably led to a looseningof the bonds betweenpeople and their"traditional" world view. This might have been done on purpose:Meiji reformersdid wantto putan end to whatthey labeled as "superstitions"( meishin âÊfit).What this word meant in themind of the reformers could be subjectto discussion,but fromthe point of view of sociologyof knowledge,a supersti- tionis a formof belief in an arbitrarycausal relation,that is to say,a relationof thetype [if A, thenB] lackingany explicative premise (Bronner 2003). Going back to theshift of calendar, we can saythat by cuttingout thecorrelative network,which played the role of a cognitiveinterface between divinatory param- etersand divinatoryresults, the reformers actually contributed to thecreation of superstitions,in thesense that all thehemerological lore, having lost its prem- ises,became a mereset of arbitrary beliefs.

REFERENCES PRIMARYSOURCES: TRIGRAM BOOKS

Arai Gensetsuff C1683 Hakkekogagami 2 vols.,1 fase. Held at the National Diet Library (ShiniõBunko Baba NobutakeJSiJHfÄ 1698 Shüekihakke zõhon-shõ И^АйШсФЁк 3 vols.,1 fase.Kyoto: Rakka Shorinř&T##, Kyõraiseki Held at theNational Diet Library (ShinjõBunko).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 362 I JapaneseJournal of Religious Studies 38/2(2011)

1703 Tsühenhakke shöchü shinan 5 vols.,1 fase. Kyoto: Rakka Shorin,Kyõraiji Heldat the International Research Center for Jap- aneseStudies, Kyoto. Chizõin 1697 Hakke( mokuroku) ketteishü A#(§§i)$:/EÄ. 10 vols.,5 fases.Kyoto: NakanoKyùemon ŤiřAíííwFI, ZeniyaShõhei íSSJEÉzŘífi. Held at the NationalDiet Library (Shinjõ Bunko). Fujiwara MitsutoyoШШтШ 1844 Gotõkegokefu kantö ippa shinke tõyõsho fr.Manuscript. Ishiyama-ke monjo 5 lil Li-7-1439.Tokorozawa Cityarchives. Higashi RintõЖ PŽS 1693 Shingonhimitsu hakke kuden AíhPíž. Manuscript,2 fases. Col- lègede France,Institut des Hautes Etudes Japonaises. HiMENTEiOnomaro 1709 HakkeHõrai-shõ 5 vols.,1 fase.Asano Yabei Osaka:Seibundõ JÜá L Heldat Tohoku University (Kanõ BunkoШЩ

Io TõsenШШЖ)'' and UtagawaKuniaki ШПВШ (ill.) 1864 Hakkekyüsei zõho tsuji ura dokuhandan 2 vols., 2 fases.Sakai: Gusokuya Jůbei Held at theInternational ResearchCenter for Tat>anese Studies. Kannonji Fudõinffiia Gon-shõsõzu Chõgyõ C1637 Töryühakke kikigaki hiden-shõ Manuscript.Kindõ Bunko HikoneCastle Museum/National Institute of Japanese Literature. Okamura Kotõken 1695 Wakanhakke shoshõ taisei fPřUAíhlf 8 vols.,4 fases.Kyoto: Kosaji Hanemon Osaka:Furubonya Seizaemon iÈ^SÍptóiíírP^; Edo: FujiyaSaburõbei ШМНЁКйй&г. Held at Kyoto University library. Seishõ Ж® 1708 HakkeAìK Foldedbook, 1 strip.Published in Kyoto,publisher unknown. Held at the Collègede France,Institut des hautesEtudes Japonaises, Paris.

PRIMARYSOURCES: UNDATED

HakkeA# n.d. Foldedbook. 1 strip.Held at the National Diet Library (Shinjõ Bunko, no 854-172).

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 363

Hakke-bonAS№ n.d. Foldedbooks. Held at theNational Diet Library(Shinjõ Bunko, nos. # 2-690- #2-696). Hakke-bon n.d. Late sixteenth-centurymanuscript. Held at theNational Diet Library (ShinjõBunko nos. if 2-688). Hakke-bon n.d. Foldedbook, 1 strip.Held at theInternational Research Center for Japa- neseStudies. Hakke-bon n.d. Foldedbook, 1 strip.Held at Shidõ Bunko, Keiõ Gijuku University. Hakkeno sho n.d. Foldedbook, 1 strip.Held at Tohoku University (Kanõ Bunko). Õzasshosanzesõ n.d. 1 fase.Edo: YoshidayaBunzaburõ pí BBIllíHÊIS. Held at the International ResearchCenter for Japanese Studies. Sanzesõkokagami H IMS/hü n.d. Suda КЕН,place of publication unknown. 1 fase.Held at the International ResearchCenter for Japanese Studies. Shintöhandan-sho Ш n.d. 1 fase.Property of the late Takase Kiyonaga МШШжЪ family. Held by the Takasefamily, Kochi prefecture. PRIMARYSOURCES WITH NO AUTHOR

Daikõyakushinsen hakke-shõ genkai 1718 2 vols.,1 fase.Publishers: Kitao Hachibei AÄfiS(place of publication unknown);Mõrita Shõtarõ ^f'JEHJEÈ^filS, Osaka. Held at the International ResearchCenter for Japanese Studies. HakkeAíh 1708 Foldedbook, 1 strip. Held at the National Diet Library (Shinjõ Bunko, nos. #2-689). Hakke-bonAí№ 1640 Folded book, 1 strip.Sahei Kyoto.Held at the NationalDiet Library. Hakketora no makiAihife^L^ 1680 3 fases.Kagiya §Ш, Kyoto.Held at theNational Diet Library(Shinjõ Bunko). Hakkezue AihBI# 1611 Foldedbook, 1 strip.Held at Tõyõ Bunko, Tokyo.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 364 I JapaneseJournal of Religious Studies 38/2(2011)

Hidenmyõhõ-ki 1703 i fase.Sudo Genbei ШШШ&Ш, Edo;Yamaguchi Mohei Ш П StÄÄr,Kyoto. Heldat the International Research Center for Japanese Studies. Jinrinkinmõ zui ЛШШШИЙ 1990 Annotatedby Asakura Haruhiko TõyõBunko 519. Tokyo: Hei- Bonsha. Kokonhakke taizen 1671 6 vols.,6 fases.Nakano Sõzaemon PI, Kyoto.Held at the NationalDiet Library (Shinjõ Bunko). Nihonkokugo daijiten 2000-2002 Secondedition. 13 volumes. Tokyo: Shõgakukan. Onmyõhakke no hõ 1628 Foldedbook, 1 strip. Published in Kyoto. Held at Tohoku University (Kanö Bunko). Onmyõ-sho( kadai ) Й1Шг1г(®11) C1528 Manuscript,1 fase. Folded book, 1 strip.Held at the National Observatory ofJapan, Mita. Õzasshosanzesõ 1852 1 fase.Publishers: Izumiya Ichibei ÄSTpÄlii, YamaguchiyaTõbei lil PS UzŘÍÍT,Kikuya Kõsaburõ ÜJÜ#HÍ!|$, Moriya Jibei ШёГрЙ®, Fujiokaya Keijirõ ® ММв?пЁ1$,Yamazakiya Seishichi ШЙМЖ-Ь, Yoshidaya BunzaburõíÉí EHSiHßß, Edo. Heldat the International Research Center forJapanese Studies. Ruijuhakke-shõ ШЗ^А^Ш 1672 2 vols.,2 fases.Held at the International Research Center for Japanese Stu- dies. Shinbanhakke früSAíh C.1680Folded book, 1 strip.Published in Kyoto,publisher unknown. Held at the Collègede France,Institut des Hautes Etudes Japonaises, Paris. Shinbanseimei hakke SMSflf ВДЛ# c.1680Held at the National Museum of Japanese History, Sakura City, Chiba prefecture. Shinseninyõ hakke narabini shõ 1667 2 vols.,1 fase.Held at theCollège de France,Institut des HautesEtudes Japonaises,Paris. Shinseninyõ hakke narabini shõ 1836 2 vols.,1 fase.Hanabusa Daisuke Edo: Mankyùdô75 Heldat theInternational Research Center for Japanese Studies.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 365

Tenmonhakke hidenshõ 1685 2 vols.,2 fases.Nakamura Sonbei Kyoto.Held at the National DietLibrary (Shinjõ Bunko). Nihonkotenseki sõgõ mokuroku. Available online at http://basei.nijl.ac.jp/~tkoten/about.html PRIMARYSOURCES

Baba NobutakeЩЩШЖ 1978 Shosetsubendan ft ШШ №. In Nihonzuihitsu shüsei dai nishü ВФРдИЩЙс #C$?, 23-71.Tokyo: Kyùko Sho'in. (Originally published 1715) Kitamura Nobuyolï#4tiHêp 1979 Кхуйshõran vol.3. In Nihonzuihitsu taisei ВФШЖ^Йс, annex 9. Tokyo:Yoshikawa Kõbunkan. (Originally published 1830) Ötsuka Mitsunobu , ed. 1998 Evora-bonnippojisho If = Vocabvlarioda lingoade Iapam. Osaka:Seibundõ Shuppan. 2006 ÕkuraToraakira nõ kyõgen shü mЖ. Vol.2, Osaka: Seibundõ Shuppan. Terashima Ryõan n.d.(C1713?) Wakan sansai zue 81vols. Osaka: Onoki Ichibë ií^TfíÂfêf.Held at theCollège de France,Institut des HautesEtudes Japonaises,Paris. Yoshida Kõichi ed. 1959 Izumi-ryûkyôgen-shù ЙАЖЯНЖ. KotenBunko 143,vol. 14. Tokyo:Koten Bunko. YoshikawaHisashi дЛ IX, ed.,and Nonomura Kaizõ (commentary) 1956 Kyögen-shü(ge) fit Ш (T) . Tokyo:Asahi Shinbunsha. Yuan TiangangÍC 1933 Sõkanbonenkin tosü sansesõ-sho Tokyo:Nihon ShoshiGakkai. Yun Lu ЛШ, ed. 1995 Qindingxieji bianfang shu wai ba chong АШ. Zhenben shushucongshu ШШШ, vols.65-66. Taipei:Shin Wen Feng Print Co. (Originallypublished 1741) secondary sources

Aston, WilliamGeorge 1908 A Japanesebook of divination. Man 8: 116-20.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 366 I JapaneseJournal of Religious Studies 38/2(2011)

Blacker, Carmen 1975 TheCatalpa Bow : A Studyof Shamanistic Practices in Japan. London: Allen & Unwin. 2000 Divinationand oraclesin Japan.In CollectedWritings ofCarmen Blacker , 65-84.Abingdon, Oxford: Routledge. Bronner,Gérald 2003 L'Empiredes croyances. Paris: Presses Universitaires de France. Frank,Bernard 1998 Kata-imiet kata-tagae: Etude sur les interditsde directionà l epoquede Heian.Paris: Collège de France,Institut des HautesEtudes Japonaises. (Originallypublished by the Maison Franco-Japonaise, Tokyo, 1958) Groupe Koten, ed. and trans. 2009 RegardsCritiques: Quatre reflexions sur la literatureclassique dans le Japon des XVIIe-XVIIIesiècles. Paris: Collège de France,Institut des Hautes EtudesJaponaises. Hayashi Makoto# fë 1987 "Shiryõ"kinsei onmyõji no katsudõto soshiki:Wakasugi kyûzô no - ichishiryõno shôkaiTÄfi-J ïr#IB AichiGakuin Daigaku Bungaku-bu Kiyö 17: 211-193. 1994 Tokugawa-perioddisputes between Shugen organizations and Onmyõji overrights to practicedivination. Japanese Journal of Religious Studies 21: 167-89. 2006 Kinsei onmyõdõno kenkyü Ж<Г>ЩЗъ. Tokyo: Yoshikawa Kõbunkan. Hayashi Makoto# Ш and Koike Junichi eds. 2002 Onmyõdõno kõgi Tokyo:Sagano Sho'in. Hayek,Matthias 2005 Chüseisetsuwa ni okerubyõin to onmyõjino yakuwari Asiayûgaku79: 170-86. Tokyo: Bensei Shuppan. 2008 Edo jidai no uranai-bon:Baba Nobutakewo chùshinni - Ф . In Nihonbunka no jinruigaku/Ibunka no min- zokugaku( komatsu kazuhiko kanreki kinen ronbun -shu) , 324-45.Kyoto: Hõzõkan. 2010a Sanoki-kõ:Chüsei kara kinsei shoki made no uranaishino jittaio sagutte - - #Ül# . Kyotominzoku 27: 1-26. 2010b L'artde jugerde l'aspectdes demeures de MatsuuraKinkaku (début XIXe): inventerune divination« traditionelle»aux accentsmodernes. In Japon Pluriel8, collectif et sous la directionde NorikoBerlinguez-Kõno et de BernardThomann, 316-28. Arles: Editions Philippe Picquier.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions HAYEK:JAPANESE EARLY MODERN DIVINATION | 367

Kalinowski,Marc 1985 La transmissiondu dispositifdes neufpalais sous les six dynasties.In Tantricand Taoiststudies in honour ofR.A. Stein vol 3.,ed. MichelStrick- mann,774-811. Bruxelles: Institut belge des hautes études chinoises. 1991 Cosmologieet divination dans la Chineancienne: le Compendiumdes cinq agents(Wuxing dayi , VIesiècle). Paris: École française d'Extrême-Orient, Adrien-Maisonneuve. Kalinowski,Marc, et al.,ed. 2003 Divinationet sociétédans la Chinemédievale: étude des manuscritsde Dunhuangde la Bibliothèquenationale de Franceet de la BritishLibrary. Paris:Bibliothèque Nationale de France. Kawasaki Rie JlIflîSS 2005 Shokiõzassho ni kansurushoshiteki kõsatsu: Karíei kara kei'an yonen made Kyotojoshi daigakudaigakuin bungaku кепкуйка kenkyúkiyõ 4: 1-32. 2008 Õzasshoni kansurushoshiteki kõsatsu: Keian kara genroku yonen made Kyõtojoshi daigaku daigakuinbungaku кепкуйка kenkyúkiyõ 5: 1-34. 2008 Õssashoni kansurushoshiteki kõsatsu: fõkyõ kara hõei-ki made Щ~$~Ъ Kyõtojoshi daigaku daigakuin bun- gakuкепкуйка kenkyúkiyõ 6: 1-28. Komatsu Kazuhiko'htMUfë - !995 Ijinron:Minzoku shakai no shinseiUÀ Im Tokyo:Chi- kumaShobõ. 2006 Minzokushakai no nakano"onmyõji" no sonzaikeitai, saikõ: Kõchiken monobemurano "izanagi-ryûtayû" no baai Hikakunihon bunkakenkyü 10: 54-76. Kornicki,Peter 1998 TheBook in Japan:A CulturalHistory from the Beginnings to theNine- teenthCentury Leiden; Boston: Brill. Masuko Masaru Ш (aka NarabaMasaru) 2006 Edo jidaino hakkeuranai to hanponni tsuite ovvf . Tõhõshükyö 107: 42-55. 2010 Kinseiekigaku кепкуй Tokyo:Ofù. MiyakéHitoshi ed. 2007 Kinseishugendõ monjo Tokyo:Kashiwa Shobõ. MurayamaShüichi tí lil 1981 Nihon onmyõdõ-shiпо кепкуй Tokyo:Hanawa Shobõ.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions 368 I JapaneseJournal of Religious Studies 38/2 (2011)

MurayamaShûichi et al.,eds. 1991-1993 Onmyõdõsõsho l^l^illlïr, vols.1-4. Tokyo: Meicho Shuppan. Nakamura Shõhachi A 2000 Nihononmyõdõ-sho no кепкуй Tokyo:Kyùko Shoin. (Originallypublished 1985) Ogawa Yõichi/WllUr- 1996 Ninsõ,teso uranai Sinica( Gekkanshinika ЯйРИЫ-^), vol. 7: 36-37. Severini,Antelmo 1874 Notiziede astrologiagiapponese. Geneva: H. Georg. Shin HyeonФ К 2002 Kanekijussü-ron кепкуй. Tokyo: Kyùko Shoin Saitõ Hideki 2007 Onmyõdõno kamigami Kyoto:Shibunkaku Shuppan. Suzuki Yoshijirõ |5 1963 Kanekiкепкуй ЩШМЗъ. Tokyo: Meitoku Shuppansha. UmenoChihiro ífeSřŤ# 2003 Onmyõdõhonsho tsuchimikado-ke no soshikitenkai Nihonshiкепкуй 487: 138-60. 2009 Kinsei onmyõdõsoshiki no кепкуй Tokyo: YoshikawaKõbunkan. Yanagita KuniofíPEHSI? 1930 Kagyü-köÍ041#. Tõkyõ: Tõkõ Shoin. Yano MichioШЖ № 1986 MikkyõSenseijutsu : Sukuyõdõ to Indo Senseijutsu Ж ¿ЛШт. Tõkyõ:Tokyo Bijutsu. Zeitlyn,David 2001 Findingmeaning in thetext: The process of interpretation in text-based divination.The Journal of the Royal Anthropological Institute 7:225-40.

This content downloaded from 132.203.227.61 on Thu, 22 Oct 2015 08:53:39 UTC All use subject to JSTOR Terms and Conditions