References and Notes1. H. J. Muller and E. Altenburg, Genetics 15, 283

(1930).2. C. P. Oliver, Science 71, 44 (1930); Z. Indukt.

Abstammun. Verebungsl. 61, 447 (1932); M.Demerec, Cytologia (Fujii Jubilee Volume)(1937),p. 1125.

3. I. H. Herskowitz,Am. Nat. 80, 588 (1946).4. M. Demerec and U. Fano, Proc. Natl. Acad.

Sci. U.S.A. 27, 24 (1941).5. D. E. Lea and D. G. Catcheside, J. Genet. 47,

10 (1945); J. B. S. Haldane and D. E. Lea, ibid.48, 1 (1947).

6. D. G. Catcheside, Adv. Genet. 2, 271 (1948).7. D. E. Lea, Actions ofRadiations on Living Cells

(Cambridge University Press, London, 1946).8. U. Fano, Science 106, 87 (1947).9. C. W. Edington, Genetics 41, 814 (1956).

10. I. H. Herskowitz, ibid. 36, 356 (1951).11. H. J. Muller and C. A. Offermann, unpublished

data, cited by H. J. Muller, J. Genet. (1940), p.401.

12. E. Novitski, Drosophila Inf. Serv. 26, 115(1952).

13. H. J. Muller, E. Carlson, A. Schalet, Genetics46, 213 (1961).

14. F. W. Stahl, The Mechanics of Inheritance(Prentice-Hall, Englewood Cliffs, N.J., ed. 2,1969).

15. E. Novitski, J. Cohen, C. Laird, Arch. JuliusKlaus Stift. Vererbungsforsch. Sozialanthropol.Rassenhyg. 37, 431 (1962).

16. G. E. Brosseau, Drosophila Inf. Serv. 40, 71(1965).

17. This is an important point. I. I. Oster reports nosuch mosaics [see F. H. Sobels, Ed., Repairfrom Genetic Radiation Damage and Differ-ential Radiosensitivity in Germ Cells (Macmil-lan, New York, 1963), pp. 179-204] after irradiat-ing a similar chromosome, but his method ofdetection depended on finding fertile F, males,

which would necessarily eliminate those caseswith a broken or fragment Y chromosome lack-ing fertility factors.

18. E. Novitski, Drosophila Inf. Serv. 38, 71 (1963).19. B. Leigh and F. H. Sobels, Mutat. Res. 10, 475

(1970); F. H. Sobels and B. Leigh, ibid. 12, 100(1971).

20. H. Bauer, Naturwissenschaften 27, 821 (1939).21. K. Burki, Arch. Genet. 47, 105 (1974).22. K. Illmensee and A. P. Mahowald, Proc. Natl.

Acad. Sci. U.S.A. 71, 1016 (1974).23. E. A. Carlson and E. Novitski, Genetics 50, 239

(1964).24. E. Inagaki and Y. Nakao, Mutat. Res. 3, 268

(1966).25. This article was written while I was the recipient

of a John Simon Guggenheim Foundation Fel-lowship in residence at the University of Lei-den. Help from Prof. F. Sobels, A. T. Natara-jan, B. Leigh, J. H. Postlethwaite, and K. San-karanarayanan is gratefully acknowledged.

In a recent article in Science (1) it wassuggested that the residents of SalemVillage, Massachusetts, who in 1692charged some of their neighbors withwitchcraft did so because of delusionsresulting from convulsive ergotism. Theauthor of the article, L. R. Caporael,argued that (i) the general features of theSalem crisis corresponded to the, fea-tures of an epidemic of convulsive ergot-ism, (ii) symptoms manifested by thegirls who were the principal accuserswere those of ergot poisoning, (iii) thesymptoms shown by other accusing wit-nesses were also those of convulsive er-gotism, and (iv) the abrupt ending of theSalem crisis suggests ergot poisoning.We shall attempt to show that these argu-ments are not well founded.

Features of Convulsive

Ergotism Epidemics

Ergot is a fungus (Claviceps purpurea)that under some conditions infests ryeand other cereal grains. When ingestedthe ergotized grain may produce a vari-ety of cardiovascular effects leading,

1390

among other things, to gangrene (gangre-nous ergotism), or neurological effectsleading, among other things, to con-vulsions (convulsive ergotism) (2-5). Epi-demics of convulsive ergotism have anumber of general features that differsubstantially from the events that oc-curred in Salem.

According to Barger (2), epidemics ofconvulsive ergotism have occurred al-most exclusively in locales where theinhabitants suffered severe vitamin Adeficiencies. Ergot poisoning in individ-uals with adequate vitamin A intakesleads to gangrenous rather than con-vulsive symptoms. Vitamin A is foundboth in fish and in dairy products. SalemVillage was a farming community andSalem Town, which bordered the village,was a well-known seaport; cows and fishwere plentiful. There is no evidence tosuggest a vitamin A deficiency in the dietof the inhabitants, and it would be partic-ularly unlikely for the so-called "af-flicted girls," some of whom came fromwell-to-do farming families. The absenceof any instance of gangrenous symptom-atology makes it highly unlikely that er-got played any role in the Salem crisis.

Young children are particularly sus-ceptible to convulsive ergotism. Bargerstates (2, p. 39):

All accounts of convulsive ergotism agree thatchildren were more liable to convulsive ergot-ism than adults; thus 56 percent in the Finnishepidemic were under 10 years of age; 60 per-cent of Scrinc's cases were under 15 years ofage....

Only 3 of the 11 afflicted girls at Salemwere under 15 years of age and only oneof those was under 10 (6, p. 57). There isno evidence either in the trial records orin eyewitness accounts to indicate a highrate of convulsive symptoms in theyoung children bf Salem Village duringthe witch crisis (7-11). In fact we couldfind references to only two cases of con-vulsions in children under ten during theperiod of the crisis. One of these was theafflicted girl mentioned above. The otherwas an 8-week-old infant that convulsedbefore it died (11, vol. 1, p. 95). An 8-week-old infant would not yet have beenweaned, and nursing infants do not suf-fer ergot poisoning even if their mothershave a very severe case of the disease (2,p. 38); it is therefore unlikely that thisinfant died from ergotism.The fact that most of the individuals

(including young children) living in thesame households as the afflicted girlsshowed no symptoms is attributed byCaporael to wide individual differencesin susceptibility to ergot poisoning.While there are wide individual differ-ences in susceptibility to gangrenous er-gotism, convulsive ergotism is anothermatter. According to Barger it was com-mon for all members of a family to devel-op symptoms of convulsive ergotism dur-ing epidemics (2, p. 27). This tendencywas so pronounced that convulsive ergot-ism was long (but erroneously) thoughtto be infectious.

Dr. Spanos is an assistant professor and Mr. Gott-lieb a doctoral candidate in the Department of Psy-chology, Carleton University, Ottawa, Ontario, Can-ada K IS 5B6.

SCIENCE, VOL. 194

Ergotism and the Salem VillageWitch Trials

Records of the events of 1692 do not support thehypothesis that ergot poisoning was involved.

Nicholas P. Spanos and Jack Gottlieb

on

Oct

ober

21,

201

0 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Symptoms of the Afflicted Girls

Convulsive ergotism characteristicallyproduces the following symptoms: (i)vomiting, (ii) diarrhea, (iii) a livid skincolor, (iv) sensations of heat and cold inthe extremities, (v) spastic muscular con-

tractions in the extremities, which in se-

vere cases may become permanent se-

quelae, (vi) severe itching and tinglingsensations, (vii) convulsions, (viii) a rav-

enous appetite following convulsions,(ix) death in severe cases (2, 3, 5, 12).Permanent dementia may also be a symp-

tom in severe cases. Perceptual distur-bances may occur, but such disturbanceswould not be expected to occur indepen-dently of the other symptoms (12).

Caporael says that "complaints ofvomiting and 'bowels almost pulled out'are common in the depositions of theaccusers" (1, p. 25). This statement isincorrect. Records of Salem Witchcraft(RSW) (11) contains 117 depositions bythe afflicted girls and 79 depositions inwhich other witnesses describe the be-havior of the girls. There are also eyewit-ness accounts by Mather (10), Lawson(9), Brattle (7), and Hale (8) which are

not contained in RSW. We examined allthese sources and were unable to findany reference to the occurrence of vomit-ing or diarrhea among the afflicted girls.In all these sources we found only threeinstances of gastrointestinal complaintsamong the girls (11, vol. 1, p. 106; vol. 2,p. 31). In one of these cases (11, vol. 1,

p. 106) the girl making the complaint(Mary Warren) lived outside the area

that Caporael suggested was exposed toergot (13). Thus 8 of the afflicted girlsdid not report any gastrointestinal symp-toms. Those who did reported only a

single instance. None of them reportedvomiting or was observed to vomit, andthere is no indication that any of themsuffered from diarrhea.We found no indication in any of the

works examined that the afflicted girlsmanifested a livid color of the skin. Wefound no reference to cold sensations inthe extremities, and only two referencesto burning sensations. In one of thosecases an afflicted girl slowly reached outand touched the hood of an accusedwitch, then immediately pulled back herhand and "cried out, her fingers, herfingers burned" (14, vol. 2, p. 110). Inthe second case the judges had obtaineda rag puppet which they believed hadbeen used by a witch to afflict people at a

distance. They burned the puppet in thepresence of the afflicted girls with thefollowing results: "A bit of one of therags being set on fire (the afflicted) criedout dreadfully (that they were burned)"24 DECEMBER 1976

(14, vol. 2, p. 216). Rather than ergotpoisoning, these descriptions suggestthat the afflicted girls were enacting-theroles that would sustain their definitionof themselves as bewitched and thatwould lead to the conviction of the ac-cused.

According to Caporael, the afflictedgirls' convulsions "appear to be epilepti-form" and their reports of being bitten,pinched, and pricked by specters "mayallude to the crawling and tingling sensa-tions under the skin experienced by er-gotism victims" (1, p. 25). There is noquestion that the girls frequently con-vulsed and reported being bitten andpinched. However, a careful look at thesocial context in which these symptomswere typically manifested belies the no-tion that they resulted from an internaldisease process. The trial testimony in-dicates very clearly that the girls con-vulsed and reported being bitten andpinched when an accused person's be-havior provided them with a social cuefor such acts.

For example, when one of the accusedwas ordered to look at an afflicted girl,"he looked back and knocked down all(or most) of the afflicted who stood be-hind him" (11, vol 2, p. 109). In anothercase, "As soon as she [the accusedwitch] came near all [the afflicted] fellinto fits" (11, vol. 1, p. 140). The court-room testimony contains a great manyinstances of the afflicted girls' convulsingen masse when the accused entered theroom, looked in their direction, movedhis chair, and so on (11). The afflictedgirls' reports of being pinched, choked,and bitten are described thus by Law-son, an eyewitness (9, p. 156):

It was observed several times, that if she [theaccused witch] did but bite her underlip intime of examination the persons afflicted werebitten on their arms and wrists and producedthe marks before the magistrates, Ministersand others.

The afflicted also produced the pins withwhich the accused purportedly pinchedthem (9, p. 156).The afflicted girls were responsive to

social cues from each other as well asfrom the accused and were thereforeable to predict the occurrence of eachother's fits. In such cases one of the girlswould cry out that she saw the specter ofan accused witch about to attack anotherof the afflicted. The other girl would thenimmediately fall into a fit (1, vol. 1, p.183). Termination of the girls' con-vulsions was also cued by social-psycho-logical factors. In some cases con-vulsions would cease when a certain Bib-lical passage was read (9, p. 158). More

commonly the girls' convulsions wouldcease as soon as they were touched bythe accused (7, p. 171).

Convulsions at the sight of a witch,alleviation of convulsions by the witch'stouch, prediction of their own and oth-ers' convulsions, and production by theafflicted of bite marks and the pins usedto pinch them were all considered stan-dard symptoms in 16th and 17th centurycases of demonic possession (15). Takentogether, these facts indicate that theafflicted girls were enacting the role ofdemoniacs as that role was commonlyunderstood in their day.

Caporael points out that one ergot al-kaloid, isoergine (lysergic acid amide),has 10 percent of the activity of LSD andmight therefore produce perceptual dis-turbances. She remarks that "the spec-tral evidence of the trials appears to behallucinogenic symptoms and perceptualdisturbances accompanying ergotism"(11, p. 25). The term "hallucination" is,unfortunately, very unspecific, and inthe psychological literature is used torefer to a wide variety of distinct experi-ences (16). Although LSD is commonlyreferred to as a hallucinogen, Barber hascorrectly pointed out that "subjects whohave ingested [LSD] very rarely report,when their eyes are open, that they per-ceive formed persons or objects whichthey believe are actually out there" (17,p. 109). Instead, they tend to report per-ceptual distortions such as persistent af-terimages, rainbow-like colors, halos onthe edges of objects, changes in depthperception, contours that appear to undu-late, and the like. None of the testimonygiven by the afflicted girls indicates per-ceptual distortions of that kind. Instead,they reported seeing "formed per-sons"-the specters of the accused-attacking, biting, pinching, and chokingthem and others.As to the remaining symptoms of ergot

poisoning, none of the work we studiedindicates that the girls experienced rav-enous appetites after their convulsions,suffered permanent contractures of thehands or feet or other signs of permanentneurological damage, suffered perma-nent dementia, or died. It should benoted that the girls often appeared to bequite healthy outside the courtroom.Even in the courtroom they did not ex-hibit the signs of chronic malaise anddebilitation that might be expected aftermonths of chronic poisoning. Thus,Brattle wrote (7, p. 187):

Many of these afflicted persons, who havescores of strange fits in a day, yet in theintervals of time are hale and hearty, robustand lusty, as tho' nothing had afflicted them. Iremember that when the chief Judge gave the

1391

on

Oct

ober

21,

201

0 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

first Jury their charge, he told them, that theywere not to mind whether the bodies of thesaid afflicted were really pined and consumed,as was expressed in the indictment; but wheth-er the said afflicted did not suffer from theaccused such afflictions as naturally tended totheir being pined and consumed, wasted etc.

In summary, while the afflicted girlsexhibited rather dramatic behavior, noneof them displayed the syndrome of con-vulsive ergotism. Instead, they showedsymptoms of "'demonic possession," aphenomenon that was fairly commonamong 16th- and 17th-century Puritans inboth England and Colonial America (18,19).

It is worth noting that the initial symp-toms of the afflicted girls were ratherambiguous, and that they began to corre-spond more closely to popular stereo-types of demonic behavior as the girlsgained increasing exposure to informa-tion about those stereotypes. The initialsymptoms included "getting into holes,and creeping under chairs and stools,and [using] sundry odd postures and an-tic gestures, uttering foolish and ridicu-lous speeches" (20, p. 342). About 2weeks after these symptoms began aneighbor had a "witch cake" baked inorder to determine whether the girlswere bewitched. Only after this eventdid the girls begin convulsing and report-ing the specters of witches (20, p. 342).As the witchcraft trials progressed, thegirls added to the repertoire. They col-lapsed en masse when looked at by theaccused during the first trial (11, vol. 1,p. 18). During the fourth examinationthey began complaining of being bittenwhenever they observed the accused ner-vously bite her lip and of being pinchedwhen she moved her hand (14, vol. 2, p.48). In later examinations they began tomimic the accused; they held their headsin the same position as that of the ac-cused (11, vol. 1, p. 87) and rolled theireyes up after the accused did so (11, vol.1, p. 142). This temporal pattern suggeststhat the demonic manifestations werelearned, that the girls' behavior was grad-ually (although perhaps unwittingly)shaped to fit the expectations for demon-ic behavior held by the community.

In Caporael's view, there is a "majordifficulty in accepting the explanation ofpurposeful fraud . . . [namely] the grav-ity of the girls' symptoms" (1, p. 22).The implication of this statement is thatthe girls' performances somehow tran-scended the volitional capacities of nor-mal, physically healthy people. There-fore it should be pointed out that numer-ous 16th-century English demoniacs whodisplayed all the symptoms manifested

1392

by the Salem girls later confessed thatthey had faked these displays (21). Theyconfirmed their confessions by publiclyenacting all of their supposedly in-voluntary symptoms. These facts cer-tainly do not prove that the perform-ances of the Salem girls consisted entire-ly of conscious faking, but they do in-dicate that the girls' behavior can beaccounted for without recourse to expla-nations based on unusual diseases.

Symptoms of Other Witnesses

Twenty-nine of the accused witcheslived in or on the fringes of Salem Villageor had moved from the village within afew years of the crisis (22, p. 375). Boyerand Nissenbaum (23) have pointed outthat most of the accused lived in one halfof the village and most of the witnesseswho testified against them lived in theother half. They hypothesize that thisgeographical split in the pattern of accu-sations was to a large extent a function ofpolitical and social factionalism withinthe village. Caporael postulates that theaccusing witnesses were exposed to er-got poisoning by their location while theaccused were not exposed by theirs. Shesuggests that not only the girls but"many of the other accusers" had physi-cal symptoms such as are "induced byconvulsive ergot poisoning" (1, p. 25).Records of Salem Witchcraft contains

Ill depositions made by 80 different wit-nesses (not including the afflicted girls)against the 29 accused village residents.Trial records compiled by Boyer andNissenbaum (22) include a depositionmade by one of these witnesses that isnot included in RSW. We examinedthese 112 depositions looking for behav-ior that, even in a broad sense, mightpossibly represent symptoms of con-vulsive ergotism. These symptoms, andthe number of individuals who sufferedfrom them, are shown in Table 1. Wit-nesses were excluded from this table ifthey reported that their symptoms oc-curred a year or more before the Salemcrisis began (five cases), or while theywere out of Salem and therefore notexposed to the supposedly ergotizedgrain (one case), or for some other rea-son could not have been exposed to er-got (one case-that of the 8-week-oldinfant referred to earlier). One of theseexcluded witnesses, John Londer, gave acolorful account of seeing a "thing" witha monkey's face and cock's feet (11, vol.1, p. 160). Caporael specifically cites thistestimony as a probable example of ergotpoisoning (1, p. 25) despite the fact that

Londer stated explicitly that he had expe-rienced the apparition 7 or 8 years be-fore the outbreak of the Salem crisis.The first fact uncovered by our exami-

nation was that 78 percent of the wit-nesses did not report suffering even asingle symptom; only 18 reported suffer-ing one or more symptoms after the er-gotism is hypothesized to have begun.Most of the testimony consisted of obser-vations made on the afflicted girls orother factual information (such as thatthe witness's cow died three days afterthe accused passed by his barn). Threewitnesses testified about the death of oneman and several testified about symp-toms of three other individuals. Altogeth-er, the testimony examined containedsymptoms for 21 individuals other thanthe afflicted girls.The first thing to note about Table 1 is

that none of the witnesses reported apattern of symptoms characteristic ofconvulsive ergotism. There is no evi-dence that any of them suffered vomit-ing, diarrhea, a livid skin color, per-manent contractures of the extremities, aravenous appetite, or perceptual distur-bances (other than apparitions). In 10 ofthe 21 cases only a single symptom wasreported. G. Cory reported a short-livedinability to say his prayers, and W. Put-nam mimicked the gestures of one of theaccused (he clenched his fist when sheclenched hers and held his head in thesame position as she did hers). These areobviously not cases of ergotism. In athird case, J. Putnam suffered brieflyfrom "strange fits." The timing of thesefits makes ergot an unlikely possibility.Caporael reasons that the village wasexposed to ergotized rye by December1691. Putnam reported having his fits inApril 1692. It is unlikely that he wouldhave been so late in succumbing to itseffects.

In a fourth monosymptomatic casetwo of the afflicted girls testified that J.Holton was "tormented" by spectersand that while they observed him thespecters left him and began attackingthem instead. Holton testified that hewas immediately cured as soon as thegirls reported that the specters had lefthim to attack them. Such an immediatealleviation of symptoms is obviously notcharacteristic of ergot poisoning.

In the other six monosymptomaticcases the witnesses each reported anapparition. These individuals all statedthat on one or more occasions they saw aspecter of some sort, usually the vividimage of an accused witch, a dead per-son, or an animal. All indicated beliefthat these imaginings were real events

SCIENCE, VOL. 194

on

Oct

ober

21,

201

0 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

rather than dreams (some occurred whilethey were in bed) or flights of fancy.However, none of these witnesses alsoreported perceptual disturbances (suchas halos around objects). As was pointedout above, apparitions or perceptual dis-tortions in the absence of other symp-toms are not characteristic of ergot poi-soning (13). The apparitions describedby these five witnesses were very similarto apparitions that five other witnesses(not included in Table 1) said they hadexperienced several years before the hy-pothesized outbreak of ergotism.The remaining 11 witnesses in Table I

each exhibited more than one symptom.In two of these cases (Bittford andGould) the witnesses' experiences con-sisted primarily of what were probablydreams or hypnagogic experiences. Bothmen reported being in bed at night whenthey saw apparitions of accused witches.Bittford testified that his experience wasaccompanied by a stiffness in his neckthat lasted several days, and Gould saidthat he was pinched twice on his side.Gould also reported a second apparition,which was followed by a pain in his footlasting 2 or 3 days.

Daniel Wilkins died after an illnessthat lasted about 2 weeks. The onlysymptom reported about his illness wasthat he appeared to be choking shortlybefore he died, and this was reportedonly after the afflicted girls testified that

they saw specters choking him. Wilkinsdid not show any sign of illness beforethe beginning of May 1692. For ergot toexplain these events he would have hadto be eating poisoned rye for 4 monthswithout exhibiting any symptoms andthen suddenly to have fallen ill and diedin 2 weeks-a highly improbable occur-rence.

Several symptoms are recorded forWilkins's sister Rebecca, but she hadnot exhibited any of them until after aphysician had diagnosed her brother'sillness as preternatural and after the af-flicted girls had reported seeing spectersattack his body.Another brother, S. Wilkins, reported

an array of symptoms which included apain in his hand, specters of a witch andof a black hat, falling off his horse, and astrong urge to run. None of them wereexperienced before June 1692.Four persons, J. Doritch, J. Indian, T.

Indian, and Mrs. Pope, displayed symp-toms during the trials similar to thosedisplayed by the afflicted girls. All con-vulsed and reported seeing specters thatafflicted them or others. Mrs. Pope con-vulsed whenever an afflicted girl "saw"her about to be attacked by specters (14,vol. 2, p. 109), and J. Indian's con-vulsions could be terminated by thetouch of a witch (14, vol. 2, p. 241). Onone occasion Mrs. Pope also reportedpain in her stomach whenever an ac-

cused witch "did but lean her Breastagainst the Seat" (9, p. 156). T. Indianeventually confessed that she had report-ed apparitions and enacted other symp-toms because her master had beaten herand otherwise threatened her until sheagreed to do so (20, p. 343).

E. Keysar is the only witness in all theSalem records whose testimony includessymptoms even vaguely resembling theperceptual distortions associated withLSD. Keysar reported that, while in a

darkened room, he saw "strange things"that quivered (22, p. 75). This was imme-diately followed by seeing a quiveringhand in his fireplace. Testimony of thistype may be associated with acute anx-

iety and a host of other factors as well as

with hallucinogenic substances. Thereare at least three reasons to infer thatKeysar's experience was due to anxietyand expectation rather than to ergot: (i)he reported no other symptoms, (ii) theexperience occurred in May 1692, fivemonths after the time he would havebegun ingesting ergot, and (iii) earlierthat same day he had been severelyfrightened because he believed that an

accused witch "did steadfastly fix [his]eyes upon me."The final case, and the only one to

exhibit as many as four of the symptomslisted in Table 1, is that of J. Bayley, whoas Caporael points out did not live inSalem Village. He and his wife had spent

Table 1. Symptoms of witnesses (other than the afflicted girls) who testified against the accused witches. A, vomiting. B, diarrhea. C, livid skin.D, permanent contractures. E, pain in extremities. F, death. G, temporary muscle stiffness. H, convulsions. I, ravenous appetite. J, perceptualdisturbances (not includijrg apparitions). K, apparitions. L, sensations of hot and cold. M, skin sensations (biting and pinching). N, stomach pain.0, choking sensations. P, temporary inability to speak. 1, symptom reported; 0, symptom not reported; ?, symptom questionable.

Reported RSW (11)Reporters avolume A B C D E F G H I J K L M N 0 P Totalsufferers and page

W. Allan 1 38 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1J. Bayley 1 113 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 1 4S. Bittford 1 108 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 2A. Booth 2 180 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1J.Childen 1 92 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1G.Cory 1 55 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1J. Doritch 1 262 0 0 0 0 0 0 0 ? 0 0 1 0 1 0 0 0 3

2 179B. Gould 2 178 0 0 0 0 1 0 0 0 0 0 1 0 1 0 0 0 3J. Holton 1 71 0 0 0 0 0 0 0 ? 0 0 0 0 0 0 0 0 1J. Hughes 1 38 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1J. Indian 1 64 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 2T. Indian 1 44 0 0 0 0 0 0 0 ? 0 0 1 0 0 0 0 0 2E. Keysar* * 0 0 0 0 0 0 0 0 0 ? 1 0 0 0 0 0 2M. Pope 1 59 0 0 0 0 0 0 0 1 0 0 1 0 0 1 0 0 3H. Putnam 1 275 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1J. Putnam 1 95 0 0 0 0 0 0 0 ? 0 0 0 0 0 0 0 0 1W. Putnam 1 96 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1D. Wilkins 2 7 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 2R. Wilkins 2 5 0 0 0 0 0 0 0 ? 0 0 1 0 0 0 1 0 3S. Wilkins 2 3 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 0 3E. Woodwell 2 178 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1

Total 0 0 0 0 2 1 2 7 0 1 16 0 3 3 2 2 39

*This testimony may be found in (22), p. 75.

24 DECEMBER 1976 1393

on

Oct

ober

21,

201

0 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

one evening there and left the next day.On their way out of the village theypassed the house of a man and wifeaccused of witchcraft. Bayley reportedthat at this point he felt a blow to hischest and a pain in his stomach. He alsothought he saw the accused witches(who were jailed at the time) near thehouse and then became speechless for abrief period of time. Shortly thereafterhe experienced another blow to the chestand thought he saw a woman in thedistance. When he looked again he saw acow rather than a woman. After arrivingat his home he reported feeling pinchedand bitten by something invisible. Hiswife experienced no symptoms. Capo-rael says Bayley's testimony "suggestsergot" (1, p. 25). It seems far more plau-sible, however, that being a fervent be-liever in witchcraft he experienced anupsurge of anxiety as he approached thehouse of two convicted witches than thathe ingested ergot during his stay in thevillage and by coincidence experiencedthe first symptoms of his poisoning as hehappened to pass the witches' house.

Thus, the testimony of the witnesseswho testified against the Salem Villagewitches does not support the ergot poi-soning hypothesis. On the contrary, ittends to disconfirm it.

The End of the Salem Crisis

Caporael says that "the Salem witch-craft episode was an event localized inboth time and space" (1, p. 25). Theimplication of this statement is that theepisode was confined to the geographicalarea hypothesized to be affected by ergo-tized grain. However, by midsummer of1692 individuals were being accused ofwitchcraft not only in Salem but also inthe neighboring towns of Amesbury, An-dover, Beverly, Billerica, Boxford,Charlestown, Gloucester, Ipswich, Salis-bury, -and Topsfield (I 1). The Salemcrisis even spurred on witch accusationsin Connecticut (24). No one has pro-posed that the spreading panic resultedfrom a concurrent spread of ergotizedrye. It is therefore worth noting that thewitnesses from neighboring towns whotestified against their own local witchesprovided the same kinds of spectral testi-mony that are found in the Salem re-cords. Andover even produced its ownafflicted girl.

Caporael cites a "commonly ex-pressed observation" that the Salemwitch hunt, after escalating through the

summer of 1692, ended abruptly "for noapparent reason" (1, p. 25). Her ownview is, apparently, that it ended abrupt-ly because the village was no longer ex-posed to ergotized rye. She points outthat, after the crisis had passed, some ofthe magistrates andjurymen experienceddeep remorse and had difficulty compre-hending their own behavior. She sug-gests that ergot may have altered theirthought processes during the crisis andafter they regained their senses theycould not understand what had happenedto them (1, p. 25).

It is impbrtant to point out that abruptendings to large-scale panics about witch-ery were the rule rather than the excep-tion. Midelfort, who has studied themany large-scale witch crises that oc-curred in 16th- and 17th-century Germa-ny (25), describes the process. Thesecrises commonly began with accusationsagainst socially deviant and lower-classindividuals. Accusations escalated quick-ly, and more and more prominent individ-uals who did not fit the popular socialstereotype of a witch were accused. Inev-itably, many people, including some ofthe prosecuting judges, became increas-ingly skeptical of the validity of the judi-cial procedures and the spectral evi-dence, and persons of standing tooksteps to bring the persecutions to anabrupt end. These crises were often fol-lowed by remorse and second thoughtson the part of some magistrates and oth-er officials. The course of the Salemcrisis was the same as that of the typicalGerman crisis.

In summary: The available evidencedoes not support the hypothesis that er-got poisoning played a role in the Salemcrisis. The general features of the crisisdid not resemble an ergotism epidemic.The symptoms of the afflicted girls and ofthe other witnesses were not those ofconvulsive ergotism. And the abrupt end-ing of the crisis, and the remorse andsecond thoughts of those whojudged andtestified against the accused, can be ex-plained without recourse to the ergotismhypothesis.

References and Notes

1. L. R. Caporael, Science 192, 21 (1976).2. G. Barger, Ergot and Ergotism (Gurney & Jack-

son, London, 1931).3. F. J. Bove, The Story of Ergot (Barger, New

York, 1970).4. W. C. Cutting, Handbook of Pharmacology:

Action and Uses of Drugs (Appleton-Century-Crofts, New York, 1972).

5. C. E. Sajous and J. W. Hundley, The Cy-clopedia of Medicine (Davis, Philadelphia,1937).

6. C. Hansen, Witchcraft at Salem (Braziller, NewYork, 1969).

7. Letter of Thomas Brattle, F.R.S., 1692, in Nar-

ratives ofthe Witchcraft Cases 1648-1706, G. L.Burr, Ed. (Scribner's, New York, 1914).

8. J. Hale, "A Modest Inquiry into the Nature ofWitchcraft" (originally published in 1702), inNarratives of the Witchcraft Cases 1648-1706,G. L. Burr, Ed. (Scribner's, New York, 1914).

9. D. Lawson, "A Brief and True Narrative" (orig-inally published in 1693), in Narratives of theWitchcraft Cases 1648-1706, G. L. Burr, Ed.(Scribner's, New York, 1914).

10. C. Mather, "The Wonders of the InvisibleWorld" (originally published in 1693), in Narra-tives of the Witchcraft Cases 1648-1706, G. L.Burr, Ed. (Scribner's, New York, 1914).

1 1 W. E. Woodward, Records ofSalem Witchcraft(privately printed, Roxbury, Mass., 1864; re-printed by DaCapo, New York, 1969).

12. A. Cerletti, in Proceedings of the First Inter-national Congress of Neuropharmacology, P.B. Bradley, P. Deniker, C. Radauco-Thomas,Eds. (Elsevier, Amsterdam, 1959), pp. 117-123.

13. Four of the afflicted girls, S. Bibber, S. Church-ill, E. Hubbard, and M. Warren, lived outside ofthe area supposedly exposed to ergot, and theplace of residence of two others is uncertain.Caporael (1, p. 24) offers post hoc speculationsconcerning how three of these girls might havebeen exposed to ergot, but provides no unam-biguous evidence to support her views. For in-stance, she cites a document indicating thatMary Warren stayed overnight in the village andmight, therefore, have eaten ergotized food. Shedoes not mention, however, that the same docu-ment (11, vol. 1, pp. 63-64) clearly indicates thatMary was experiencing her symptoms beforespending the night in the village. In short, thereis no evidence that 6 of the 11 afflicted girls wereexposed to the supposedly ergotized grain be-fore they began exhibiting symptoms.

14. C. W. Upham, Salem Witchcraft (Wiggins &Lunt, Boston, 1867; reprinted by Ungar, NewYork, 1959), vols. 1 and 2.

15. T. K. Oesterreich, Possession; Demoniacal andOther (University Books, Hyde Park, N.Y.,1966; originally published in German in 1921);N. P. Spanos, paper read at the annual meetingof the American Psychological Association, Chi-cago, 1975.

16. T. R. Sarbin, J. Pers. 35, 359 (1967).17. T. X. Barber, in Imagery: Current Cognitive

Approaches, S. J. Segal, Ed. (Academic Press,New York, 1971), pp. 102-124.

18. F. Hutchinson, An Historical Essay ConcerningWitchcraft (London, 1720).

19. The great popularity ofIncrease Mather's "Illus-trious Providences" may be cited as an exampleof the widespread Colonial interest in demonicpossession. In this work Mather discussed thecase of Ann Cole of Connecticut in 1662 and thatof Elizabeth Knapp of Groton, Mass., in 1671.He also discussed numerous other "pre-ternatural happenings" in both old and NewEngland ["An Essay for the Recording of Illus-trious Providences" (originally published in1684), in Narratives of the Witchcraft Cases1848-1706, G. L. Burr, Ed. (Scribner's, NewYork, 1914)].More or less simultaneous occurrences of "de-

monic possession" in a group of children oradolescents were by no means rare in the 16thand 17th centuries. The most famous of suchcases in England involved about a dozen af-flicted youths and led to the execution of threealleged witches [The Most Strange and Admir-able Discovery of the Witches of Warboys(London, 1593)].

20. R. Calef, "More Wonders of the InvisibleWorld" (originally published in 1700), in Narra-tives of the Witchcraft Cases 1648-1706, G. L.Burr, Ed. (Scribner's, New York, 1914).

21. S. Harsnett, Declaration of Egregious PopishImpostures (London, 1603); Discovery of theFraudulent Practices of John Darrel (London,1599).

22. P. Boyer and S. Nissenbaum, Eds., Salem Vil-lage Witchcraft: A Documentary Record of Lo-cal Conflict in Colonial New England (Wads-worth, Belmont, Calif., 1972).

23. _ _, Salem Possessed: The Social Origins ofWitchcraft (Harvard Univ. Press, Cambridge,Mass., 1974).

24. M. L. Starkey, The Devil in Massachusetts(Knopf, New York, 1950).

25. H. C. E. Midelfort, Witch Hunting in South-western Germany 1562-1684 (Stanford Univ.Press, Stanford, Calif., 1972).

26. We thank B. Jones, D. K. Chambers, M. E.Marshall, and A. B. Laver for their helpfulcomments on the manuscript.

SCIENCE, VOL. 1941394

on

Oct

ober

21,

201

0 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from