Contemporary Hypnosis ( I 998)

Vol. 15,No. I, pp.3-21

A WORKING MODEL OF THE NEUROPHYSIOLOGY OF HYPNOSIS:

A REVIEW OF EVIDENCE

John Gruzelier

Imperial Collegeof Science,Technologyand Medicine,London, UK

Abstract

Neuropsychophysiological evidenceis reviewedtestinga three-stage, top-downwork-

ing model of the traditional hypnotic relaxationinduction involving: (1) a thalamo-
cortical attentional network engaginga left frontolimbic focusedattention control
systemunderpinningsensoryfixation and concentrationon the induction;(2) instate-
ment of frontolimbic inhibitory systemsthrough suggestionsof tirednessat fixation
and relaxationwhereby anterior executivefunctionsare suspendedand directedby
the induction;(3) engagementof right-sidedtemporoposteriorfunctionsthroughpas-
siveimageryand dreaming.A selectivityof actionin high susceptibles was a hallmark
of the studies.IncreasedStroop interferencecoincidedwith maintenanceof error
detectionand abolition of error evaluationpotentials,interpretedas dissociationof
cognitive and affective executive systemsof the anterior cingulate. Verbal, category
and design fluency tasks were dissociatedwith hypnosiscentring on left anterior
processes as seenin left lateral and medial reducedEEG connectivity.Limbic modu-
lated electrodermalorienting responsesand frontal modulatedmismatchnegativity
waveswere inhibited.Asymmetriesin electrodermaland electrocorticalresponsesto
tonesshiftedto favour the right hemisphere,an asymmetryalso seenin visual sensi-
tivity. Haptic processingand visual sensitivitydisclosedmore distributedchangesin
medium susceptibles,while low susceptibleswere characterizedby poorer attentional
functionsat baselineand improvementsthroughthe induction.

Key words:EEG, ERPs,attentionlaterality,frontal limbic cingulate

Introduction
This article provides a review of experiments,thematicallyrather than chronologi-
cally structured,that were carried out in the Charing Crosslaboratory of Cognitive
Neurosciencewith the aim of understandingthe neuropsychophysiological basisof
hypnosis.Hypnosiswas defined operationallyon the basisof the traditional relax-
ation procedure,which beganwith eyefixation, suggestions of relaxationand eye clo-
sure,and wasfollowed by imageryassociated with deeprelaxationand a dream.This
was applied in all experimentsexceptone where the active-alertprocedure(B6nyai
and Hilgard,I976) wascompared.During the inductionthere were behaviouralchal-
lengesto assessdepth of hypnosis,and at the end questionsabout memory for the
induction,a post hypnotic suggestionand subjectiveratings.The purposewas to
unravel someof the changesin brain activity that accompanythe hypnotic induction
in both high and low susceptibleindividuals,and featuresthat may differentiatethem
in the baselinestate.
4 Gruzelier
The history of localizing neurophysiologicalmechanismsin hypnosisbegan by
likening hypnosisto sleep.Pavlovdemonstratedthe existenceof hypnosisin his con-
ditionedreflex studiesin dogsand attributedthis to a partial and spreadinginhibition
of the cortex, lessextensivethan occursin sleep,though Heidenhainhad reasoned
earlier that more than the cortex was involved becauseanimalsdid not behaveas if
they were decorticate(Windholz,1996).Although sleepbecamean increasinglypop-
ular analogy,to date electroencephalographic (EEG) studieshavenot found similari
ties betweenhypnosisand sleep(for a review seeCrawford and Gruzelier,1992).At
the sametime sleepcanbe inducedby instructionsof hypnosis,but conventionalpro-
ceduresstop short of sleepinduction.Hernandez-Peon (1977)proposedthat hypnosis
wascloserto wakefulnessthan sleep,involving alterationsin both consciousness and
executivefunctions, which he localized to the midbrain, pons and medulla.
Nevertheless, dependingon the nature of the induction,dream-likefeaturesare close
to the hypnoticexperienceof many high susceptibles.
Recordingsfrom intracranial electrodesin epileptic patients have disclosedthe
importanceof limbic structuresin hypnosis.Craiselneck,McCranie and Jenkins
(1956)reportedanecdotallythat hypnosiswasterminatedeachtime the hippocampus
was electricallystimulated.De Benedittisand Sironi (1986)demonstratedin a high
hypnotizable that there was a reduction in interictal focal abnormalities in the hip-
pocampusduring hypnosisand an increasein alpha activity.In a secondpatient they
concludedthat hypnosiswas associatedwith functional inhibition of the amygdala,
becausestimulationof the amygdalaarousedthe patient from hypnosisunlike stimu-
lation of the adjacentand reciprocallyconnectedhippocampus.Electricalactivity in
the amygdalabecamesynchronizedwith hypnosiswhereasactivity in the hippocam-
pus becamedesynchronized (De Benedittisand Sironi,1988).
In the 1960sparallelswere drawn betweenhypnosisand right hemisphericpro-
cessingand high hypnotic susceptibleswere assumedto be characterizedby right
hemisphericity.This point of view has been popularizedand has been found to be
important in the clinical useof hypnosis(Pedersen,1984).Aside from evidencefrom
cognitivestudieswith putative hemispherespecifictasks,there were also neuropsy-
chophysiologicalmeasureswhich took into account brain anatomy such as EEG,
dichotic listening and conjugatelateral eye movements(Crawford and Gruzelier,
1,992).Experimental evidencewas, however,conflictual,but the methodologiesof
many studieswere questionable(Gruzelier, 1988).Methods of inducing hypnosis
were various and often poorly described.Often there was a failure to distinguish
betweenhigh and low susceptiblesor to attribute changesin low susceptiblesthat
were absentin high susceptibles to the hypnotic process.Seldomwas evidencecited
of test-retestreliability, or of replication,or of validationof hypnoticlevel during the
experimentalprocedure.
In integratingthe resultsof a decadeago a three-stageworking model of the
inductionprocesswasproposed(Gruzelier,1988,1990).

StageI: The initial instructionsof fixating on a small object and listeningto the
hypnostist'svoice waspositedto involve an attentionalnetwork includingthalam-
ocorticalsystemsand parietofrontalconnectionswith engagementof a left ante-
rior focusedattention control system.This underpins the focused,selective
attentioninherentin fixation and listeningto the hypnotist'svoice,processes that
together require left hemisphericfrontotemporal processing.
StageII: The first stageis then replacedby eye closure,suggestions of fatigue at
continued fixation, and tirednesstogether with deep relaxation.This sets in
 Neurophysiologyof hypnosis 5
motion frontolimbic inhibitory processesunderpinningthe suspensionof reality
testing and critical evaluation,and the handing over of executiveand planning
functionsto the hypnotist;the 'letting go' componentof the hypnoticinduction.
. StageIII: The third stageinvolvesinstructionsof relaxed,passiveimageryleading
to a redistributionof functionalactivity and an augmentationof posteriorcortical
activity,particularlyof the right hemispherein high susceptibles.Simplifyingthe
verbal contentof the inductionmessagemay alsofacilitateright hemisphericpro-
cessing,as does emphasizingpast experienceand emotion. In contrast,low sus-
ceptiblesfail to show engagementof left frontal attentionalcontrol mechanisms,
or if there is focal attentional engagement,low susceptiblesfail to undergo the
inhibitory, letting go process.This working model will serve to structurethe
review of findings.Here emphasiswill be placedon the cognitiveneuroscience of
hypnosis;implicationsfor socio-cognitiveapproacheswill be found in a commen-
tary (Gruzelier,1998)on a theoreticalpaperby Wagstaff(1998).

Thalamo-frontal-limbic attentional processes

Electrodermal o rienting, habituation, sensitization and tonic reactivity

We first investigatedthe basic attentional processesof orienting, which represents
focusingof attention,and of habituationwith stimulusrepetition,which allowsatten-
tion to be redirected,and which involve modulationby the limbic system,in particu-
lar the amygdalaand hippocampus(e.g.Pribram and McGuinness,1975;Gray,1982).
The electrodermalresponsewas chosenbecauseamongphysiologicalmeasuresit has
the advantageof indexing sympatheticnervous systemactivity unconfoundedby
competingparasympatheticinfluences.Using a standardizedtone orienting and
habituationparadigm(Gruzelierand Venables,1972)the effectsof hypnosison ori-
enting responsesto tones that were interspersedwith the hypnotic induction were
monitoredin normal and patientvolunteers(Gruzelierand Brow, 1985).Subjectstook
part in three sessionsseparatedby four weeksto avoid carryovereffectson habitua-
tion. They were first monitored to provide a baselinemeasureand to equategroups
for individual differencesin rate of habituation. Then with sessionorder counterbal-
ancedthey experienceda hypnosissessionand one of two control conditions.The
control conditionsconsistedof either a story read by the hypnotist,or relaxinglisten-
ing to a story for a period equivalentin length to the hypnoticinductionprior to the
introduction of the tones,which were presentedwithout any accompanyingverbal
message. Hypnoticsusceptibilitywasmonitoredthroughoutthe experimentalsession.
The outcomewas clear and dependedon level of susceptibility.As shown in
Figure 1, in the group distributionsof orienting and rate of habituationthe hypnosis
conditionwas distinguishedfrom the three control conditionsthrough a higher inci-
denceof both non-respondingand non-habituation,a bimodal distribution.It wasthe
high susceptibles who showeda reductionin orientingand/or fasterhabituationwith
hypnosis,whereaslow susceptibles showedretardedhabituationwith hypnosis,ascan
be seenin Figure2.
Comparisonswith other featuresof electrodermalactivity shed light on arousal
and attentionalprocesses. In hypnosisfor the group as a whole there was an absence
of sensitizationto a test tone presentedtowardsthe end of the sessionwhile through-
out the inductionthere was an increasein electrodermalnon-specificresponses. It is
important to note that both theseeffectswere sharedwith the control story condition
and did not vary with hypnoticsusceptibility.From this it canbe inferedthat listening
to the story and to the hypnotic induction produced a similar degreeof autonomic
 Gruz.elier
TRIALSUMIL HABITUATION ffi Patients
ANDPATIEMS
STUDENIS l--1 Students

TONESALONE HYPNOSIS
WITHTONES
AFTER
INDUCTION
t2
10
R

4 8

IONESALONE TONESWITH
AFITRSTORY STORY
e

0
4 8 12 16 4 8 t216

Figure 1. Electrodermalorienting responsehabituation in baseline (tones alone), hypnosis,

relaxationafter listeninsto a storv.and storv conditions.

Patients
Students
Increase
in Habituation
q ?
ia 3 -
'E-
0

o 9

o
O
L
o
L
SameHabituation

2 o
Decrease
in Habituation

0 2 t 4 6 8 1 0
Susceptibil
ity Score

Figure2. Subjectscategorizedaccordingto increase,decreaseor no changein habituationfrom

baselineto hypnosisas a functionof inductionsusceptibilityscore.

arousability and attentional engagement and these did not vary with susceptibility.
Turning to the relaxation condition, this was insufficient to change either habituation
(see also Teasdale, 1972) or sensitization, so that the facilitation of habituation could
not be explained away as a function of relaxation. At the same time some compo-
nents of arousal reduction were associated with hypnosis. This was shown by fewer
 Neurophysiologyof hypnosis 7
non-specificresponsesduring hypnosisboth in the first half and secondhalf of the
hypnotic induction as well as by reductionsin tonic levels of skin conductance.
However, one high susceptiblehad levels of skin conductancetwo standarddevia-
tions abovethe control group mean,indicatingthat a reductionin tonic arousalis not
a necessarypart of the hypnotic processas will be confirmedlater in an experiment
with the active-alertinductionprocedure.
The facilitation of habituation with hypnosiswas replicated in an experiment
designedto compare hypnosiswith simulating hypnoslsin medium/high hypnotiz-
ables(Gruzelier et al., 1988).Subjectswere examinedfirst in a baselinesessionand
then assignedwith the Barber SuggestibilityScale(Barber, 1969)to the simulatoror
hypnosisgroups matchedfor suggestibility,electrodermalreactivity and sex.
Hypnotic susceptibilitywasmonitoredthroughoutthe secondsessionasin the former
experiment.Levels of susceptibilityall fell within the moderate to high range. As
before, rate of habituation was faster with hypnosisin the susceptiblesubjects
whereasin simulatorshabituationwas slower.Simulatorswere more arousedduring
the inductionprior to presentationof the tones(p <0.005)but subsequentlythere was
no differencein the number of non-specificelectrodermalresponses.Support was
found for reports that simulatorsare characlerizedby exaggeratedcompliance
(Williamsenet al., 1965;Hilgard et al., 1978),for in compliancewith instructionby
the hypnotistto forget about the tones,all but one simulatingsubjectclaimedat the
end of the sessionnot to have heard the tones,whereasall but one in the hypnosis
group admittedto hearingthe tones.
Neuroanatomicallythe influenceon electrodermalorienting and habituationwas
compatiblewith the evidenceof De Benedittisand Sironi (1988)arisingfrom record-
ings of intracranialelectricalactivity. They found that hypnosisinvolved functional
inhibition of the amygdalaand activation of the hippocampus.The amygdalahas
been shown to exert mainly excitatoryinfluenceson orienting activity whereasthe
inhibitory action of the hippocampusfacilitatesthe habituationof the orienting
responsewith stimulus repetition (Gruzelier and Venables, 1972;Pribram and
McGuinness ,I975).

Electrocortical event-relatedattentional components and frontal inhibition

Electrocorticalprocedureshave also demonstratedalterationsin attentionalprocess-
ing with the induction of hypnosisand, like electrodermalratesof habituation,these
changedin oppositedirectionsfrom baselineto hypnosisfor high and low suscepti-
bles (Gruzelier,1996).Cortical evokedpotentialswere measuredto infrequenttones
mixed with frequent tonesin a standardP300paradigmwith particularinterestin a
negativegoing (N2a) attentionalcomponent.The differencebetweenthe wave to the
infrequent target or deviant when subtractedfrom the frequent non-targetor stan-
dard belongsto the classof phenomenatermed MisMatchNegativity(MMN). This is
thought to involve a preattentivesensoryspecificprocessgeneratedin the auditory
cortex (superiortemporalgyrus).Aside from bilateral temporalmaximait has a pre-
dominant singlemaximum over the frontal cortex suggestiveof frontal involvement
(Naatanen,1992;Naatanenand Mitchie, 1979).
Before the topographicalEEG recordingsessionsubjectswere assignedwith the
Harvard Group scale(Shor and Orne, 1962)to high (9-12) and low (0-4) susceptibil-
ity groups.Baselinemeasureswere first recordedand thesewere repeatedfollowing
the hypnotic induction as in the electrodermalstudies,and repeateda secondtime
following an extendedinduction.Susceptibilitywas recordedthroughoutthe session
as in previousstudies.As can be seenin Figure 3 high susceptiblesshoweda large
8 Gruzelier
magnitude difference wave at baselineand a progressivereduction in MMN with
eachstageof the induction and in keepingwith frontal inhibition. By the later stage
of the induction MMN was negligiblein both the lateral frontal placements.
Importantly oppositechangeswere manifestedby the low susceptiblegroup.
Whereas at baselinetheir differencewave was absent,there was a progressive
increasein MMN throughthe experiment,until in the last conditionthe magnitudeof
at baseline,sug-
the differencewavewas on a par with the resultsin high susceptibles
gestingan increasingenhancementof attentionalprocessing.

HIGHSUSCEPTIBLES

lH1
[f H2

LOWSUSCEPTIBLES

tr--t R

IH1
t--1 H)

Figure 3. Mismatchnegativityscoresin baseline(B), induction (H1) and extendedinduction

(H2) in high and low susceptiblegroups.
 Neurophysiologyof hypnosis 9
Summary
There was a consistencybetweenthe electrocorticalMMN and electrodermalmea-
sures in depicting opposite changesfrom baseline to hypnosisin susceptibleand
unsusceptible subjects.Congruentopposingeffectson attentionhave alsobeenfound
in a recent Finnish study involving a computerizedvigilance task. High susceptibles
showedan increasein omissionerrorsand greatervariability in RTs from baselineto
hypnosis,while low susceptiblesshowed a reduction in errors and RT variance
(Kallio et al., 1998).Together these studiesshow that whereassusceptiblesubjects
evincedinhibitory influenceson attention with hypnosis,unsusceptiblesubjects
improvedattentionalperformanceasthe inductionprogressed.

Anterior inhibitory processes

Fr o nto -limb ic sup erv iso ry attentio nal sy st em

We went on to examineevidenceof frontal inhibition in the contextof contemporary
models of anterior functions that focus on attentionalcontrol systems(Posnerand
Peterson,1990;Shalliceand Burgess,1991).A supervisoryattentional systemthat
involvesthe frontal lobesand limbic systemmonitorsongoingactivity and modulates
behaviour in responseto novelty, as in orienting, and when environmentalstimuli
conveyconflictinginformation.We testedthis with a behaviouraland electrophysio-
logicalparadigmthat requiredthe monitoring of errorsin performance(Kaiseret al.,
1997).We utilized a Stroop-liketask involving a simpletwo-choicereactiontime task
in which a button waspressedaccordingto the sidein which a greenarrow waspoint-
ing (congruentcondition).This wascontrastedwith a complexfour-choicetask where
in addition to the green arrow condition red arrows were randomly presented,in
which casethe button must be pressedin the oppositedirection to the arrow (incon-
gruent condition). Electrophysiologicalevidencehas shown that following an erro-
neousresponsein a reactiontime task there is a large negativegoing wave at about
100ms, referred to as error-relatednegativity and termed an error detectionwave,
which is not elicited following correctresponses(Falkensteinet al., 1990;Gehring et
al., 1993).This negativewave is followed by a positivewave that varieswith a range
of task-relatedfactors and may represent context updating, error evaluation and
adjustmentof responsestrategies,an error evaluationwave(Falkensteinet al., 1995).
With hypnosisthe medium/highsusceptibilitygroup showedan increasein errors
on incongruenttrials - the stroop interferenceeffect - but no changein errors on
congruenttrials, whereasthe performanceof low susceptiblesremained constant.
Reaction times were not influenced by hypnosisin either group. Therefore in
medium/high susceptiblesthere was a failure to inhibit the automatic responsein
keepingwith an inhibition of frontal attentionalcontrol. F{owever,the large negative
going error detection wavesthat were elicited were non-significantlylarger in Ihe
medium/highthan the low susceptibilitygroup, and thesewaveswere unalteredby
hypnosis.In other words an error detectionsystemwhich operatesat an early and
possiblypre-consciousstageof processingwas not compromisedby hypnosis;to wit
the unconscioushidden observerof Hilgard, Morgan and McDonald (1975).In con-
trast we found that the positive going error evaluation wave following the error
detection wave was reduced in amplitude with hypnosisbut only in medium/high
susceptibles. Resultsfor the susceptiblesubjectsare found in Figure 4. This was in
10 Gruzelier
keepingwith inhibition of a frontal error evaluationprocessand wascompatiblewith
the behavioural data showing a higher error rate on incongruoustrials in the
medium/highsusceptibles with hypnosis.
The error detectionwave hasbeen localizedto a midline anteriorcingulategener-
ator (Dehaeneet al., 1994),a promisingcandidatefor involvementin hypnosis.The
anterior cingulateperformsexecutivefunctionsthat havebeen subdividedinto affec-
tive and cognitivecomponents(Devinskyet al., 1995).The cognitiveexecutivecom-
ponent is involved in responseselectionin advanceof any movement and in
cognitivelydemandinginformation processingsuchas Stroop interference,localized
by blood flow imagingand lesionstudiesto the anterior cingulate(Pardo et al., 1990;
Georgeet a1.,1.993; Vendrell et al., 1995),and by someto the right anterior cingulate
(Bench et al., 1993).The affectiveexecutivefunctionsare involved in regulationof
autonomicand endocrinefunctions,assessment of motivationalcontext and signifi-
canceof sensorystimuli and emotionalvalence.Theseare mediatedthrough exten-
sive connectionswith the amygdalaand periaqueductalgrey and autonomic
brainstem nuclei. Our results have indicated that the monitoring of motor perfor-
mance carried out by the cognitive executivecomponent remained intact, for the
error detectionwave and RTs were unchangedby hypnosis.Rather it would appear
that the affect systeminvolving connectionswith the rostral limbic systemincluding
the amygdalawas unresponsive,as shown by the absenceof the error evaluation
wave and apparentlymotivationalinfluenceson performance.This interpretationis
also in keepingwith the reducedelectrodermalorienting activity reflectinga reduc-
tion in amygdaloidexcitatorymodulatoryinfluences.Dissociationbetweencognitive
and affectiveanterior cingulateexecutivesystemswould explain the increasein the
Stroopinterferenceeffectwith hypnosis.

-to.@

Site: Cz
-4.@
high susceptibles

-6.08

a,Ba

a z-gs

4.SS

pre-hgpnos is

-399 -zqq -t80

[i | | iieconds

Figure 4. Error detection and error evaluation waves in medium/high susceptiblesin hypnosis
and pre-hypnosis baseline.
 Neurophysiologyof hypnosis 11
Left anterior inhibition
Someevidencehassuggestedthat the anterior inhibition may be laterally asymmetrical
and biasedtowardsthe left hemisphere.This was disclosedby measuringright and left
hemisphereprocessingtimes in dextral subjectswith a haptic object sorting task com-
paringleft and right hands(Gruzelieret al., 1984).This wasplannedinitially to validate
asymmetriesdisclosedby the bilateral monitoring of electrodermalorienting and habit-
uationprocesses describedin the next section;in the haptictaskthe mediationof hemi-
sphericinfluencesis unambiguouslycontralateral.Subjectssorted objectsby classwith
each hand separatelywhile blindfolded.Hand order was counterbalanced and there
was control for movement time. The task was done prior to hypnosisand again after
the hypnotic induction, with susceptibilitymonitored through the experiment.As
shown in Figure 5 high susceptiblesshowedan increasein right hand processingtimes
with hypnosiswhile there was no changein their left hand times, nor were there bilat-
eral changesin low susceptibles.The increasein the processingtimes of the right hand
(indexingthe left hemisphere)correlatedpositivelywith the hypnotic susceptibility
score.We replicated the slowing of left hemisphericprocessingwith hypnosisin high
susceptiblesin contrastto low susceptiblesin a follow-up experimentwith middle aged
subjects,which was performed away from the laboratory and which included a non-
hypnosiscontrol group, (Gruzelier et al., 1984).The combiningof both experiment
samplesdisclosedevidenceof a left hemisphericpreferencein high susceptiblesat base-
line ashad beenshownin the electrodermalstudyoutlinedin the next section.
Left-sidedinhibition of somatosensory functionswas replicatedfurther by exam-
ining haptic processingwith an active-qlertinduction (Cikurel and Gruzelier, 1990).
Following B6nyai and Hilgard (1976)subjectspedalleda stationaryexercisebicycle
againsta load and with instructionsof mental alertnessin the hypnotic induction.
Dextral subjectswere selectedwith the Barber scaleto form groupsof high and
medium susceptibles.Each subjectparticipated in two sessions,one involving the
conventionalrelaxation induction and the other the active-alertinduction. In both
conditionsa baselinemeasurewas obtained either while seatedor while pedalling.
Sessionorder was counterbalancedand in each sessionhypnotic susceptibilitywas
monitored to validate the group designationand to compare the induction proce-
dures,which in the event producedsimilar influenceson susceptibility(r = 0.79,

(J
20
o)
3
q)
E
IE
,a
C

c,
(J
o
o-
L

ro
post pre
Hypnos
is

Figure 5. Haptic processing times for right and left hands for pre-hypnosis baseline and with
hypnosis in high (left figure) and low (right figure) susceptibles.
12 Gruzelier
p <0.001).As in previousexperimentsthere was a hypnosisx hand interaction such
that for the group as a whole there was a slowing of right hand processingtimes
with hypnosisand no changewith the left hand. Subdivision of the subjectsinto
high and medium groups showed that whereasboth groups sharedthe right hand
slowing with hypnosis,the left hand of high susceptiblesgave faster sorting times
with hypnosis,whereasin medium susceptiblesthere was a bilateral slowingof pro-
cessing.Comparinginduction procedures,the active-alertinduction sharedwith the
conventionalprocedure the slowing of left hemisphericprocessingbut the active-
alert induction was solely responsiblefor the improvement in right hemispheric
processing.
Lateralized anterior inhibitory functions were examined further with a small bat-
tery of neuropsychological tasks(Gruzelierand Warten, 1993).Theseincludedword
and categoryfluency, which are both left hemispheretaskswith word fluency left
frontal and categoryfluencyleft temporal.Designfluencywasincludedto index right
anterior processing.Left and right hand finger tapping was included to examine
motor and pre-motor functions.Dextral subjectswere first selectedwith the Barber
scaleand divided into high and low susceptibilitygroups.Susceptibilitylevel was also
monitored throughout the experiment.The influence of hypnosison word fluency
differedsubstantivelybetweenthe groups.While low susceptibles showedan increase
in fluency from the pre-hypnosisbaselinewith hypnosis,high susceptibles showeda
decrease.While the groupsdid not differ at baselinein word fluency (althoughthere
was a mean advantageto high susceptibles) with hypnosisthere was a highly signifi-
cant differencebetweenthem. The semanticcategorytest showedsimilar mean
changesbut did not disclosesignificanteffects.Theseresultsare contrastedin Figure
6. With designfluency both groupsshowedan improvement with hypnosiswhereasin
finger tappingonly the low susceptibles showedimprovementin finger tapping- the
high susceptiblesshowedan impairment.The word versuscategoryfluency effects
have now been replicatedin the Finnishstudy mentionedabove(Kallio et al., 1998),
which has brought togetherin one investigationoppositechangesin susceptibleand
unsusceptiblesubjects,first in vigilance performanceinvolving a thalamo-cortical
(parieto-frontal)attentionalnetwork mentionedabove and secondin word genera-
tion involving the left prefrontal cortex.

Categoryfluency

,< unsusceptible

suscePtible

pre-hypnosis hypnosis

Figure 6. Word fluency for letters and categoriesin high and low susceptiblesin pre-hypnosis
baselineand with hypnosis.
 Neurophysiologyof hypnosis 13
Patternsof correlationsbetweenthe fluency taskssupportedan inhibitory influ-
ence of hypnosiscentring on the left anterior word fluency performance.First, the
three fluency tests did not correlate at baselineimplying an independenceof func-
tion. Second,correlationsbetween baselineand hypnosisconditions were signifi-
cant for categoryfluency and designfluency but not for word fluency,in support of
the word fluency test alone being altered by hypnosis.In contrast,significantcorre-
lations were obtainedwith hypnosisbetweenword fluency and both designfluency
and categoryfluency, but not between designand categoryfluency. This could be
interpreted as showingthat an underlyingprocesssuch as distributedinhibition
was most at work with word fluency, and while centring on left anterior processes
was also having some impact on design and categoryfluency which involved right
anterior and left temporal processingrespectively.Correlations also showed that
the significantrelation betweenleft and right hand finger tapping dexterity before
hypnosiswas reducedwith hypnosisin keepingwith a tendencytowards lateral
dissociation.
Anterior disconnectionwith hypnosisin susceptiblesubjectswas recently dis-
closedby Kaiser in an unpublishedexperimentinvolving EEG topographicalmap-
ping with a 32 electrodeanay in which we examinedregionalconnectivitywith EEG
coherence.EEG coherence,a putative measureof connectivity,was examined
betweenbipolar pairs of electrodes.This discloseda significanthypnosisx group x
conditioninteractionin high alphaactivity,as shownin Figure 7. In high susceptibles
with hypnosisthere was a reductionin connectivitywithin the left prefrontalregion-
specificallybetweenleft lateral (FP1and F7) and medial (F3 and FTCI) placements-
whereasthe oppositeeffect,namelyan increasein connectivitywasfound in low sus-
ceptibles.In baselinethere was also a highly significantdifferencebetweensuscepti-
bility groupsin the direction of greaterleft anterior connectivityin high rather than
low susceptibles.

o I
a
I
o I
o
c-l -I-
I

?
O

I
I prehypnotic

I
I hypnotio
hish

hypnoticsusceptibility

Figure 7. Left anterior EEG coherencein baselineand hypnosisfor medium/high and low sus-
ceptibles(high short distancecoherencerepresentslow connectivityand vice versa).
1,4 Gruzelier
Importantly the coherenceresult did not generalizeto all bandsbut wasrestricted
to high alpha activity.This representsan EEG band that relatesto cognitiveas dis-
tinct from connativeprocessing,variouslydescribedas indexinghigh workload, sus-
tained motor control and long-termmemory (Mecklinger and Bosel, 1989;Sterman
eI a1.,7994; Klimesche,1996;Burgessand Gruzelier,1998).The importanceof selec-
tivity in narrow band EEG power has also been demonstratedby our differentiating
hypnosisfrom baselinein high and low susceptibles(Williamsand Gruzelier,1998).

Summary
action of hypnosiswas shown
Further evidenceof a selectivityof neurophysiological
throughexaminationof anteriorinhibitory influences:

O the dissociationbetweenerror detectionand error evaluationwaves;

a the left lateralized influenceson haptic processing,and the improvement in right-
sidedprocessingthat wasspecificto the active-alertinduction;
the specificitywithin the left hemispherefor the effects on verbal fluency thal
were restrictedto letter and not semanticdesignatedcategories;
a the localizationof the changesin EEG coherenceto within the left frontal lobe;
a the restrictionof the EEG coherencechangesto the high alphaband.

Thesefactorsserveto introducea note of cautionin attemptsto interpret changesin

brain blood flow and metabolismin hypnosis,which do not permit such fine grained
interpretation,nor do they discriminatebetweenfacilitatoryand inhibitory functional
systems.

Right-sided processing

Focal versusdistributedinfluencesof hypnosishave also been demonstratedin

experimentsinvestigatingthe popular view of right hemisphericinvolvementin hyp-
nosis(Pedersen,1984).Firstly, in our original experimenton electrodermalorienting
and habituationprocesses(Gruzelier and Brow, 1985)bilateral recording disclosed
an asymmetryin the amplitude of orienting responsesfavouring the right hand in
hypnosisin high susceptibles whereasthere wasno reliableasymmetryin low suscep-
tibles.This was in contrastto the baselinesessionwhere after the initial tonesthere
was an asymmetryfavouringthe left hand in high susceptibles and againno reliable
asymmetry in low the
susceptibles; initial stage of orienting is thought to involve the
right hemisphere,which governs states of broadened attention, after which the focal
and selectiveattentionalabilitiesof left hemisphere take over as would be exempli-
fied in the high suseptibleswith their left preference after the initial trials (seeFigure
8). At the time the psychophysiological experiment was undertaken the mediation of
hemisphericinfluenceson electrodermal activity was considered controversial.
Subsequentlythe influence of limbic modulation has been clarified by intracranial
stimulationstudies(Mangina and Beuzeron-Mangina,1996), which have supported
the original interpretation of the dominanceof ipsilaterallimbic modulatory influ-
enceson the passiveorientingprocesses (Gruzelier,1973).Thereforethe resultsmay
be interpretedas showinga left hemisphericpreferencein high susceptibles at base-
line. While this resultwasnot predictedand ran counterto right hemisphericitytheo-
ries of hypnotic susceptibility,support was forthcomingfrom the subsequent
investigationof lateralizedhapticprocessingdescribedabove(Gruzelieret al., 1984).
Suchbaselineleft hemisphericadvantages may simply reflect greatercognitiveagility
 Neurophysiologyof hypnosis 15

I BLE
SUSCEPT UNSUSCEPTIELE
_Len

-3.0 -3.0

-1., -7.'
Log.
Log. Amplltude
Amplitude
-4.0

l0
l0
Irial
Trlal
-3.0 -3.0

-).5 -3.'
109. Log.
AmDlitude Amplitude
-4.0

l0 l0
Trial Irial

Figure8. The amplitudeof bilateralelectrodermalorientingresponsesshowingreversalof

asymmetry frombaseline(tones)to hypnosis andno consistent
in highsusceptibles changesin
lowsusceptibles.

in line with task demandsby high susceptibles(Crawford, 1989).Accordingly it

would appearthat it was the hypnotic induction that instatedthe right hemispheric
functionalpreferencein susceptiblesubjects.
An enhancementof right posteriorfunctionswas found in an experimentinvolv-
ing the divided visual field presentationof flashesrequiring brightnessjudgements.
The experimentinvolved three sessionsin the order baseline,hypnosis,baseline.It
wasconductedwith eyesopen with sufficienttrials to perform signaldetectionanaly-
sis to give estimatesof perceptual sensitivity independentof cognitive bias
(McCormackand Gruzelier,1993).Blocks of trials were interspersedwith a live hyp-
notic induction. Susceptibilitywas monitored throughout the experimentwith sub-
jects divided into medium and high susceptibles. Resultsare shown in Figure 9.
Perceptualsensitivitywas found to be enhancedin the hypnosiscondition compared
with the control conditions.Comparisonof the susceptibilitygroupsindicated a
1,6 Gruzelier
bilateral increasein perceptualsensitivityin medium susceptibleswhereasin high
susceptibles there was no changein susceptibilityin the left hemisphere(right visual
field) in contrast to a focal right hemisphericenhancement(left visual field). An
improvementin perceptualsensitivityis accordingto signaldetectiontheory indica-
tive of an increasein signalto noiseratio as will occurwith a reductionin centrallev-
els of arousal.Analysis of the cognitivebias variable showedthat judgementswere
more conservativefor the right hemispherethan the left for the group as a whole (p
<0.05).However,there wasno effectof group,providingno evidencefor the possibil-
ity that the influenceof hypnosismay be due to a shift in attitude suchas an adoption
of a lax responsecriterion leaving perceptualsensitivityunaffected (Naish, 1985).
The demonstrablechangesin perceptualsensitivitywith hypnosiswere consistent
with other reports (Segaland Fusella,1970;Miller and Leibowirz, 1976;Farthing et
al., 1982).Thus the results,which disclosedan enhancementof right posteriorpro-
cessingwith hypnosis,showedthat only in high susceptibles was this strictly lateral-
ized. It was more widely distributed in medium susceptiblesto include a bilateral
processingenhancementand was of lessermagnitude.A similar conclusionwas
reachedin the hapticsortingtask experimentabove(Cikurel and Gruzelier,1990).
Turning to central and temporal regions,an electrophysiological study was per-
formed measuringevoked potentialsto tone probes presentedsimultaneouslywith
the hypnotic induction and comparedwith their presentationduring a story read by
the hypnotist.Event-relatedpotentialsin both conditionswere referredto a baseline
condition giving three conditionsin all (Jutai et al., 1993).Bilateral electrodeplace-
mentsincludedcentraland temporalsitesand the analysiscentredon the N100 atten-
tional component.According to assessment of hypnotic susceptibilitymonitored
throughout the study subjectswere categorizedas low or medium/highsusceptibles.
Specificto the hypnosiscondition there was a right>left asymmetryat the temporal
location (electrodesT3l4) in the medium/highgroup. In contrastthere was an oppo-
site left>right asymmetryin the story condition of medium/highsusceptibles, as well
asin both the conditionsof the low susceptibles. This reversalof asymmetryto favour
the right hemispherein susceptiblesubjectsdid not extend to the lateral central
(C314)placements.The resultsdemonstratedthat right anterior temporal lobe activ-
ity wasraisedin medium/highsusceptibles with hypnosis.

LHem. .
xigtt
---.--- RHem. )
.a.'"'' -.-.... LHem. .
.,"'' Medium
nHem. l

fr;;;':'il:L--

Control Hypnosis Control

Figure9. Visual sensitivityin baselinecontrol conditionsand hypnosisin high and mediumsus-

ceptiblesfor left and right hemispheres.
 Neurophysiologyof hypnosis l7
Summary
The asymmetriesin electrodermalorientingresponsesand the corticalevokedpoten-
tial N100, both to auditory stimuli presentedduring the latter part of the induction
(Gruzelierand Brow, 1985;Jutai et al., 1993),indicatea shift in the balanceof tempo-
ral-limbic activity to favour the right hemisphere.In contrast the asymmetryof visual
sensorysensitivityassessed with signaldetectionanalysis(McCormackand Gruzelier,
1993)depictedan enhancementof right hemisphericprocessingin high susceptibles
and no changein the left hemispherewhereasmedium susceptibles showedbilateral
improvement.

Conclusion

This seriesof experimentshas shown a number of reproducible changesin brain

function that distinguishedmedium/highsusceptiblesafter instructionsof hypnosis
The attemptswere modest
both from their baselinestate and from low susceptibles.
in scopeand must be confined to the traditional hypnotic relaxation induction.
Continuingsupportwasprovidedfor associations betweenhypnosisand:

o activationof anteriorfronto-limbicinhibitory processes,

a anterior inhibition or disconnection,either lateralized to left hemisphericregions
or bilateraldependingon the processes examined,
a involvementof right temporoposteriorprocessing,
a evidenceof superiorattentionalabilitiesin high susceptibles,
a evidenceof poor attentional abilities in unsusceptiblesubjectswith progressive
improvementthrough the induction,and
. no evidenceof right hemisphericityin the baselinestatein susceptiblesubjects.

Across electrodermal,electrocorticaland behaviouraldomainssusceptiblesubjects

evincedinhibitory influenceson attention with hypnosiswhereasunsusceptiblesub-
jects improved attentional performanceas the induction progressed.Evidencewas
also found for bilateral alterationsof function in medium susceptiblesin situations
where changesin high susceptibles were lateralized,suggestingmore diffuse or dis-
tributed changesin medium susceptibles and more focal changesin high susceptibles.
Togethertheseresultsindicatethe importanceof stratifyinggroupsinto low, medium
and high susceptibles.
In our neuropsychological translationof the traditional hypnotic induction, hyp-
nosiswas initiated by engaginganterior executivecontrol systems.Aside from alter-
ations in cortical functions along anterior-posteriorand lateral axes,these will
orchestratetop down changesinfluencingthalamic and brain stem mechanisms.
Currently there is renewed interest in the electrophysiologyof thalamocortical
mechanismsin perceptualbinding,consciousperceptionand alteredstatesof aware-
ness(Llinas and Pare, 1991;Singer,1993).Llinas, Ribary, Joliot and Wang (1994)
have proposed that consciousness is a noncontinuousevent determined by simul-
taneity of activity in specific thalmocortical nuclei, which provide the content of
experience,and the non-specificdiffusethalamicprojection systemthat providesthe
context and alertness.In this regard the anterior and posterior cingulateappear of
particularpromisein the top down control of thalamicactivity relevantto hypnosis.
'One of the unique featuresof
Devinsky, Morrell and Vogt (1995)have remarked
anterior cingulatecortex circuitry is its diverse thalamic afferentsand consequent
ability to sampleinputs from more thalamic nuclei than any other cortical region.
18 Gruzelier
The ability to samplefrom a wide rangeof thalamicinputs may be crucialfor its con-
tributions to motor responseselectionfunctions.p 280.' The samecould be said for
conativefunctions and the limbic thalamus(Bentivoglio et al., 1993).Hypnosis
researchwould benefit from examiningthe interplay betweencortical and thalamo-
cortical systems,for which the methodologyof fast frequencyEEG transientsholds
muchpromise.
In Llinas's model dreamingis regardedas a stateof hyperattentiveness to intrin-
sic activity without the registrationof sensoryinput, a statewith an obviousaffinity
with hypnosis.This servesto acknowledgethat the dream analogyremainsappealing
for aspectsof the hypnotic experience.ConsiderFuster's(1995)descriptionof cog-
nitive featuresof dreaming,which include the altered senseof time and absenceof
temporality,the lack of guiding reality and critical judgement,the anchoringin per-
sonal experience,affective colouring, dissociationfrom sensoryinput and context.
'The fragmentednetworks activatedin the dream seemto lack the associativelinks
to a time frame, anchoredas they are in the present,without time tags and refer-
ences.'This could equallybe a descriptionof the hypnotic state as high susceptibles
experienceit.
As the studiesunfolded a selectivityof central action increasinglybecamea hall-
mark of hypnosisproviding undeniableevidenceof neurophysiologicalchangesin
susceptiblesubjects,which distinguishedthem from unsusceptiblesubjects.As
Schopenhauerremarked 'All truth passesthrough three stages.First it is ridiculed.
Secondit is violently opposed.Third it is acceptedas being self-evident'.Application
of the rapid advancesin cognitiveneuroscienceto hypnosisresearchmay make the
reality of the third stageevermore likely.

Acknowledgements

The researchwassuppportedby the SaugstadFund, the Institut fiir Grenzgebieteder

and studentshipsfrom the Medical ResearchCouncil
Psychologieund Psychohygiene
and the WellcomeTrust.

References
B6nyaiEI, HilgardER. A comparison of active-alerthypnoticinductionwith traditionalrelax-
ationinduction.
Journalof AbnormalPsychology 1976;85:21.8-224.
BarberTX. Hypnosis: A Scientific Approach. NewYork:Von Nostrand, 1969.
BenchCJ, Frith CD,GrasbyPM, FristonKJ, Paulescu E, Frackowiack RSJ,Dolan RJ.
Investigations of the functionalanatomyof attentionusing the Stroop test.
Neuropsychologia 1993; 3l: 907-922.
BentivoglioM, Kultas-Ilinsky K, IlinskyL In: Vogt BA, GabrielM, eds.Neuurobiology of
CingulateCortexand Limbic Thalamus:A Comprehensive Handbook.Boston:
Birkhauser,1993:7l-122.
BurgessA, GruzelierJH. Shortdurationpowerchanges in the EEG duringrecognition mem-
oryfor wordsandfaces. Submitted, 1998.
CikurelK, GruzelierJ. Theeffectof an active-alerthypnoticinductionon lateralasymmetry in
hapticprocessing.BritishJournalof Experimental andClinicalHypnosis1990;7:1725
CraiselneckHB, McCranieEJ, JenkinsMT. Specialindicationsfor hypnosisas a method
Journalof theAmericanMedicalAssociation
anaesthesia. 1956:162:1606-1608.
CrawfordHC. Cognitiveandphysiological flexibility:Multiplepathwaysto hypnoticrespon-
sieness.In: GhorghiuV, NetterP, EysenckH, RosenthalR, eds.Suggestion and
TheoryandResearch.
Suggestibility: Berlin:Springer-Verlag, 1989;155*168.
 Neurophysiology of hypnosis 19
Crawford HJ, Gruzelier J. A midstream view of the neuropsychophysiologyof hypnosis:
Recent researchand future directions.In: Fromm W, Nash M, eds. Hypnosis;Research
Developmentsand Perspectives, 3rd edn.New York: Guildford Press,1992;227-266.
De Benedittis G, Sironi VA. Deep cerebral electrical activity in man during hypnosis.
InternationalJournalof Clinicaland ExperimentalHypnosis1986;34:63-:70.
De BenedittisG, Sironi VA. Arousal effectsof electricaldeep brain stimulationin hypnosis.
InternationalJournalof Clinicaland ExperimentalHypnosis1988;36:96-106.
DehaeneS, PosnerMI, Tucker DC. Localisationof a neural systemfor error detectionand
compensation. PsychologicalScience1994;5: 303-305.
Devinsky O, Morrell MJ, Vogt BA. Contributionsof anterior cingulatecortex to behaviour.
Brain 1995:ll8: 279-306.
FalkensteinM, Hohnbein J, Hoorman J, Blanke L. Effects of errors in choicereactiontime
taskson the ERP under focussedand divided attention.In: Brunia CHM, Gaillard AWK,
Koh A, eds.Psychophysiological Brain Research.Tilburg: Tilburg University Press,1990;
192-195.
Falkenstein,M., Hohnsbein,J. and Hoorman,J. Event-relatedpotentialcorrelationof errorsin
reaction time tasks.In: Kamas G, Molnar M, CsepeV, Czigler I, Desmedt JE, eds.
Perspectivesof Event-RelatedPotentialsResearch(EEG supplement44). Amsterdam:
Elsevier, 1995;287296.
FarthingGW, Brown SW, Venturio M. Effectsof hypnotisabilityand mentalimageryon signal
detectionsensitivityand responsebias.InternationalJournalof Clinical and Experimental
Hypnosis1982;30:289*305.
FusterJM. Memory in the CerebralCortex.Boston:MIT Press,1995.
GehringWJ, GossB, ColesMGH, Meyer DE, Donchin E. A neuralsystemfor error detection
and compensation. Psychological Science1993;4:385-390.
GeorgeMS, Ketter TA, Gill DS, Haxby JV, UngerleiderLG, HerschovitchP. Brain regions
involved in recognisingfacial emotion or identity: an oxygen-l5 PET study. Journal of
Neuropsychiatryand Clinical Neuroscience1993;5: 384-394.
Gray JA. The Neuropsychology of Anxiety. Oxford: Oxford UniversityPress,1982.
GruzelierJH. Bilateral asymmetryof skin conductanceorienting activity and levelsin schizo-
phrenics.BiologicalPsychology1973;I:2t41
GruzelierJH. The neuropsychology of hypnosis.In: Heap M, ed. Hypnosis:Current Clinical,
Experimentaland ForensicPractices.London: Croom Helm, 1988;68-76.
GruzelierJH. Neuropsychological investigationsof hypnosis:Cerebrallaterality and beyond.
In: Van Dyck R, SpinhovenPh, Van der Does AJW, eds.Hypnosis:Theory, Researchand
ClinicalPractice.Free UniversityPress,1990;38-51.
GruzelierJH. The stateof hypnosis:Evidenceand applications.QuarterlyJournalof Medicine
1996;89:313-317.
GruzelierJH, VenablesPH. Skin conductanceorientingactivityin a heterogeneous sampleof
schizophrenics. Journalof Nervousand Mental Disease1972;155:277-287 .
GruzelierJH, Brow TD. Psychophysiological evidencefor a statetheory of hypnosisand sus-
ceptibility.Journalof Psychosomatic Research1985;29:287-302.
GruzelierJH, Warren K. Neuropsychological evidenceof left frontal inhibition with hypnosis.
Psychological Medicine1993;23:93-101.
Gruzelier JH, Brow TD, Perry A, Rhonder J, Thomas M. Hypnotic susceptibility:A lateral
predispositionand altered cerebralasymmetryunder hypnosis.International Journal of
Psychophysiology 1984;2: 137-139.
Gruzelier JH, Allison J, Conway A. A psychophysiological differentiationbetweenhypnosis
and the simulationof hypnosis.InternationalJournalof Psychophysiologyl988; 6: 331-338.
Hernandez-PeonR. Las basesneurofisiologicas de la hipnosis.NeurologiaNeurocirugia
Psiquiatria(Mexico) 1977; 18:7-26.
Hilgard ER, Morgan AH, MacDonald H. Pain and dissocaitionin the cold pressortest: A
studyof hypnoticanalgesiawith 'hidden reports'through automatickey pressingand auto-
matic talkins.Journalof Abnormal Psvcholosv1975'.84:280-289.
20 Gruzelier
Hilgard ER, McDonald H, Morgan AH, JohnsonLS. The reality of hypnoticanalgesia: a com-
parisonof highly hypnotisables with simulators.Journalof Abnormal Psychology1978;81:
239-246.
Jutai J, GruzelierJH, GoldsJ, ThomasM. Bilateralauditory-evokedpotentialsin conditionsof
hypnosisand focusedattention. InternationalJournal of Psychophysiology 1993:15:
167-\76.
KaiserJ, Barker R, HaenschelC, BaldewegT, GruzelierJ. Hypnosisand event-relatedpoten-
tial correlatesof error processingin a stroop-typeparadigm:a testof the frontal hypothesis.
InternationalJournalof Psychophysiology 1997; 27: 215-222.
Kallio S, RevonsuoA, HamalainenH, Markela J, GruzelierJ. Changesin anterior attentional
functionsand word fluencyassociated with hypnosis.Submitted,1998.
KlimescheW. Memory processes, brain oscillationsand EEG synchronisation. International
Journalof Psychophysiology 1996;24:61-100.
Llinas RR, PareD. Of dreamingand wakefulness. Neuroscience 1991;44:521-535
Llinas R, Ribary U, Joliot M, WangXJ. Contentand contextin temporalthalamocorticalbind-
ing. In: Buzsaki G, ed. Temporal Coding in the Brain. Berlin: Springer-Verlag,1994;
25r-273.
Mangina CA, Beuzeron-ManginaJH. Direct electricalstimulation of specifichuman brain
structuresand bilateral electrodermalactivity. InternationalJournal of Psychophysiology
1996;22:l-8.
McCormack K, Gruzelier JH. Cerebral asymmetryand hypnosis:A signal detection analysisof
divided visual field stimulation. Journal of Abnormal Psychology1993;102:352-357.
MecklingerA, Bosel R. Veraenderungenim EEG-Frequenzspektrum bei verscheidenPhasen
mentalerbelastung.Z.Exp. U. Angew.Psychologie1989;36:.453475.
Miller RJ, Leibowitz HK. A signal detectionanalysisof hypnoticallyinduced narrowing of
peripheralvision.Journalof Abnormal Psychology1976;85:44H54.
NaatanenR. Attention and Brain Function.New York: Erlbaum,1992.
Naatanen R, Mitchie PT. Early selectiveattention effects on the evoked potential. A critical
reviewand reinterpretation.BiologicalPsychologyL979;8: 81-136.
Naish P. The trance describedin signal detection terms. British Journal of Experimental and
ClinicalHypnosis1.985 ; 2: 133-138.
Pardo JV, Pardo PJ, Janer KW, Raichle ME. The anterior cingulatemediatedprocessingselec-
tion in the Stroop attentionalconflict paradigm.Proceedingsof the National Academy of
Science1990;USA 87:256-259.
PedersenD. Hypnosisand the right hemisphere.Proceedings of the British Societyof Medical
and Dental Hypnosis1984;5:2-14.
P o s n e r M , P e t e r s o nS . T h e a t t e n t i o n s y s t e mo f t h e h u m a n b r a i n . A n n u a l R e v i e w o f
Neuroscience1990;13:2542.
Pribram KH, McGuinnessD. Arousal, activationand effort in the control of attention.
PsychologicalReview I975; 82:-116.
SegalSV, FusellaV. Influenceof imagedpicturesand soundson detectionof visual and audi-
tory signals.JournalofPsychology1970;83:458464.
ShalliceT, BurgessP. Deficits in strategy application following frontal lobe damagein man.
Brain 1991;Il4: 727-741.
Shor RE, Orne EC. The Harvard Group Scaleof Hypnotic Susceptibility. Form A. Palo Alto:
ConsultingPsychologists' Press,1962.
Singer W. Synchronisationof cortical activity and its putative role in information processing
and learning.Annual ReviewofPhysiology1993;55:349-374.
StermanMB, Mann CA, KaiserDA, SuyenobuBY. Multiband topographicEEG analysisof a
simulatedvisuomotor aviation task. InternationalJournal of Psychophysiology 1994;16:
49-56.
TeasdaleJD. Relaxationand habituation:A theoreticaland experimentalanalysis.PhD thesis,
Universityof London, 1972.
 Neurophysiology of hypnosis 21'

Vendrell P, JunqueC, Pujol J, JurandoMA, Molet J, GrafmanJ. The role of prefrontalregions

in the Strooptask.Neuropsychologia 1995;33:.341-352.
WagstaffG. The semanticsand physiologyof hypnosisas an alteredstate:Towardsa definition
of hypnosis.ContemporaryHypnosis,1998,in press.
Williams JD, Gruzelier JH. Differentiation of hypnosis and relaxation by analysisof narrow
band theta and alphafrequencies. Submitted,1998'
WilliamsenJA, JohnsonHJ, Eriksen CE. Somecharacteristics of post-hypnoticamnesia.
Journal of Abnormal Psychology 1965:74: 123-131.
Windholz G. Hypnosisand inhibition as viewed by Heidenhain and Pavlov.Integrative
Physiologicaland BehaviouralScience1996;31:155-162.

Address for correspondence:

Professor J ohn Gruzelier,

Department of Behavioural and Cognitive Sciences,
Imperial College School of Medicine,
St Dunstans Road,
London W6 8RF, UK