The British Society for the Philosophy of Science

Towards a Theoretical Conceptualisation of Superstition

Author(s): Karl E. Scheibe and Theodore R. Sarbin
Source: The British Journal for the Philosophy of Science, Vol. 16, No. 62 (Aug., 1965), pp.
143-158
Published by: Oxford University Press on behalf of The British Society for the Philosophy of Science
Stable URL: http://www.jstor.org/stable/686154 .
Accessed: 04/07/2011 00:37

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=oup. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Oxford University Press and The British Society for the Philosophy of Science are collaborating with JSTOR to
digitize, preserve and extend access to The British Journal for the Philosophy of Science.

http://www.jstor.org
 TOWARDS A THEORETICAL
CONCEPTUALISATION OF SUPERSTITION *

and THEODORER. SARBIN

KARL E. SCHELBE

" Manneedsa rulefor his will, andwill invent

one if one be not givenhim"-William James

I. Introduction

THEpurposeof this paperis to suggesta generalway of thinkingabout

superstitions. While it is common to describesome beliefs and be-
havioursas 'superstitions', it is uncommonfor philosophers,psycho-
logists, and other systematicwritersto be explicit about the meaning
they intend to convey by the use of such a descriptivelabel. A wide
range of phenomena, including magic, ritual, myth, and occult
practices,is traditionallydenoted by the term superstition,and by cus-
tom the whole lot is summarilyassignedto the provinceof anthropo-
logical inquiry. While a greatamountof concernwithin psychology
has centredon such relatedtopics as prejudiceand attitudinalbias, not
muchefforthasbeen directedtowardsan understanding of superstitions
as such. Few and limited, indeed,have been the attemptsto integrate
a conceptof superstitioninto psychologicaltheory.1
To the theorist,the topic is immediatelydiscouraging. The term
'superstition'is so frequentlyusedthat common experienceand com-
mon senseseem to provide an adequatebasisfor understanding. We
know that superstitionsare to be regarded as erroneous beliefs:
irrational, primitive, and based on inadequate information. In
addition,we subscribeto the assertionthat superstitions,even though
manifestlyfalse,provide some want-satisfying,or utilitarian,function.
* ReceivedIo.ii.65
1A notableexceptionis Willoughby(1935). A checkof the PsychologicalAb-
stractsrevealsthatfewerthanthreearticlesperyearhaveappeared underthe' belief'
indexheading. Even fewer arelistedundersuperstition.There seemsto be no
quantitativetrendoverthe lastthirty-sixyears. In a recenthistoricalreviewof the
topic,Cohen(I960)hasshownthatonlyfive of a sampleof fiftypsychological text-
bookshavedevoteda chapterto beliefs. Theresistance of psychologists
to investigate
beliefsis worthyof studyin its own right.
143
 KARL E. SCHIEBE AND THEODORE R. SARBIN

Then, too, we are aware that the amount of descriptive material existing
on superstitions is vast by any standards. Frazer's (1925) encylopedic
GoldenBoughand Tylor's (1888) classic work on primitive societies are
but two early examples of the bulk of ethnographic work on this topic.
Lewinsohn's (1961) book on prophecy and prediction, Willoughby's
(1935) lengthy article on magic, and Cohen's (1960) work on gambling
behaviour are likewise replete with examples of superstitious practices.
In short, the concept is too familiar to seem very strange. But upon
exanminationthis familiarity acquires a fascinating strangeness.
The focus of this fascination is the functional significance of invalid
beliefs. A number of writers have been impressed with this idea.
Malinowski's (1926) analysis of magical practices among the Trobriand
Islanders centres about the thesis that every such practice and corres-
ponding superstitious belief,' fulfills some vital function, has some task
to accomplish, representsan indispensablepart within a working whole'
(p. 132). Malinowski established the importance of superstitions by
viewing them in theirfunctional context. The serious student of cul-
ture and behaviour could no longer disdainfully dismiss superstitions
as futile primitive beliefs or esoteric phenomena. Malinowski con-
sidered superstitions to be current vital elements within a culture rather
than vestigial remnants of the past.
This conception is both tempered and extended toward psychology
in Merton's discussion(1957) of the distinction between the manifest and
latent functions of social behaviour. The application of latent function
analysis clarifies apparentirrational, paradoxical, and superstitioussocial
patterns. For instance, once the latent function of conspicuous con-
sumption is seen to be the acquisition of status through expensive
purchases, this pattern of social behaviour is better understood, and
no longer appears paradoxical. Superstitions are established in this
sense as continuous with normal or rational modes of believing and
behaving.
A third and more purely psychological point of view is contained in
Rotter's Social Learning Theory. Rotter's conception of' Generalised
Expectancies' is inclusive of a class of beliefs that might well be labelled
'superstitions'. An expectancy is defined by Rotter (1954) as, 'The
probability held by the individual that a particular reinforcement will
occur as a function ofa specific behavior on his part in a specific situ-
ation or situations' (p. 10o7). Expectancies derived from specific situ-
ations tend to generalise to similar situations. Rotter views the failure
to cross-validate generalised expectancies as responsible for superstitions.
144
THEORETICAL CONCEPTUALISATION OF SUPERSTITION

The remainingroot of the presentdiscussionis a generalinference

model developedby Sarbin,Taft, and Bailey (1960) as a psychological
descriptionof the predictionprocess. Sarbin,et al. in consideringthe
informationalbasisof categoricaljudgments(whichare calledinstanti-
ations), suggest that, 'Where adequateinput is lacking, the cognitive
organisationsfunctionto " fill in the gap " and instantiationsare then
largely a functionof the modules(predicates)selectedby factorsother
than sensory input' (p. I31). Propositionsin the form of major
premisesstandreadyto give meaningto sensoryinputs. Propositions
that are empiricallyderived or inductively confirmedearn the label
'scientific' or 'rational'. Others, developed through analogy or
fantasyor acquiredthroughauthority,until examinedthroughinduc-
tion, mustbe consideredirrationalor superstitious. Regardlessof how
derived, these premisesserve as guides to action.
Both Rotterand Sarbinsuggest,in effect,that a person'sconductis
predicatedon more information than is provided through current
inputs. Malinowskiand Merton argue that culturesand societies do
the same thing. In all these sources, the thesis is emphasisedthat
invalid beliefshave functionalsignificance. We directthe rest of this
essay to an elaborationof this thesis. In the end we hope to provide
the basis for a useful set of conventions for the employment of the
term 'superstition'.

2. SomeDefinitionsandSomeQuestions
Anticipatingthe model which will later be presented,we define a
superstitionas a specialkind of expectation(or subjectiveprobability,
hypothesis, premise, belief). A superstitionmay be said to exist
whenever an individualpersistentlyor repeatedlybehavesas if his sub-
jectiveestimateof theresultof thatbehaviouris significantly
diferentfroman
objective(scientific)estimateof the effectof that behaviour. Some un-
settledproblemsas to the inclusivenessof this definitionare contained
here, but it will do for a start. However, it shouldbe clearlyunder-
stood that the term superstitionis not meant at this time in its usual
pejorativesense. The full reasonsfor this will become clearlater.
Questions that deserve examination regarding these special ex-
pectanciesincludethe following. What determineswhetheror not an
individual subscribesto a given superstition? Some people carry
charmsand amulets,consultastrologersand palmists,and some do not.
Is it the casethat some people are superstitiousand some arenot? Or
K 145
 KARL E. SCHEIBE AND THEODORE R. SARBIN

may it be that everyoneis superstitiousin nearlylike degree,only the

specificcontentof the superstitionvaryingfromindividualto individual
andfrom groupto group? Thatcollegestudentssubscribeto numerous
superstitionshad been demonstrated(Gilliland,1930; Dockeray and
Valentine, 1935). Is it not appropriateto describethe postulatesof
scientifictheoreticiansand of theologiansas somewhathighly evolved
superstitions?1
At the moment, we have no systematicstudiesto supporthypo-
theses regarding differential distributionsof superstitionsamong
differenteducationalclasses.
These problemsare approachable,but it seemsthatthey may all be
subsumedunderthe more generalquestion: What are the conditions
under which superstitionsare formed, maintained,and abandoned?
It is to this questionthatwe addressourselves,but not until we present
a theoreticalconceptualisationof superstitionin more detail.

3. A Cognitive Model
What does a man need in orderto participatein a culture? In the
simplestterms,he needsa body with musclesthatwork. In addition,
he needs some desires,or values. These two propositionsmay be
combinedinto the following sentence: Man wants a great variety of
things; for example,food, shelter,and clothing; children,happiness,
satisfaction,security,and a senseof self-esteem,andis capableof action.
But he lacks the sine.quanon to participatein a culture, to behave
effectively. He needsto know how to directhis actionstowarddesired
ends. In Tolman's(I958) terms,he needs to have a cognitive map of
means-endsrelationships. In the terms of the present analysis,he
needs a set of beliefs, some of which may be denoted by the term
' superstitions'.
The behaviour model here suggestedis in part borrowed from
Rotter (1954). The potentialto behaveis some functionof the expec-
tancy regardinga favourableoutcomeof thatbehaviour,and the value
of the outcome of that behaviour. Expectanciesderive from past
observationsof contiguously occurring events, and may be said to
represent perceivedcause and effect relationships between behaviours
and the outcomes of behaviours. To complete the skeleton of the
model for superstition,it need only be observedthat real cause and
1
senseis not hereintended,anaffirmative
Sincethe pejorative answerto thislast
needoffendno one.
proposition
146
THEORETICAL CONCEPTUALISATION OF SUPERSTITION

effectrelationshipsareobscure,andoften quitedifferentfromperceived
causeand effect relationships.
We are now brought up to the psychologicalproblem of epistomo-
logy. How do we come to know what we know? What is the
relationof the real world to the known world? How do we know
thatwe know? Hume argues that cause and effect relationshipsare
not ultimately compelled upon us in toto, but rather are inferences
basedupon contiguity. He saysin his Treatiseof Human Nature,' We
find only thatthe effectdoes, in fact, follow the cause. The impactof
one billiardball upon anotheris followed by the motion of the second.
There is here contiguity in space and time, but nothing to suggest
necessaryconnection.' Yet, it is often over-looked that though this
inferencebe logically unnecessary,it is psychologicallyvery powerful.
Cause and effect relationshipsare psychologicallycompelled upon us
in toto, as Heider (1944), Michotte (I963), and Piaget (1955) have amply
demonstrated. In short,it is possibleto believe that the known world
is the realworld, andto act asif it is, even thoughit might be impossible
to be ultimatelycorrectin this belief.
Consideran example: FarmerBrown wants-toplanthis corn. He
wantsto do it so as to mnaximise his autumnharvest. What is to deter-
mine when and how he plantsand cultivateshis corn? He certainly
does not proceedon a trialand errorbasis. He usesa numberof rules.
These will have to do with the calendardate, the stateof the weather,
the moistureof the soil, the type of seed, and perhapsthe phaseof the
moon or the pain in his bunions. He has probably derived few if any
of these rules by his own empirical observations, but has acquired them
from his 'cultures '. But only some of them would be called super-
stitions, and these would be so called because of some demonstration
that the application of the rule hasno effect on the outcome of the crop.
However, even if Farmer Brown were to be shown compelling
statisticswhich contradicted one of his rules (expectancies) he still might
apply the rule. Under these circumstances, we would feel justified
in calling Farmer Brown's behaviour superstitious.
To rephrase the conceptual scheme: superstitions come to exist
because of imperfect knowledge of antecedent-consequent relation-
ships in the ecology. On the assumption that an elaborate and fairly
definite set of beliefs is necessaryto support active behaviour, we con-
sider superstitions to 'fill the gap', as it were, between the objective
probability estimates available to a person, and the degree of subjective
probability required to execute behaviours. The beginning fact is
'47
 KARL E. SCHEIBE AND THEODORE R. SARBIN

objective uncertainty. In the face of this uncertainty, beliefs must be

adopted, so that a man is not transfixed for lack of Truth. Where he
lacks truth and needs it, he makes it up, somehow. We shall now
consider how this construction occurs.

4. The Developmentof Superstitions

In an experiment with pigeons, Skinner (I948; also Morse and
Skinner, 1957) has demonstrated the means of development of a ' sort
of superstition '. Using his operant conditioning technique, he was
able to establish firmly those behaviour patterns which happenedto
precede the presentation of the reinforcement. The bird behaves as if
'there were a causal relation between its behaviour and the presentation
of food, although such a relation is lacking' (Skinner, 1948, P. I70).
The superstitious rituals are due not only' to the fact that a reinforcing
stimulus strengthens any behavior it may happen to follow, even
though a contingency has not been explicitly arranged, but also to the
fact that the change in behavior resulting from one accidental con-
tingency makes similar accidents more probable' (Morse and Skinner,
1957, p. 308).
Harry Stack Sullivan quotes as a prime example of primitive
'parataxic' thinking a case portrayed in one of Franz Kafka's short
stories. A dog who lived in a kennel surrounded by a high fence was
urinating one day when a bone was thrown over the fence. The dog
thought, 'My urinating made that bone appear'. Thereafter when-
ever he wanted something to eat he lifted his leg.x Sullivan believes
that much of our thinking does not advance beyond the level of
parataxis; that we see causal connections between experiences that
have nothing to do with one another.
The essential features of these animal illustrations is that accidental
contingencies are mis-represented as essential contingencies. There is,
of course, a wealth of anecdotal human evidence which fits the same
description. For example (from Diprose's Anecdotes of Curious
Superstititionsand Omens):
In Devonshire, the appearanceof a white-breastedbird has long been
consideredan omen of death. This belief has been tracedto a circum-
stance said to have happenedto the Oxenham family in that county,
and relatedby Howell, in his FamiliarLetters; wherein is the following
monumental inscription: Here lies John Oxenham, a goodly young

1 We are indebtedto Hall and Lindzey(1957) for this example.

148
THEORETICAL CONCEPTUALISATION OF SUPERSTITION
man,in whosechambers, as he wasstrugglingwith the pangsof death,
a bird with a white breast,was seenflutteringabouthis bed, and so
vanished. He diedimmediately. The samecircumstances arerelated
of his sisterMary,andtwo or threeothersof the family.
The paradigmis the same. First is the condition of uncertainty
regardinga valuedoutcome: Skinner'spigeonsuncertainas to when or
whether they were going to be fed, and John Oxenham's relatives
uncertainas to when or whetherJohn would expire. Then, immedi-
ately prior to the revelation of the outcome, something fortuitous
happened. Moreover, this fortuitouspairingof events was repeated,
at least once. Since both the pigeon and the Oxenhamsare ignorant
of any more powerful causalagent, the fortuitousevents are taken as
causes. In the future, the pigeon has the habit of making swiping
motions with its head. The Oxenhams have the habit of avoiding
white-breastedbirds. The sameis true of Kafka'sdog, which has the
habitof raisingits leg at inappropriatetimes.
This type of explanationdoes well enough for the individualscon-
cerned. But how do we explainthe fact that the white-birdsupersti-
tion becomes widely disseminatedthroughoutthe culture, while the
colony of pigeonsis probablynot so takenin? We would do well at
thispoint to heed a warningmadelong ago byJastrow(1900oo) regarding
the development of superstitions. He warns againstthe practice of
analogicalreasoning: The inferenceof a furtherdegreeof resemblance
from an observeddegree of resemblance. Jastrowarguedthat many
superstitions,in particularVoodoo practices,derive from this faulty
logical procedure. Sarbin,et al. (1960), have formalisedthis process
of analogicalmediationas follows:
A hascharacteristic
x.
x andy aredefiningcharacteristics
of speciesM.
Therefore,A hascharacteristic
y andis a memberof speciesM (p. 62).
The conclusion,of course,is false. Yet the Voodoo Priestpersists
in the belief that mutilationof the representationof a personwill do
harmto thatperson,andour more primitivecomparativepsychologists
persistin the belief that what is true of learningin the pigeon is also
true of human learning.
We should concludethat the pigeon can carryus no fartherin our
understandingof superstition. We are left with the majority of
human superstitions quite unexplained. Skinnerian conditioning
must involve the empiricalcontiguity of the perceivedcauseand effect
149
 KARL E. SCHEIBE AND THEODORE R. SARBIN

for each individual who bears the superstition. Fortunately or un-

fortunately, there seems to be something else than this involved in the
ways people acquire superstitions. In fact, for human beings, even
the first-order demonstration of the efficacy of superstition is not
necessary. That is, they need not ever be witnesses to the original
contiguity between the fortuitous event and the outcome. For
example, Leonardo da Vinci, in The Last Supper,painted Judas Iscariot
in the act of knocking over the salt-cellar. This is held to be the origin
of the still extant superstition that it is bad luck to spill the salt. Now
most people who have this superstition are probably not even aware
of its origin. Likewise, astrological superstitions may have had their
origin in the fortuitous pairing, of a given stellar orientation and a
particular noteworthy earthly event. But such origins are buried in
antiquity. Yet people continue to be superstitiousabout the placement
of stars.
It is to be noted that popular astrology has enjoyed a recent boom
and revival. Lewinsohn (1961) offers the following speculation on
this phenomenon:
While the revolution which has been changing the Newtonian world
picture ever since 1900oo has made some people more skeptical, it has
simply made others more credulous. Scientificlaws which our text-
books have taught us to look upon as eternaltruthshave collapsedlike
ninepins, and new ones still lack the authority that their predecessors
enjoyed. Scientists have grown a little more circumspect and no
longer dare to proclaim general laws with the facility of their I9th
century predecessors. Because of these developments, laymen have
begun to feel that it is impossibleto distinguishtruth from hypothesis.
Now, laymen have alwaysbeen sticklersfor certainty,and where truth
is uncertain,they will acceptanythingthat is presentedas being simple
and sure (pp. 93-94).
In this vein we might also suggest that the current resurgence of
mysticism in psychology, as exemplified by some of the existential
psychologies, and the popular issue of consciousness-expanding, or
psychodelic drugs, might be traceable to the lack of definitive state-
ments within the field of psychology, to the disillusionment with
psychoanalytic absolutes, and to the great difficulty of understanding
current psychological orientations.
Returning to the problem of the acquisition of superstitions, we
shall no longer avoid the most simple declaration: that is, people may
acquire their superstitions by direct verbal communication, as well as
150
 THEORETICAL CONCEPTUALISATION OF SUPERSTITION

by first-handperceptionof fortuitouspairingsof events. It is asif each

culture presentsits participantswith a rule book which statesstandard
beliefs about what leads to what. These standard beliefs may be re-
garded as major premises that stand ready-under conditions of un-
certainty-to instantiatea potentiallythreateningor disturbingevent
in such a way as to bring it undercontrol.

5. The Maintenanceof Superstitions

Any objectwhich remainsuncontradicted is ipsofactobelievedand
positedas absolutereality-WilliamJames
It is the thesis of this section that if an individualalreadyhas an
explanationregardinga given event, that explanationwill tend to
stand until replacedby some 'stronger magic', and that this replace-
ment is contingenton severalconditions. It is commonplacein science
that a theory is rarely,if ever, negated by its own terms, but rather
must be replacedby a bettertheoryifitis to bereplacedat all. Perhaps
this is why Freudiantheory has endured. Volumes of criticismhave
been directed towards it, but nobody seems to have publicised a
competingtheory as generaland elaborateand dramaticfor the under-
standingof psychologicalcause-effectrelationships. Perhapssome of
the newer psychodynamictheories are just as general, but still the
Freudianmonument stands, and will continue to stand through its
own inertia, unlessit is challengedby a significantlystronger(more
dramatic,more need-fulfilling)theory. Littleis gainedin substituting
one belief for a pragmaticallyequivalentone.
One might point to a set of beliefsheld by laymen and by many
professionals in the mentalhealthfieldasanexampleof the maintenance
of a superstition. Some disorderedpersons(i.e. thosewhose behaviour
does not follow rules consideredappropriateby certainmembersof
the society) are 'diagnosed' as sick personsand are given a label-
schizophrenia. This labellingis coterminouswith what goes on when
a person is sick as the result of microbialinvasion, virus infection,
fracturedbones, etc. Becauseof superficialsimilaritiesin the conduct
of some disordered persons-disturbed speech, invalid judgments,
'hallucination', etc.--with the conduct of persons with known
disease processes(toxins, brain tumours, virus infections, etc.), the
illicit analogicalleap is made that all disordermay be accountedfor
through applicationof the disease model. The failure to establish
uniform diagnostic criteria for schizophrenia,the failure to find
i i 151
 KARL E. SCHEIBE AND THEODORE R. SARBIN

language and thought disorders substantially different from rates in

the normal population (when the effects of commitment and hospital-
isation procedures are controlled), and the arbitrarinesswith which a
person is ckssified as mentally ill, have not eliminated the belief that
disordered persons are 'sick', must be 'diagnosed', and subjected to
legal and medical procedures-presumably for the' cure' of the patient
(Sarbin, 1962). The operation of such a superstition to 'fill the gap
is epitomized in Cumming and Cumming's (1955) account of the in-
effectiveness of a community mental-health educational programme.

Li our attemptsto producea more permissiveclimatefor formermental

patients,we conveyed the ideathat they were prettymuch like everyone
else, and that therewas no sharpline dividing the sane from the insane,
but rathera continuousrange of behavior. In stressingthis idea we
were hammeringdirectlyat the core of the community'sown solution
to the problem of the mentallyill. Our problem was not theirs. We
were concernedwith the cure of the mentally ill, the people of Prairie
Town with the stability and solidarityof their own community ...
Prairie Town's pattern of beliefs and attitudes toward mental illness
was not merelya patchworkof half-truths,fallacies,and inconsistencies.
. . it played an important part in preserving the well-being of the
community and the peace of mind and self-esteem of the average
individual(pp. 63-64).
The basis for the longevity of superstitions need no longer be
shrouded in mystery. If we accept the general statement of the Law
of Effect-a response that is followed by reinforcements will probably
be employed again under the same or similar circumstances-then the
events that support superstitions may be ordered through the inference
model (Sarbin, et al. 1960). Certain events occur. They provide
sensory inputs. If these inputs cannot be readily instantiated(assignedto
a class) by the participantsin a culture, cognitive strain follows. This
is characterisedby large increasesin cognitive activity-an effort to find
a major premise that contains the same middle term as the so-far-unin-
stantiated input. Increasesin cognitive activity frequently' spill over
into physiological activity, sometimes identified as dysphoric affects.
The solution to the problem is through the employment of ritual or
other procedures dictated by the superstition. Filling the gap in know-
ledge through provision of a usable major premise reduces cognitive
strain. This may be reinforcing in either or both of two ways (I) it
may be arousal reducing (drive reduction in the Hullian sense) and (2)
the persons carrying out the ritual are given approval (generalised social
152
THEORETICAL CONCEPTUALISATION OF SUPERSTITION

reinforcement in the Skinnerian sense). The solution--no matter

how irrationalfrom our point of view-is thus self-perpetuating.
Following are some instancesof this generaltheme.
Among the Polynesiansit is very common to resolvethe questions
of guilt or innocenceof a defendantby spinninga coconut,andobserv-
ing the directionin which it comes to rest. It is common knowledge
that witch trials were often conducted by the similar method of
'ordeal', in which the accusedwas thrown into water, or otherwise
placed in a perilous situation. Survival was accepted as proof of
innocence, while death gave certaintyof guilt. After the ordeal is
over, the acceptedconclusionis a sourceof satisfactionto all concerned.
If the accusedsurvives,the accusersarerelievedthattheyhavenot taken
a life in vain. If the accuseddies, the actionsof the accusersare vin-
dicated, and again relief, or a reinforcingstate of affairs,is the con-
sequence.
Or considerthe exampleof prayerto believed-ingods. Since the
prayer is believed to be efficaciousin altering the course of future
events, the worshipperfeels relievedand comfortedimmediatelyafter
having prayed. He may now restbecausehe has placedcontrolin the
hands of a greaterpower. An appealhas been made to the powers
which controlall uncertainty,andhumanworry is no longernecessary.
Even if the prayeris not answeredthe prayeris still, on the whole, a
source of relief and comfort. Further,his actionsmay be approved
by significantothersin the community.
Or consider the more homely and current example of filter-tip
cigarettes. It is now fairlywell establishedthat the presenceof a filter
does little to alter the probabilityof smoking leading to lung cancer
(SmokingandHealth,1964). Yet filter cigarettessucceedin reducing
some of the tensionin smokersarousedby the cancerresearchfindings.
Additionalexamplescould be cited by the score, but the essential
point is that the superstitiondoes indeed have an effect, even though
the perceived causal relation is entirely false. Once installed, and
investedwith a high degreeofconfidence,the superstitionis completely
self-maintaining,and requiresno furtherconfirmationin the way of
fortuitouspairingsof events,so long as the superstitionis not subjected
to conflict and challengedby an alternativeand functionallymore
appropriatebelief.
The Hutteritecommunitiesof the easternUnited Statesprovide a
revealing example on this point1: the culture remainsquite distinct
1 The sourceof informationhereis a reporton' folk obstetrics' by Eaton(1958).

153
 KARL E. SCHEIBE AND THEODORE R. SARBIN
from that of the surrounding communities, and this distinction is
cherished by the Hutterites. A remarkable people, they have success-
fully resisted the intrusion of science into the majority of their customs,
and do not seem much the worse for this rigidity. In fact, their child-
birth traditions function so well for them, that while the average
mother has 1I children, a severe complication is very rare. The
Hutterites have numerous superstitions about the' evil eye' and about
sexual taboos, and these are accepted from generation to generation
without much question. Alcoholism, crime, dope-addiction, and
mental disorders are present only to a negligible degree. In fact, the
current problems of this culture can be traced directly to the partially
successful attempts of outsiders to have them institute more modern
practices in childbirth and child rearing. The beliefs of this culture
have worked for them in a very efficient and satisfying fashion and are
thereby self-sustaining. The truth of this statement is demonstrable
even though we would not hesitate to label a large number of their
beliefs as' superstitions'.
Of course, not all persistent superstitions have such an obvious
functional significance. Saunders (1958) reports that beliefs regarding
medical practices came from four widely disparate sources for the
Spanish-Americansof the south-west. These people seem to be able to
accept all of these beliefs, even though many of them seem contra-
dictory. Characteristically, they try their own folk-medicine first,
and if these techniques are ineffective, they simply turn to something
else. 'A remedy is tried, and if it works no surprise is evinced, since
that is what was expected. If it does not work, the failureis rationalized
and something else is tried' (Saunders, I958, p. 191). Clearly these
folk beliefs exhibit a fair amount of inertia in the face of challenge.
Beliefs which seem inconsistent and dissonant to us are not necessarily
so for the people who hold those beliefs.
Another basis for the maintenance of a superstition is the process of
insulation against or distortion of objective information to conform to
the superstitious commitment. For example, for some individuals
agreement between the conclusion of a syllogism and one's own beliefs
functions as a criterion forjudging the soundness ofsyllogistic reasoning
(Thouless, 1959; Morgan and Morton, 1944; Thistlethwaite, 195o).
Highly prejudiced individuals are resistant to change in general
(Rokeach, 1943) and individuals with highly conservative beliefs are
resistant to contrary information (Frenkel-Brunswik, 1949).
It has been experimentally demonstrated that the extent to which

I54
 THEORETICAL CONCEPTUALISATION OF SUPERSTITION
beliefs about matters of fact deviate from the truth is highly related to
the level of information the respondents possess about those facts.
Moreover, the distortions in beliefs are consistent with the evaluative
significance assigned by the respondents to those beliefs. That is,
individuals tend to believe what they want to believe where lack of
information makes this possible (Scheibe, I963).
In sum, superstitious beliefs seem to have a fair amount of stability.
If they are not challenged they are maintained. Even if moderate
degrees of challenge are made to established superstitions,the individual
is likely to prefer distorting facts or logic to discarding the superstition
(Cohen, Brehm and Latane, I959). Of course, there are interesting
individual differences in this regard: 'The passion for unity and
smoothness is in some minds so insatiate that, in spite of the logical
clearnessof... reasonings and conclusions, many will fail to be influen-
ced by them' (James, I, I890, p. I62).

6. AnotherFunctioinof Superstitions
A tertiary functional significance of muncertaintymay give rise to
superstition. James has said that uncertainty must exist in order for
man to have a sense of personal significance. Many 'classical'
psychologists would agree with him. (For instance, France, 1902).
The essence of Kierkegaard's philosophy is that man must cope with
uncertainty and unpredictability. If man were to have true vision of
cause and effect relationships, there would follow, according to some,
a complete abrogation of human responsibility. It is the uncertainties
which give life its vigour and interest. Others of a less romantic
persuasion view this line of thinking as a severe restriction on the
progress of scientific knowledge, and openly propose that we seek to
establish a completely controlled and predictable universe. The
flavour of this issue is conveyed in this statement by Skinner (Rogers
and Skinner, 1956).
A world in which people are wise and good, without trying, without
'having to be' or without 'choosing to be', could conceivably be a
far betterworld for everyone. In such a world we should not have to
'give anyone credit '-we should not need to admire anyone-for
being wise and good. From our present point of view we cannot
believe that such a world would be admirable. We do not permit
ourselvesto imagine what it would be like (p. 10o60).
We do not propose to discuss this issue here. However, it should
serve as a caution to those who would cavalierly declare that human
11 I155
 KARL E. SCHEIBE AND THEODORE R. SARBIN

beings shouldbe relievedof theiruncertaintiesand of the superstitions

which are theirproduct.
To recapitulate:
First,a superstitionmay be consideredto be a specialkind of ex-
pectancythat can be codified as a majorpremise. Startingwith the
assumptionof objectiveuncertainty,it hasbeenarguedthatsome extra-
logical propositions,suchas superstitions,arenecessaryguidesto action.
Thoughinitiallysuperstitionsmay be acquiredin a manneranalogous
to the processof operantconditioning,sucha model doesnot do well in
describingmost superstitions,especiallythose for which the original
confirmationis historicallyobscure. We havesuggestedthe possibility
that superstitionsmay be acquiredby direct tuition on condition of
uncertaintyof outcomes,where the outcomesare highly valued.
If a superstitiongoes unchallengedby another and 'stronger
superstition,it will not just wither away. Superstitionsdo have the
effectof reducing coglnitive strain andthusareself-maintaining. They
alsoexhibita high degreeof inertia,in thatonce they areinstalled,they
then remainuntildislodgedby a muchstrongerbelief. Individualsand
culturesmay differin theextentto whichthey cantoleratetheambiguity
of contradictorysuperstitions.
It shouldbe clearthatthe necessaryconditionsfor the applicationof
the superstitionlabel are located bothin the actor and in the observer.
If the actorhasan operativebeliefthatthe observerbelievesto be false,
then the observermay say that the actor'sbelief is a superstition. If
the beliefsof the observerare not derivedfrom empiricalor inductive
procedures,his allegationthata belief is a superstitionwill not hold up
amongstscientists. We, as scientists,are wont to labelas superstitions
the beliefsthat differfrom scientificallysupportedpropositions.
How about scientifictheoriesabout which there is disagreement?
Or how about scientistswho persistin believing discreditedtheories?
Shouldthe label' superstition' be appliedin thesecases? In one sense,
it certainly should. Beliefs in invalid theories that belong under the
institutionalaegis of Science are of precisely the same logical and
psychologicalsort as beliefsin the most primitiveand mundaneritual.
But this bringsup the point of the socialpsychologicalsignificance
of calling a belief superstitious. Since the term superstitioncarriesa
strong pejorativemeaning in our culture, one may run the risk of
offendinganotherif he chargeshim directlywith being superstitious.
In actualpracticeit is not the custom to apply the term to individuals
with whom some social contact must be maintained,or who are

I56
 THEORETICAL CONCEPTUALISATION OF SUPERSTITION

presentin person. The termnis usuallyappliedto absentpersons,to

the hypotheticalMasses,or to distantcultures. This usageconvention
involves little risk or retribution,and is properlydiscreet.
A final point shouldbe made that may seem surprising,in view of
the title of this paper. Insteadof using the term superstitionmore
frequently,as would be necessaryif we were to apply the term at all
consistently,it should be used less frequently,and perhapsnot at all.
Accordingto the major points outlined earlierregardingthe psycho-
logical nature of superstitionsas a category of beliefs, it should be
apparentthatthey areubiquitous;thateven scientistshavethemthough
here they are often clothed in the dignified fabric of irreproachable
mathematicalor verbal symbolism. Moreover, superstitions,i.e.
beliefs that are not inductively derived and not empiricallychecked,
are psychologicallynecessaryfor existence. Thus, even though the
category of beliefs called superstitionsis a distinct and meaningful
category, it turnsout to be too inclusivea classof beliefs to bear the
socially reprehensiblelabel of Superstition. For the sake of equity
and fairness, the term should be used not more, but less.

Wesleyan University and University of California,

Berkeley

REFERENCES

A. R. Cohen, J. V. Brehm and D. Latane, 'Choices of Strategy and Voluntary

Exposureto Information',J. Personality,1959, 27, 63-73-
J. Cohen, Chance,Skill andLuck,Penguin Books, I960.
Elaine Cumming andJ. Cumming, ' Mental IealthEducationin a CanadianCom-
munity ' in B. D. Paul(ed.) Health,Culture,andCommunity,RussellSageFounda-
tion, New York, I955.
F. C. Dockeray and W. L. Valentine. An Analysis of the ElementaryPsychology
Course at Ohio State University,J. Appl. Psychol.,1935, 19, 503-520.
J.Eaton,' Folk Obstetricsand Pediatrics: A Case Study of SocialAnthropologyand
Medicine'inJaco,E. (ed.),Patients, andIllness. FreePress,Glencoe,
Physicians,
1958.
C. J. France,'The GamblingImpulse', Amer.J.Psychol., 1902, I3, 364-407.
J. G. Frazer,TheGoldenBough: A Studyin MagicandReligion(abridgeded.),Mac-
millan,LondonandNew York,I925.
Else Frenkel-Brunswik,'Intoleranceof ambiguityas an emotionaland perceptual
personalityvariable'. J. Personality,
I949,I8, Io8-143.
of CollegeStudents', J. Abn.& Soc.
A. R. Gilliland,'A Studyof the Superstitions
Psychol., I930, 24, 472-479.
C. S. Hall and G. Lindzey, Theoriesof Personality,Wiley, New York, I957.

157
 KARL E. SCHEIBE AND THEODORE R. SARBIN
F. Heider, 'Social Perceptionand PhenomenalCausality', Psychol.Rev., 1944, 51,
358-374.
W. James,Principlesof Psychology,Vols.I andII, Henry Holt, New York, 1890o.
J.Jastrow,FactandFablein Psychology,Macmillan,New York, 19oo.
R. Lewinsohn,ProphetsandPrediction, Secker & Warburg,London, I96r.
B. Malinowski, 'Anthropology', Encyclopedia Britannica,First SupplementaryVol-
ume, London and New York, 1926.
R. K. Merton, Social TheoryandSocialStructure. Glencoc,Ill. : FreePress,I957.
of Causality(trans.by T. R. andElaineMiles),Methuen,
A. E. Miclhotte,ThePerception
London, 1963.
J.J.B. MorganandJ. T. Morton,' The Distortionof SyllogisticReasoningProduced
by PersonalConvictions', J. Soc.Psycho!l.,1944, 20, 39-59.
W. H. Morse and B. F. SkiInner,' A Second Type of Superstitionin the Pigeon',
Amer.J. Psychol., 1957, 70, 308-311.
of Realityiil the Clhild(trans.by MargaretCook), Basic
Jean Piaget, The Construction
Books, New York, I955.
Carl R. Rogers and 13.F. Skinner,'Some IssuesConcerningthe Control of Human
Behavior : A Symposium', Science, 1956, 123, lo57-1o66.
M. Rokeach,' GeneralizedMentalRigidit) as a Factorin Ethnocentrism', J. Abn. &
Soc. Psychol., 1948, 43, 259-278.
J.B. Rotter, SocialLearningand ClinicalPsychology,Prentice-Hall,New York, I954.
T. R. Sarbin,R. Taft and D. Bailey, ClinicalIifcrenceand CognitiveTheory,Holt,
Rinehart & Winston, New York, 1960.
T. R. Sarbin,Role-TheoreticalAnalysisof Schizophrenia,Ms. 1962.
L. Saunders,'Healing Ways in the SpanishSouthwest'. In E. Jaco (ed.), Patients,
Physicians,andIllness,FreePress,Glencoe, 1958.
K. E. Scheibe, 'On the Relationship Between Beliefs and Values, unpublished
doctoraldissertation,University of California,Berkeley, 1963.
B. F. Skinner,' Superstitionin the Pigeon ',J. Exp. Psychol.,1948, 38, 168-172.
SmokingandHealth,reportof the Advisory Committeeto the SurgeonGeneralof the
Public Health Service,1964, Washington,D.C.
D. Thistlethwaite,'Attitude and Structureand Factorsin the Distortionof Reason-
ing ', J. Abn. & Soc. Psycho!., 1950, 45, 442-458.
R. H. Thouless, 'The Effect of Prejudiceon Reasoning', Brit.J. Psychol.,1959, 50,
289-293.
E. C. Tolmnan,BehaviorandPsychological Man, University of CaliforniaPress, 1958.
E. B. Tylor, Primitive
Culture
(2 vols.), Holt, New York, 1888.
R. R. Willoughby,'Magic and CognatePhenomena: An hypothesis',in C.
(ed.),Handbook
Murclhison ClarkUniversityPress,Worcester,
of SocialPsychology,
Mass., 1935.

rS8