You are on page 1of 26

International Journal of Public Opinion Research Vol.  No.  oµj¡–:8µ:/µµ $¡.

Dietram A. Scheufele and Patricia Moy
Numerous studies have been conducted on the spiral of silence since Noelle-Neumann
(+µy¡) formulated the theory a quarter of a century ago. As a whole, these studies draw
upon different conceptualizations, employ inconsistent operationalizations, and give
short shrift to important macroscopic variables. Such inconsistencies potentially account
for substantial proportions of the variance in spiral of silence effects. This paper
examines these three areas in greater detail. First, we review key assumptions and
theoretical statements of the spiral of silence. Second, we examine how these conceptual
issues translate into operational ones. Finally, we outline areas that have remained
largely unexplored over the last :j years. Specifically, we contend that spiral of silence
studies in different cultures have failed to take into account culture-specific variables
that may mitigate the importance of opinion perceptions as predictors of individual
behavior or attitudes. In other words, cross-cultural differences are key factors in
predicting speaking out, the key dependent variable in spiral of silence research. As a
result, we call for the return to a more macroscopic focus in spiral of silence research.
When Noelle-Neumann (+µy¡) formulated the spiral of silence theory, she in
essence posited that an individual’s willingness to express his or her opinion
was a function of how he or she perceived public opinion. After all, individuals’
ideas, attitudes, and behaviors are often influenced by their perception of what
others do or think (Cooley +µjô). As O’Gorman and Garry (+µyô, p. ¡¡µ)
noted: ‘These cognitive attributions . . . may be widespread, firmly held and
provide common understanding, but their accuracy is not to be assumed.’
Researchers have examined individual perceptions of public opinion and their
impact on political behaviors and attitudes (for an overview, see Mutz +µµ¡),
∗ A previous version of this paper won the MAPOR Fellows Student Paper Competition at the annual
meeting of the Midwest Association for Public Opinion Research, Chicago, IL, November, +µµ8.
© World Association for Public Opinion Research 
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn ¡
and some have concluded that ‘the spiral of silence is the only theory of public
opinion possessing the attributes of depth, range, and to more limited extent,
accuracy’ (Neuwirth +µµj, p. ++). Since its inception a quarter of a century
ago, the spiral of silence has generated considerable empirical research with
inconsistent findings, leading Glynn, Hayes, and Shanahan (+µµy) to examine
the key relationship between perceptions of majority opinion and an individual’s
willingness to express his or her opinion. Their meta-analysis of over :o
empirical studies revealed a small but significant effect.
Contradictory results and the inconsistently strong findings across spiral of
silence studies have their origin in a number of sources: (+) conceptual problems,
e.g. inconsistencies in how key concepts in the spiral of silence are explicated;
(:) problems regarding the measurement of key variables in the process; and
(¡) varying levels of attention paid to testing important macroscopic variables.
This paper examines all three areas in greater detail, looking back over the
history of spiral of silence research. First, we review key assumptions and
theoretical statements of the spiral of silence, grounding the theory in public
opinion and situating it in the more general framework of theories on opinion
perceptions. Based on this framework, we address problematic assumptions and
theoretical statements in spiral of silence research. Second, we examine how
these conceptual issues translate into concrete operational ones. Finally, we
outline areas left largely unexplored by previous research on the spiral of silence.
Specifically, we address a concern raised by McLeod and Blumler (+µ8y) who
argued for a more macroscopic focus in communication sciences that will enable
researchers to make cross-cultural comparisons. While we argue that inconsistent
findings across studies in different cultures (for an overview, see Glynn et al.
+µµy) stem from the inapplicability of a given measure in different cultures or
general operational differences among studies, cross-cultural research suggests
that cultures differ in their socially shared meanings, norms of interpersonal
relationships, and conflict resolution. Hence, the question for spiral of silence
research becomes: to what degree can culture-specific variables offset or increase
the importance of opinion perceptions as predictors of individual behavior or
attitudes? As a result, we call for a return to a more macroscopic focus in spiral
of silence research.
Against a backdrop of philosophers and researchers grappling with the task of
defining public opinion (see Davison +µô8 for an overview), Noelle-Neumann
(+µµj) differentiates two concepts of public opinion: (+) public opinion as
rationality, which makes it ‘instrumental . . . in the process of opinion formation
and decision making in a democracy’; and (:) public opinion as social control,
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr j
where ‘its role is to promote social integration and to ensure that there is a
sufficient level of consensus on which actions and decisions may be based’
(Noelle-Neumann +µµj, p. ¡¡).
Pturr c Orr ×r o× +s R++r o×+rr +v
Public opinion, according to a rational model, results from political raisonnement
in the public sphere (Habermas +µô:). In its purest form, public opinion as
rationality ‘is the social judgment reached upon a question of general or civic
import after conscious rational public discussion’ (King +µ:8, p. xxiii). Similarly,
the rational formation of public opinion occurs when many or most members
of the citizenry agree on judgments that individuals have arrived at, due to
either reflection or knowledge of an issue (Palmer +歓). The rational model is
based on the notion of an enlightened, rational public that is willing to and
capable of participating in political processes (Childs +µôj, Wilson +µ¡¡). In
this sense, the rational concept of public opinion is a necessary condition for
generating social change.
Pturr c Orr ×r o× +s Socr +r Co×+aor
It is, however, the model of public opinion as social control (Bentham +8¡8–
+8¡¡/+µô:) that underlies the spiral of silence theory. Because successful social
systems ‘must have ways to institutionalize consensus’ (Csikszentmihalyi +µµ+,
p. :8µ), they threaten individuals with social isolation in order to ensure these
necessary levels of cohesion (Noelle-Neuman +µµj). At the same time, individuals
constantly scan their environment for present and future distributions of public
opinion ‘in order to see which opinions and modes will win the approval of
society and which will lead to their isolation’ (Noelle-Neumann +µµj, p. ¡:).
Public opinion as social control is thus defined as opinions that can be expressed
without risking sanctions or social isolation, or opinions that have to be expressed
in order to avoid isolation (Noelle-Neumann +µ8¡).
Rrco×cr rr ×o +nr Co×crr+s
Childs (+µôj) wrote that ‘it is rather unlikely that students of public opinion
will ever discover a method for separately identifying the rational as distinguished
from the irrational opinion, for the process of opinion formation and expression
is a complicated and mysterious process’ (p. +8). However, recent work makes
it possible to identify at least three distinct differences between the two models
of public opinion (Merton +µô8, Noelle-Neumann +µ8¡, +µµ¡, +µµj).
First, the rational model of opinion formation and the model of social control
differ substantially in their modes of opinion expression. The former views
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn ô
public opinion in terms of political discussion and decision making, and
emphasizes verbal expression of rational arguments or opinions, i.e. raisonnement
in the public sphere. Public opinion as social control, on the other hand, works
through various forms of opinion expression like verbal and facial expressions,
gestures, and publicly visible symbols. The display of yard signs, bumper
stickers, or campaign buttons, often used as a measure of political participation
(e.g. Verba et al. +µµj), falls under the spiral of silence model—at least in those
cases where the intention is to influence perceptions of opinion distribution
rather than to convey a political message.
Second, the two models differ with respect to the effort required for opinion
expression. In a rational model, participation in political discussion and decision
making is a function of two groups of individual-level factors: the ability to
acquire information about and discuss political issues rationally and the motivation
to discuss these issues with others. Both ability and motivation require conscious
effort. As a result, ‘although all citizens may potentially participate in the
discussion, there is in fact only a small group of informed and interested citizens
who actually do participate’ (Noelle-Neumann +µµj, p. ¡¡). The model of
public opinion as social control, however, views processes of opinion formation
and expression as mostly unconscious. Individuals scan their environment for
present and future distributions of opinion to a large degree unconsciously.
Speaking out in this model is therefore less of a conscious decision to enter a
discussion than an unconscious reluctance to express one’s opinion.
Finally, the two models of public opinion differ with respect to the con-
ceptualizations of ‘public’. ‘Public’ in a rational model of public opinion always
refers to a particular group of politically interested and knowledgeable citizens
in the sense of a public sphere (Habermas +µô:, Noelle-Neumann +µ8¡, Wilson
+µ¡¡). In contrast, the model of social control defines ‘public’ as involving
everybody; thus public opinion, or rather public pressure, affects all members
of society. The spiral of silence clearly falls under the model of public opinion
as social control. It can also be viewed as a theory of social perception.
Perceptions of public opinion matter not only because individuals attend to
their social environment, but also because these perceptions potentially influence
individual behavior and attitudes.
Wnv I×nr vr nt+rs A++r×n +o +nrr a Socr +r E×vr ao×·r×+
If human beings are naive psychologists as Heider (+µ¡o, +µy8) concluded, then
they cannot understand the world in all its complexity. As a result, they constantly
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr y
observe their environment to make inferences from sensory information to
the underlying causal relationships. Indeed, subjects exposed to movies with
abstract movements of geometrical shapes not only tended to interpret these
movements as actions of human beings, but also made inferences to certain
underlying motivations (Heider and Simmel +歭).
How do people process and organize incoming information? Drawing upon
Gestalt psychology (e.g. Koffka +µ::, Wertheimer +µ:j), Sherif (+µôy) argued
that individuals use frames of reference, the most important one being social
groups. His argument, supported by experimental evidence, that individuals
use their social environment as a frame of reference for interpreting new
information has important implications for public opinion research. Under the
assumptions of the Gestalt psychology, individuals feel uneasy if faced with
information without an appropriate frame of reference by which to interpret it.
Public opinion or judgments of others therefore provides the most powerful
frame of reference. One subject in Sherif’s (+µôy) experiments summarizes the
power of public opinion as a frame of reference nicely: ‘If you tell me once
how much I am mistaken, all my judgments will be better’ (Sherif +µôy,
p. +¡ô).
Wnv Pracrr+r o×s M+++ra
The spiral of silence theory rests on the assumption that individuals constantly
scan their environment in order to assess the climate of opinion, i.e. the
aggregate distribution of opinions on a given issue. This assumption aligns the
spiral of silence with other theoretical approaches that consider public opinion
a process and provide information on how that process operates (Glynn et al.
Their accuracy notwithstanding, perceptions—or misperceptions—of opinion
climates matter from a normative standpoint as an increasing body of literature
points to the power of perceptions in shaping, among other things, one’s
perception of the self, emotions, judgments, and behavior (Miller and Prentice
In the political realm alone, numerous studies have examined the effects of
opinion polls on voter turnout and the direction of the vote (e.g. Ansolabehere
and Iyengar +µµ¡, Atkin +µôµ, de Bock +µyô, Cook and Welch +µ¡o, Lavrakas
et al. +µµ+, Noelle-Neumann +µµ¡, Skalaban +µ88, West +µµ+). For instance,
experimental data from Atkin (+µôµ) and Mutz (+µµ:) revealed that (fictitious)
poll results influenced preferences of candidates and issues. Bartels’ (+µ88)
study of momentum in presidential primaries showed that prospective voters’
perceptions of the candidates’ popularity influence their vote choice. Similarly,
other researchers have studied the impact of exit polls and early election returns
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn 8
on voter turnout and vote choice (e.g. Delli Carpini +µ8¡, Sudman +µ8ô).
Evidence from these studies, however, is inconsistent, even in the case of
bandwagon effects, i.e. effects based on people’s desire to be on the winning
side (e.g. Gartner +µyj, McAllister and Studlar +µµ+, Navazio +µyy, Straffin
+µyy, Zech +µyj).
Effects of opinion perceptions on vote choice aside, a second group of variables
deserves closer attention: opinion expression. Experimental findings by Latane´
and Darley (+µô8) very clearly demonstrated the importance and societal
relevance of speaking out. Subjects who worked alone on a questionnaire were
more likely than subjects who worked in groups to report the presence of smoke
that had been induced into the room in which they worked. This finding is
especially striking, considering that the experimental situation suggested po-
tential physical threat. Still, this concern for physical well-being was obviously
offset by what Miller and Prentice (+µµ¡) call ‘social fear’ or the ‘fear of
appearing naive or foolish’ (p. j¡j) if their concerns proved to be unjustified.
Nisbett and Kunda (+µ8j) suggest that even in cases of discrepancies between
one’s own opinions and public opinion, the mere knowledge of this discrepancy
could lead an individual ‘to steer away from controversial topics so as to avoid
offending others who hold different opinions’ (p. :µy). This notion is the same
one underlying Noelle-Neumann’s model of the spiral of silence.
With an emphasis on the formation, functions and effects of public opinion,
the spiral of silence presents an approach that integrates what some consider a
fractured concept of public opinion and offers the possibility to test it empirically
(Salmon and Kline +µ8j, p. ¡). This section outlines the assumptions and key
theoretical statements of the spiral of silence, and evaluates conceptual issues
that have arisen since the theory was introduced.
Asst·r+r o×s
The assumptions of the spiral of silence can be outlined in the form of five
major hypotheses
(Noelle-Neumann +µµ+).
+ Threat of Isolation: ‘In the social collective cohesion must be constantly ensured
by a sufficient level of agreement on values and goals’ (Noelle-Neumann +µµ+,
This description follows the most recent publications of Noelle-Neumann (+µµ+, +µµ¡, +µµ¡), which
differ significantly from the original publication (Noelle-Neumann +µy¡) in the systematic description of the
premises and necessary conditions underlying the process of the spiral of silence.
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr µ
p. :j8). To guarantee this agreement, society threatens with isolation those
individuals who violate the consensus.
: Fear of Isolation: The formation of individual opinion and action is char-
acterized by individuals’ fear of becoming ‘social isolates’ (Sanders et al.
+µ8j, p. xvi). This second assumption of the spiral of silence derives
from experimental studies on conformity.
Social conformity can be either
informational social influence, reflected in individuals accepting information
from others as evidence about reality, or normative societal influence, in
which individuals ‘conform with the . . . expectations of others’ (Deutsch
and Gerard +µjj, p. ô:µ). Noelle-Neumann (+µµ¡) uses the latter to explain
and provide evidence for the assumed impact of fear of isolation on willingness
to speak out. She refers to experiments in which subjects conformed with
the majority in performing relatively straightforward tasks such as selecting
a line that matched another in length (Asch +µjj, +µôj) and selecting the
longer of two acoustic tones (Milgram +µô+). Noelle-Neumann (+µµ¡) uses
the fact that subjects in Asch’s experiment saw ‘with their own eyes that the
line selected by the majority as the best match is not the best match’ as
evidence that fear of isolation is the dominant factor influencing conformity
(Noelle-Neumann +µµ¡, p. ¡o).
¡ Quasi-statistical Sense: As a result of fear of isolation, individuals constantly
monitor their environment to check on the distribution of opinions as well
as the future trend of opinion. Such monitoring can involve attending to
media coverage of an issue, direct observation of one’s environment, or
interpersonal discussion of issues.
The quasi-statistical sense is probably the most widely misinterpreted
concept in the spiral of silence. Critics have challenged the notion that a
quasi-statistical perception of the climate of opinion is usually accurate. For
example, some have suggested that ‘Noelle-Neumann’s model predicts the
complete accuracy of a respondent’s perception of majority opinion’ (Neuwirth
+µµj, p. ¡y; see also Salmon and Kline +µ8j, Neuwirth and Ilundain +µ8¡)
and therefore cannot account for phenomena such as pluralistic ignorance or
As much of the literature on conformity uses the terms ‘conformity’ and ‘compliance’ interchangeably,
we follow Hardy’s (+µjy, p. :8µ) definition of conformity that is equivalent to the definition underlying
Noelle-Neumann’s assumptions: ‘Conformity is defined as the public avowal of a belief or attitude at variance
with one’s prior position, which avowal tends to correspond to the position approved by the group in which
the avowal occurs’.
Whereas Noelle-Neumann used the term ‘quasi-statistical organ’ in her early writings (+µy¡), her later
pieces address broadly a ‘quasi-statistical ability to perceive the climate of opinion’ (+µµ¡, p. +j). These two
sources of perceptions have to be understood in this broader sense as perceptions of reality through mass
communication, on the one hand, and interpersonal communication and observation, on the other hand
(+µµ¡, p. ::µ).
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn +o
the looking-glass perception.
As Salwen, Lin, and Matera (+µµ¡) point out,
however, ‘the hypothesis focuses on people’s perceptions of the climate of
opinion, rather than a real climate of opinion’ (Salwen, Lin, and Matera
+µµ¡, p. :8¡, emphasis in original). Misperceptions of public opinion like
pluralistic ignorance or the looking-glass perception are an integral part of
the spiral of silence theory with respect to phenomena like the dual climate
of opinion (Noelle-Neumann +µµ¡), i.e. the misrepresentation of public
opinion in mass media. However, Glynn et al.’s (+µµô) meta-analysis identified
a relatively large number of ‘researchers [who] fail to operationalize perceived
congruency as such, instead focusing on ‘‘objective’’ congruency’ (Glynn et
al. +µµô, p. ++, emphasis in original).
¡ Willingness to Speak Out and Tendency to Remain Silent: Individuals tend to
publicly express their opinions and attitudes when they perceive their view
to be dominant or on the rise. In contrast, when people sense their view is
in the minority or on the decline, they become cautious and silent.
j Spiral of Silence: The interaction of these four factors leads to a process of
formation, change and reinforcement of public opinion. ‘The tendency of
the one to speak up and the other to be silent starts off a spiraling
process which increasingly establishes one opinion as the prevailing one’
(Noelle-Neumann +µy¡, p. ¡¡). This process is illustrated in Figure +.
Over time, changing perceptions of the opinion climate influence people’s
willingness to express minority opinions and establish one opinion as the
predominant one. Public opinion is transformed from a morally-loaded
question or from the ‘liquid’ state (To¨nnies +µ::) to a ‘solid’ norm or dogma
(Noelle-Neumann +µ8¡).
Three additional factors need to be taken into account to fully explain the
process of the spiral of silence.
+ Moral Component of Public Opinion: The process of the spiral of silence only
works for issues with a moral component, or value-laden issues ‘by which
the individual isolates or may isolate himself in public’ (Noelle-Neumann
+µµ¡, p. :¡+). It is only from that moral or normative element that public
opinion asserts its threat of isolation (Noelle-Neumann +µ8¡, p. +¡+).
Studies making this criticism generally do not refer to Noelle-Neumann’s original theorizing but to
Donsbach and Stevenson’s (+µ8ô) claim that people are capable of ‘perceiving the distribution of opinions
in their environment quite accurately’ (p. ¡).
This does not mean that unanimity of opinion is the ‘logical conclusion’ (Glynn et al. +µµô, p. :j) of
the spiral of silence model. ‘Hardcores’, individuals who resist public pressure or lack the ability to perceive
or attend to public pressure, and ‘avant-gardes’, individuals who introduce and voice new ‘minority’ ideas,
explicitly contradict this notion (Noelle-Neumann +µµ¡).
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr ++
Frotar + The process of the spiral of silence
: Time Factor:
The perceived climate of opinion and its future development
are critical factors in the spiral of silence. Which point of view is dominant
or likely to become dominant, and which is held by a minority and likely to
lose? This perception, accurate or not, influences individuals’ willingness to
speak out, which in turn influences their perception of the climate of opinion.
The result is a spiral process which establishes one opinion over time as the
predominant public opinion (Noelle-Neumann +µyµ, +µµ:).
¡ Role of the Media: The processes described above can only take place if the
media take an identifiable position in the conflict (Donsbach +µ8y). Any
scientific approach to the spiral of silence disregarding the media as a critical
factor ‘refute[s] the spiral of silence theory whenever the tone of the media
diverges greatly from public opinion’ (Noelle-Neumann +µµ¡, p. :oo).
To summarize, if public opinion is a form of social control and involves the
perceptions of others, then it can be defined as ‘attitudes or behaviors one must
express in public if one is not to isolate oneself; in areas of controversy or
Allport (+µ¡y) points to the importance of the process character of public opinion, i.e., to ‘view the
phenomenon as a process with a time dimension’ (p. +ô).
Neuwirth’s (+µµj) assumption that ‘once a spiral is set in motion, all persons favoring the minority
viewpoint eventually should fall silent’ (p. :j) is not justified by Noelle-Neumann’s theoretical assumptions.
The spiral of silence establishes one opinion as the prevailing one, not as the only one. According to
Noelle-Neumann (+µµ¡), there is always one minority ‘that remains at the end of the spiral of silence process
in defiance of the threats of isolation’ (p. +yo).
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn +:
change, public opinions are those attitudes one can express without running
the danger of isolating oneself’ (Noelle-Neumann +µµ¡, p. +y8). Human behavior,
particularly the willingness to speak out, is ‘heavily directed by a fear of isolation
that makes . . . sanctions of denial of sympathy, and so forth, very powerful
forms of influence’ (Glynn and McLeod +µ8j, p. ¡y).
Willingness to Speak Out
Criticism of the spiral of silence as a macro theory has mainly focused on three
areas (for an overview, see Donsbach and Stevenson +µ8ô). First, researchers
have questioned whether fear of isolation adequately and sufficiently explains
one’s willingness to speak out in experimental and survey designs (Glynn and
McLeod +µ8j). Others have suggested factors other than fear of isolation that
potentially influence people’s willingness to speak out (Lasorsa +µµ+, Salmon
and Neuwirth +µµo). Second, studies examining the spiral of silence empirically
have criticized Asch’s (+µjj, +µôj) and Milgram’s (+µô+) conformity experiments
as providing an adequate basis for the spiral of silence theory (e.g. Glynn and
McLeod +µ8j, Price and Allen +µµo, Salmon and Kline +µ8j). Third, and
somewhat related, researchers have suggested that cues about opinion dis-
tribution may come from sources other than the national climate of opinion
(e.g. Glynn and Park +µµy).
These criticisms need to be reevaluated as recent studies have responded in
part to these critiques and investigated fear of isolation in greater detail (e.g.
Hallemann +µ8µ, Noelle-Neumann +µµ¡). In addition, some arguments identify
less of a substantive problem, suggesting that the theoretical framework of the
spiral of silence has been misinterpreted. The identification of deficits in the
spiral of silence and its theoretical premises requires examining criticisms of
the experimental evidence and fear of isolation.
Criticism of Experimental Evidence
Criticisms of Asch’s conformity experiments as appropriate experimental evidence
for testing the hypotheses formulated in the spiral of silence (Glynn and McLeod
+µ8j, Salmon and Kline +µ8j) concentrate on six aspects, all of which involve
external validity and refer to Hovland’s (+µjµ) notion of a captive audience.
First, subjects in the experiment were exposed to unambiguous stimuli, asked
to differentiate between objectively observable ‘right’ and ‘wrong’ answers
(Pollis and Cammalleri +µô8, p. yo).
This point is especially important given
Crutchfield (+µjj) provides experimental data which includes, among other judgments, ‘estimates of the
opinions of others, [and] expression of his own attitudes on issues’ (p. +µ+). The external validity of his
findings, however, is limited in that all subjects were engaged in a profession in which leadership is one of
the expected qualifications. Moreover, subjects did not have direct face-to-face contact with each other.
Argyle (+µjy) points to the influence that public, i.e. face-to-face experimental situations have on the results
of tests of conformity.
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr +¡
Noelle-Neumann’s (+µµ:, +µµj) distinction between public opinion as rationality
and public opinion as social control. With respect to conformity and the model
of social control, ‘public opinion does not depend on what is right or wrong’
(Noelle-Neumann +µµj, p. ¡¡) but on what is good or bad. Accordingly, it is
this moral component that makes public opinion so powerful in inducing
Second, important independent variables like interpersonal communication
between the experimental subjects during the experiment were not adequately
controlled for (Glynn and McLeod +µ8j; see also Price and Allen +µµo, Salmon
and Kline +µ8j).
Third, Asch’s experimental settings with one stimulus at a single point of
time cannot explain the process character of public opinion, from which the
spiral of silence takes its name.
Fourth, researchers have suggested that positive motives like being on the
winner’s side can explain conformist behavior (Glynn and McLeod +µ8j,
Salmon and Kline +µ8j). Others have argued that the conformity rate in
Asch-type experiments is a result of two opposing forces—a pressure to conform,
and the pressure to remain independent (Ross et al. +µyô).
Fifth, good theories should be able to cross national boundaries (Salmon and
Kline +µ8j). Milgram’s (+µô+) tests of conformity showed higher levels of
conformity among Norwegians than the French. Expanding on this research,
Meeus and Raijmakers’ (+µ8µ) experiment found relatively consistent results
for the USA, Europe, Australia, and the Far East. There is no evidence,
however, that these experiments can be considered an operationalization of the
concept of conformity.
Finally, the extent of conformity in Asch’s experiment clearly depended on
the unanimity of majority pressure. When subjects were supported by a single
confederate, conformity rates dropped to one-fourth the level of that under
conditions of unanimity (Asch +µjj). These data suggested that there are
‘sufficient grounds for concluding that unanimity—that is, the sequence of
responses reflecting interindividual consistency—is more important than the
mere number of persons adopting a common response’ (Moscovici +µ8j,
p. ¡ô¡).
Criticism of Fear of Isolation as the Only Motivation
The second area of criticism focuses on the fear of isolation as the single factor
in explaining an individual’s willingness to speak out. Researchers have examined
various contingent conditions other than fear of isolation under which individuals
are in fact willing to express their opinions in public (Lasorsa +µµ+, Neuwirth
+µµj, Noelle-Neumann +µy¡, Salmon and Neuwirth +µµo). The problem is
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn +¡
that while Noelle-Neumann (+µy¡) does not provide a systematic control for
antecedents of the willingness to speak out other than fear of isolation, others
have controlled for other factors but not for fear of isolation (Lasorsa +µµ+,
Salmon and Neuwirth +µµo, Weimann +µµ+, +µµ¡).
Reference Groups vs. Anonymous Publics
Finally, critics have questioned whether reference groups affect one’s willingness
to express an opinion (Glynn and Park +µµj, +µµy, Oshagan +µµô). This
criticism, to some degree, is related to the criticism of Asch’s experiments on
conformity. As described earlier, a single person agreeing with the subject’s
judgment was enough for the subject to withstand majority pressure. However,
this criticism addresses only one of two problems regarding reference groups.
First, do reference groups influence an individual’s fear of isolation or is fear
of isolation, as Noelle-Neumann (+µy¡) argues, a personality trait? Second, and
perhaps more importantly, can reference groups serve as proxies for the climate
of opinion? In other words, do people make inferences from their reference
groups to the general climate of opinion? Are the opinion distributions within
reference groups a better predictor of the willingness to speak out than the
national climate of opinion?
Based on the premises outlined and the criticisms raised over the last quarter
of a century, we outline a research agenda for future spiral of silence research
in the following sections. This agenda includes, in the first section, an outline
of key variables, including the criterion ‘opinion expression’, and related
measurement issues. In the second section, the importance of reference groups
is examined in the process of the spiral of silence. The focus here is on
differences within a given society with respect to the social groups or entities
that are used as reference for climate assessment. Finally, the spiral of silence
is examined in the context of cross-cultural research.
Srr+kr ×o Ot+ +s +nr Krv V+ar +urr: Tnr Co×crr+ +×n I+s
Orra++r o×+rr z++r o×
Various operationalizations of public opinion expression on controversial issues
have been suggested (Glynn and McLeod +µ8j, Noelle-Neumann +µµ¡, Salmon
and Rucinski +µ88, Taylor +µ8:). Generally, public expression has been op-
erationalized by some measure of the respondent’s willingness to express his or
her opinion in a hypothetical situation.
A valid measure, however, needs to take into account at least six factors.
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr +j
+ Cross-national differences: major problems have emerged in comparative
studies due to the inapplicability of empirical tests of willingness to speak out
in different cultural settings (Donsbach and Stevenson +µ8ô, Noelle-Neumann
+µµ¡, Tokinoya +µ8µ, +µµô). Therefore operationalizations have to be de-
veloped which can be applied across cultures.
: Public exposure: Taylor (+µ8:) uses a measure of willingness to donate money
for a group that supports one’s own position as an indicator of public
outspokenness. Financial contributions, however, lack the public element that
produces conformity (Tyson and Kaplowitz +µyy). In other words, public
opinion needs to be operationalized as ‘the tribunal at which judgment is
passed’ (Noelle-Neumann +µµj, p. ¡ô).
¡ Anonymous public: even if a public element is included in the measures of
public opinion expression, tests might neglect the anonymous character of
the public. Glynn and McLeod (+µ8¡), for example, operationalized public
as ‘a social gathering of people you know’ (p. y¡¡). Salmon and Rucinski
(+µ88, p. µ) used a hypothetical conversation ‘with a group of friends’ to
operationalize public. Essentially private situations, however, i.e. out-
spokenness in the family or with friends, are not suited for empirical tests
of the concept ‘for it is the willingness to speak out in public (coram publico)
which is important to the process of public opinion’ (Noelle-Neumann +µµ¡,
p. +o).
¡ Size of the public: when operationalizing outspokenness, experimental designs
or hypothetical questions in surveys should be chosen in which the size of the
public is kept constant and as small as possible. According to Noelle-Neumann
(+µµ¡), ‘the larger the public, the more personal characteristics—security,
self-confidence, practice in speaking, education, role—will influence re-
sponses, independently of the climate of opinion’ (p. +++).
j Survey data: criticisms of the weak external validity of most conformity
experiments make survey data the appropriate method for data collection for
evaluating the willingness to speak out.
ô Moral loading: the issue under study has to be a controversial one with a
clearly identifiable moral aspect attached to it.
A hypothetical question used in a telephone survey by Donsbach and Stevenson (+µ8ô) with an extremely
large public will therefore not be appropriate for measuring outspokenness. They asked: ‘Suppose a TV
reporter with a camera and a microphone were asking people on the street whether there should be a
constitutional amendment to ban abortion. Would you be willing to say what you think or would you prefer
not to?’ (p. :µ).
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn +ô
Wr +nr ×-Ctr+tar Dr rrrar×crs r × Orr ×r o× Pracrr+r o×s:
Rrrrar×cr Gaotrs vs. A×o×v·ots Pturr cs
Embedded in mass society theory, in which mass media are seen as powerful
and affecting a universe of atomized individuals, the spiral of silence has been
criticized for theorizing that individuals fear isolation from society at large. The
emphasis, these critics argue, should be on reference groups (Kennamer +µµo,
Salmon and Kline +µ8j).
Research in social psychology has documented the effects of reference groups.
For instance, Newcomb’s (+µ¡8) study of Bennington College students revealed
an increasing conformity to group norms over time. Data from a similar, but
shorter, field experiment conducted by Siegel and Siegel (+µj¡) offer additional
support for how attitude change over time is related to the group identification
of the subject (in this case, female college students as well). So while the power
of reference groups has been acknowledged in other fields relatively long ago,
it has until recently been noticeably absent from spiral of silence research.
Katz (+µ8¡) wrote that ‘the spiral of silence theory itself cannot so easily
dismiss the direct influence of actual membership and reference groups’
(p. µy). While that statement may have stemmed from his work during the era
of ‘limited effects’ (Katz and Lazarsfeld +µjj, Klapper +µôo), others concerned
with linking levels of analysis within the field of public opinion also call for
greater attention to the role of reference groups (Price +µ88). Indeed, reference
groups (Shibutani +µô+/+µµ+) may play a more important role in the process
of the spiral of silence than Noelle-Neumann had originally assumed (Glynn
and McLeod +µ8¡, Price and Allen +µµo, Salmon and Kline +µ8j, Salmon and
Neuwirth +µµo, Salmon and Rucinski +µ88, Taylor +µ8:). Critics have charged
that ‘Noelle-Neumann has ignored the role of these small reference groups and,
instead, has maintained that individuals perceive and are subject to the dominant
‘‘public’’ opinion, although it is unclear what this ‘‘public’’ represents or is
composed of’ (Salmon and Kline +µ8j, p. +o). Rather, individual perception of
opinion climates is heavily influenced by the opinion distributions in important
reference groups. Edelstein (+µ88) calls for an empirical answer to the question
‘what kinds of reference groups and what kinds of anonymous publics contribute
most to the spheres of communication in which fear of isolation is played out’
(Edelstein +µ88, p. j+j).
Empirical research does not provide any conclusive evidence for the importance
of reference groups for individual perceptions of opinion climates. Salmon and
Rucinski (+µ88) examined the congruency of respondents’ opinions with the
perceived opinions of their friends, the population at the state level, and the
opinions portrayed in mass media as antecedents of their willingness to speak out.
Congruency between each of these climates of opinion and one’s own opinion was
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr +y
not related to willingness to speak out. For congruency with friends, the coefficient
was in the wrong direction and significant. These findings, however, are very
likely a result of the operationalization of the dependent variable. Rather than
constructing a hypothetical situation with an anonymous group of people, they
measured respondents’ willingness to express their opinions to friends.
Salmon and Neuwirth (+µµo) examined the impact of perceived community
congruity and perceived national congruity on respondents’ willingness to
engage in a conversation with a stranger about abortion. Results showed that
perceived community congruency did not have an impact on willingness to
speak out. Consistent with Noelle-Neumann’s reasoning, perceived national
congruency predicted positively to respondents’ willingness to engage in such
a conversation. Oshagan (+µµô) reported somewhat different results from a
similar study, finding that when reference and societal majority opinions are
made equally salient, the former becomes a more important influence.
Examining issues relevant to four separate communities, Glynn and Park
(+µµj, +µµy) found that self-specified reference groups did not influence
respondents’ willingness to express their opinions on a given issue, whereas the
perceptions of the opinion distribution in the community as a whole did have
an effect, ‘suggesting that the dominant opinion among the ‘‘generalized other’’
reference group makes a difference, while perceptions of opinion dominance
within a specific reference group does not’ (Glynn and Park +µµj, p. +8). These
findings need to be interpreted in light of two limitations. First, their measure
of willingness to speak out involved subjects circling the statement they
supported most from a list of statements. This measure does not meet the
criteria for operationalizing willingness to speak out as outlined earlier. Second,
as a result of the community-specific issue, the study could not make any
inferences to a national climate of opinion.
Br+wrr×-Ctr+tar Dr rrrar×crs r × Orr ×r o× Exrarssr o×: Is +nrar
+ Dr rrrar×cr r × Ot+srokr××rss?
A question that has remained virtually unanswered concerns whether the spiral
of silence is a universal, cross-cultural phenomenon. Salmon and Kline (+µ8j)
suggest that as a macro theory, the spiral of silence should be applicable across
national boundaries. Indeed, the theory has been tested in a number of countries
including Germany (see Noelle-Neumann +µµ¡), Japan (Ikeda +µ8µ, Tokinoya
+µ8µ, +µµô), Korea (Yang +µµy), and the USA (see Scheufele +µµµ).In their
meta-analysis of over :o published and unpublished studies from six countries,
Glynn et al. (+µµy) found small but significant effects of perceived congruency
between own and perceived public opinion on people’s willingness to express
their opinion on controversial issues. Their meta-analysis, however, is plagued
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn +8
by the same problems of commensurability confronting cross-national com-
parisons of spiral of silence research. Genuine comparability, of course, ‘can be
achieved only by the application of some carbon-copy like process’ (Halloran
+µ8+, p. ++). Comparisons require that one key variable be reasonably similar
with respect to how it is conceptualized and operationalized: the hypothetical
situation commonly used to measure respondents’ willingness to speak out.
As noted earlier, willingness to speak out has been measured in a number of
ways, posing threats even to the comparability of studies within a single country
or culture. Even more confusing is a finding reported by Donsbach and
Stevenson (+µ8¡) who found that the same indicators for people’s willingness
to speak out might not be equally appropriate across different cultures. They
conclude that ‘the problem continues to exist of adapting the ‘‘public situation’’
and the ‘‘readiness to speak up’’ to the cultural and social conditions in a society’
(Donsbach and Stevenson +µ8¡, p. +j).
Research suggests that beyond the more methodological problem of finding
appropriate indicators for concepts in a given culture, there is a substantial
difference in personality traits for people living in different cultures (Cushman
and Sanderson King +µ8j, Gudykunst et al. +µµ:, Hui and Triandis +µ8ô, Hsu
+µ8j, Ito +µµ¡, Ting-Toomey +µ88, Triandis et al. +µ88, Trubinsky et al. +µµ+).
The concept ‘culture’, and differences between cultures, are hard to grasp and
even more difficult to operationalize (e.g. Triandis et al. +µ88). Cushman
and Sanderson King (+µ8j) define culture in terms very similar to symbolic
interactionism, in which an important aspect of culture is a ‘symbolically
integrated framework that regulates social interaction’ (p. ++j).
Cultures differ ‘in the extent to which cooperation, competition, or in-
dividualism . . . are emphasized’ (Triandis et al. +µ88, p. ¡:¡). The concept of
individualism seems to be a key variable in differentiating social behavior,
particularly communicatory behavior, across cultures (e.g. Gudykunst et al.
+µµ:, Ito +µµ¡, Ting-Toomey +µ88). Ting-Toomey (+µ88) distinguishes between
‘individualistic, low-context cultures, and collectivistic, high-context cultures’
(p. :+¡). Countries like Australia, Germany, or the USA can be considered
individualistic cultures, while Asian countries exemplify collectivistic cultures.
In individualistic cultures, the consistency between private self-image and public
self-image is of utmost importance (Ting-Toomey +µ88). In other words, one
has to be true to himself or herself. As a result, ‘members of individualistic
cultures tend to stress the value of straight talk and tend to verbalize overtly
their individual wants and needs’ (Trubinsky et al. +µµ+, p. ô8). What other
people think of them is of only marginal importance to individuals in in-
dividualistic cultures (Ito +µµ¡). It seems, however, that the idea of individualism
is limited to certain cultures. Moscovici (+µ8j) writes: ‘To the best of my
knowledge, it was the West, and the West alone, that produced and refined
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr +µ
the concept of humanity as autonomous, rational, self-directed individuals’
(Moscovici +µ8j, p. ¡¡y).
In contrast, the ‘self’ collectivistic culture is situationally based and depends
heavily on the social environment at the time the social interaction takes place.
Wei-ming (+µ8j) suggests that this is partly a function of a widespread Confucian
school of thought in Asian countries that ‘has undermined the autonomy of
the individual self’ (Wei-ming +µ8j, p. :¡¡). Consequently, individuals in
collectivistic cultures display ‘discretion in voicing . . . opinions and feelings’
(Trubinsky et al. +µµ+, p. ô8). Hui and Triandis (+µ8ô) summarize what can
be called the ‘collectivist personality’:
Collectivists are more likely to pay more attention to the influencing agent than are
individualists. As a result, collectivists are more conforming than individualists . . .
It may be safe to say that the former are more willing to go along with the group,
to avoid being rejected (Hui and Triandis +µ8ô, p. :¡o).
This distinction between individualism and collectivism is highly relevant
for future spiral of silence research. If it is indeed possible to identify personality
characteristics common to citizens of a given culture, these characteristics might
prove to be important long-term predictors of people’s willingness to speak out
beyond more temporally-bound perceptions of opinion climates.
In order to answer this question, the distinction between individualism and
collectivism at the socio-cultural level must be translated into equivalents at
the psychological level: allocentrism and idiocentrism, respectively (Triandis et
al. +µ88). However, this dichotomy might need to be refined for cross-cultural
comparisons. In examining conflict styles in the USA, Yugoslavia, and Japan,
Cushman and Sanderson King (+µ8j) concluded that different cultures ‘develop
their own myths, rituals, and social dramas for conflict resolution’ (Cushman
and Sanderson King +µ8j, p. ++y). They identified distinctively different
patterns of conflict resolution for the three cultures: the Japanese manifested a
pattern of collaboration, based on their ‘strong respect for hierarchy [and]
homogenous values’ (Cushman and Sanderson King +µ8j, p. +::); the USA
showed a pattern of competition, based on a strong sense of individual freedom
and achievement; and Yugoslavia as a culture was characterized by a climate of
compromise, which may stem from the ethnic diversity created somewhat
artificially this century.
Cushman and Sanderson King’s (+µ8j) scheme, however, lacks the abstraction
that makes it useful as a general classification scheme for types of cultures based
on the collectivism–individualism continuum. Moreover, at the psychological
level, it does not allow us to determine the levels of allocentrism or idiocentrism
that might be important factors in predicting people’s willingness to speak out.
More directly applicable approaches may be adapted from Rahim’s (+µ8¡)
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn :o
work on interpersonal conflict styles and Putnam and Wilson’s (+µ8:) research on
communicative strategies in organizational conflicts. By classifying respondents
along their concern for themselves and for others, Rahim generated five different
conflict styles: integrating (high concern for self and others), dominating (high
concern for self and low concern for others), obliging (low concern for self and
high concern for others), avoiding (low concern for self and others), and
compromising as a middle category. Based on a similar reasoning, Putnam and
Wilson (+µ8:) suggested a trichotomous scale that orders respondents and their
conflict resolution strategies hierarchically. Their scale can be used as a measure
of allocentrism vs. idiocentrism and, at the same time, explain the phenomenon
of avant-gardes or hard cores, or those respondents whose willingness to speak
out seems mostly unaffected by their perceptions of the climate of opinion.
Putnam and Wilson (+µ8:) suggested the following three dimensions of conflict
+ Nonconfrontation: indirect strategies for handling a conflict; choices are to
avoid or withdraw from a disagreement; communication behaviors include
silence, glossing over differences, and concealing ill feelings.
: Solution-orientation: direct communication about the conflict; behaviors that
aim to find a solution, to integrate the needs of both parties, and to give in
or compromise on issues.
¡ Control: direct communication about the disagreement; arguing persistently
for one’s position, taking control of the interaction, and advocating one’s
These scales may be used to identify persons who presumably react more or less
strongly to social pressure, and can be employed in internationally comparative
research to identify societies that have a particularly pronounced desire for
consensus. Accordingly, the mechanisms of speaking out and falling silent
described within the framework of the spiral of silence ought to be clearly
evident in these societies.
Looking back over :j years of spiral of silence research, this paper has attempted
to summarize and clarify some of the conceptual issues that spiral of silence
research has suffered from in recent decades. More specifically, we developed
a set of general assumptions underlying the spiral of silence theory and theoretical
statements describing its most important processes. Based on this conceptual
work, we outlined a research agenda, identifying deficits or inconsistencies in
previous research and necessary areas of future research. In a third step, rather
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr :+
than focusing on existing theorizing, we argued for the inclusion of more
macroscopic variables in spiral of silence research.
A stronger focus of macroscopic variables and a stronger emphasis on
cross-cultural research has the potential to promote theory-building in this area.
Putnam and Wilson’s (+µ8:) scale of conflict dimensions, for example, was not
designed to tap cross-cultural differences. If applied to spiral of silence research
in a cross-cultural setting, however, it might be useful in three areas.
First, measures of conflict styles can serve as measures of a phenomenon that
Noelle-Neumann calls hardcore or avant-gardes, tapping the degree to which
individuals are susceptible to perceptions of opinion climates. In other words,
are some people more concerned about themselves than about others, or vice
versa, and are there personality characteristics that make some more likely to
express their opinions, independent of their perceptions of their environment?
Previous research has treated the concepts of ‘hardcores’ or ‘avant-gardes’ as
assumptions rather than variables. We strongly argue for including measures of
these concepts in future empirical research on the spiral of silence.
Second, and closely related, this measure might serve as a measure of fear
of isolation. As we mentioned earlier, Glynn and McLeod (+µ8j) suggested
that fear of isolation should be treated as a variable rather than an assumption.
So far, research has widely ignored this suggestion (exceptions are Neuwirth’s
(+µµj) study in Mexico, Yang’s (+µµy) research in Korea, and Moy, Domke,
and Stamm’s (+µµµ) work in the USA).
Third, and finally, a measurement of conflict styles could serve as a control
for cross-cultural comparisons. If cultures truly differ with respect to how
individuals handle conflict and deal with public pressure, this variable is crucial
in revealing these cultural differences and providing further insights into the
process of the spiral of silence.
In sum, we argue for a return to a more macroscopic focus in spiral of silence
research. A number of researchers (e.g. Lasorsa +µµ+, Mutz +µ8µ, Price and
Allen +µµo) have approached the theory from a more micro perspective. While
very fruitful with respect to testing single theoretical statements within the
theory, these approaches have done little to account for overall differences in
effects or to examine the more societal-level processes that the spiral of silence
theory predicts. A more macroscopic focus is what makes the spiral of silence
theory most appealing for the field of public opinion research, which traditionally
has limited its explanations to micro- and meso-theoretical approaches. Not
only can macro-theoretical approaches, like the spiral of silence, explain micro-
inconsistencies on an empirical level but more importantly, a macroscopic focus
is very likely the key to reconceptualizing and defining concepts like hardcores
and avant-gardes that have been somewhat neglected in previous research on
the spiral of silence.
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn ::
Allport, F. H. (+µ¡y): ‘Toward a science of public opinion’, Public Opinion Quarterly,
1, y–:¡.
Ansolabehere, S. and Iyengar, S. (+µµ¡): ‘Of horseshoes and horse races: Experimental
studies of the impact of poll results on electoral behavior’, Political Communication,
11, ¡+¡–¡o.
Argyle, M. (+µjy): ‘Social pressure in public and private situations’, Journal of Abnormal
and Social Psychology, ¡¡, +y:–j.
Asch, S. E. (+µjj): ‘Opinions and social pressure’, Scientific American, 1q¡, ¡+–j.
Asch, S. E. (+µôj): ‘Effects of group pressure upon the modification and distortion of
judgements’. In Campbell, J. H. and Hepler, H. W. (eds.) Dimensions in Communication:
Readings, Belmont, CA, Wadsworth, pp. +:j–¡y.
Atkin, C. K. (+µôµ): ‘The impact of political poll reports on candidate and issue
preference’, Journalism Quarterly, ¡6, j+j–:+.
Bartels, L. (+µ88): Presidential Primaries and the Dynamics of Public Choice, Princeton,
NJ, Princeton University Press.
Bentham, J. (+8¡8–+8¡¡/+µô:): ‘The constitutional code’. In Bowring, J. (ed.) The
Works of Jeremy Bentham. Vol. , Book  New York, Russel & Russel, pp. ¡+–ô.
Childs, H. (+µôj): Public Opinion: Nature, Formation, and Role, Princeton, NJ, D. v.
Cook, S. W. and Welch, A. C. (+µ¡o): ‘Methods of measuring the practical effect of
polls on public opinion’, Journal of Applied Psychology, z¡, ¡¡+–j¡.
Cooley, C. (+µjô): Human Nature and the Social Order, New York, Scribner.
Crutchfield, R. S. (+µjj): ‘Conformity and character’, The American Psychologist, 1o,
Csikszentmihalyi, M. (+µµ+): ‘Reflections on the ‘‘spiral of silence’’ ’. In Anderson, J.
A. (ed.) Communication Yearbook, , Newbury Park, CA, Sage, pp. :88–µy.
Cushman, D. P. and Sanderson King, S. (+µ8j): ‘National and organizational cultures
in conflict resolution: Japan, the United States, and Yugoslavia’. In Gudykunst,W. B.,
Stewart, L. P., and Ting-Toomey, S. (eds.) Communication, Culture, and Organizational
Processes, Newbury Park, CA, Sage, pp. ++¡–¡¡.
Davison, W. P. (+µô8): ‘Public opinion: Introduction’. In Sills, D. L. (ed.) International
Encyclopedia of the Social Sciences,New York, Macmillan, pp. +88–µy.
de Bock, H. (+µyô): ‘Influence of in-state election poll reports on candidate preference
in +µy:’, Journalism Quarterly, ¡¡, ¡jy–ô:.
Delli Carpini, M. X. (+µ8¡): ‘Scooping the voters? The consequences of the networks’
early call of the +µ8o presidential race’, Journal of Politics, ¡6, 8ôô–8j.
Deutsch, M. and Gerard, H. B. (+µjj): ‘A study of the normative and informational
social influences upon individual judgment’, Journal of Abnormal and Social Psychology,
¡1, ô:µ–¡ô.
Donsbach, W. (+µ8y): ‘Die Theorie der Schweigespirale’ [The theory of the spiral of
silence]. In Schenk, M. (ed.) Medienwirkungsforschung, Tu¨bingen, Mohr, pp. ¡:¡–
Donsbach, W. and Stevenson, R. L. (+µ8¡): Challenges, Problems, and Empirical
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr :¡
Evidence of the Theory of the Spiral Of Silence. Paper presented to the annual
convention of the International Communication Association, San Francisco, May.
Donsbach, W. and Stevenson, R. L. (+µ8ô): ‘Herausforderungen, Probleme und em-
pirische Evidenzen der Theorie der Schweigespirale’ [Challenges, problems, and
empirical evidence of the theory of the spiral of silence], Publizistik, ¡1, y–¡¡.
Edelstein, A. S. (+µ88): ‘Communication perspectives in public opinion: Traditions and
innovations’. In Anderson, J. A. (ed.), Communication Yearbook , Beverly Hills,
CA, Sage, pp. jo:–¡¡.
Gartner, M. (+µyj): ‘Endogenous bandwagon and underdog effects in a rational choice
model’, Public Choice, z¡, 8¡–µ.
Glynn, C. J. and McLeod, J. M. (+µ8¡): ‘Public opinion du jour: An examination of
the spiral of silence’, Public Opinion Quarterly, ¡8, y¡+–¡o.
Glynn, C. J. and McLeod, J. M. (+µ8j): ‘Implications of the spiral of silence theory
for communication and public opinion research’. In Sanders, K. R., Kaid, L. L.,
and Nimmo, D. (eds.) Political Communication Yearbook , Carbondale and
Edwardsville, Southern Illinois University Press, pp. ¡¡–ôj.
Glynn, C. J. and Park, E. (+µµj): Reference Groups, Opinion Thresholds, and Public
Opinion Expression: A Modification of the Spiral of Silence. Paper presented at the
annual meeting of the American Association for Public Opinion Research, Fort
Lauderdale, Florida, May.
Glynn, C. J. and Park, E. (+µµy): ‘Reference groups, opinion intensity, and public
opinion expression’, International Journal of Public Opinion Research, q, :+¡–¡:.
Glynn, C. J., Ostman, R. E., and McDonald, D. G. (+µµj): ‘Opinions, perception, and
social reality’. In Glasser, T. L. and Salmon, C. T. (eds.) Public Opinion and the
Communication of Consent, New York, Guilford Press, pp. :¡µ–yy.
Glynn, C. J., Hayes, A. F., and Shanahan, J. (+µµô): Willingness to Speak Out and the Spiral
of Silence: A Meta-Analysis. Paper presented at the annual meeting of the Association
for Education in Journalism and Mass Communication, Anaheim, CA, August.
Glynn, C. J., Hayes, A. F., and Shanahan, J. (+µµy): ‘Spiral of silence: A meta-analysis’,
Public Opinion Quarterly, 61, ¡j:–ô¡.
Gudykunst, W. B., Gao, G., Schmidt, K. L., Nishida, T., Bond, M., Leung, K., Wang,
G., and Barraclough, R. A. (+µµ:): ‘The influence of individualism-collectivism,
self-monitoring, and predicted-outcome value on communication in ingroup and
outgroup relationships’, Journal of Cross-cultural Psychology, z¡, +µô–:+¡.
Habermas, J. (+µô:): Strukturwandel der O
ffentlichkeit: Untersuchungen zu einer Kategorie
der bu¨rgerlichen Gesellschaft [The structural transformation of the public sphere: An
inquiry into a category of Bourgeois society], Neuwied, Luchterhand.
Hallemann, M. (+µ8µ): Peinlichkeit: Ein Ansatz zur Operationalisierung von Isolationsfurcht
im sozialpsychologischen Konzept o¨ffentlicher Meinung [Embarrassment: An approach
to measuring fear of isolation in the social-psychological concept of public opinion],
Unpublished Doctoral Dissertation, Johannes Gutenberg-Universita¨t Mainz.
Halloran, J. (+µ8+): ‘An exercise in international research co-operation’. In United
Nations Educational, Scientific and Cultural Organization (ed.) Communication in the
Community, Paris, UNESCO, pp. j–+¡.
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn :¡
Hardy, K. R. (+µjy): ‘Determinants of conformity and attitude change’, The Journal of
Abnormal and Social Psychology, ¡¡, :8µ–µ¡.
Heider, F. (+µ¡o): ‘Die Leistung des Wahnehmungssystems’ [The functions of the
perceptual system], Zeitschrift fu¨r Psychologie, 11¡, ¡y+–µ¡.
Heider, F. (+µjµ): The Psychology of Interpersonal Relations, :nd edn, New York, Wiley.
Heider, F. (+µy8): ‘U
ber Balance und Attribution’ [About balance and attribution]. In:
Go¨rlitz, D., Meyer, W.-U., and Weiner, B. (eds.) Bielefelder Symposiumu¨ber Attribution,
Stuttgart, Klett, pp. +µ–:8.
Heider, F. and Simmel, M. (+µ¡¡): ‘An experimental study of apparent behavior’,
American Journal of Psychology, ¡y, :¡¡–jµ.
Hovland, C. L. (+µjµ): ‘Reconciling conflicting results derived from experimental and
survey studies of attitude change’, The American Psychologist, 1¡, 8–+y.
Hsu, F. L. K. (+µ8j): ‘The self in cross-cultural perspective’. In Marsella, A. J., DeVos,
G., and Hsu, F. L. K. (eds.) Culture and Self: Asian and Western Perspectives, New
York, Tavistock, pp. :¡–jj.
Hui, C. H. and Triandis, H. C. (+µ8ô): ‘Individualism and collectivism: A study of
cross-cultural researchers’, Journal of Cross-cultural Psychology, 1y, ::j–¡8.
Ikeda, K. (+µ8µ): ‘ ‘‘Spiral of silence’’ hypothesis and voting intention: A test in the
+µ8ô Japanese national election’, KEIO Communication Review, 1o, j+–ô:.
Ito, Y. (+µµ¡): ‘New directions in communication research from a Japanese perspective’.
In Gaunt, P. (ed.), Beyond Agendas: New Directions in Communication Research,
Westport, CT, Greenwood Press, pp. ++µ–¡j.
Katz, C. and Baldassare, M. (+µµ:): ‘Using the ‘‘L-Word’’ in public: A test of the spiral
of silence in conservative Orange County, California’, Public Opinion Quarterly, ¡6,
Katz, E. (+µ8¡): ‘Publicity and pluralistic ignorance: Notes on ‘‘the spiral of silence’’ ’.
In Wartella, E., Whitney, D. C., and Windahl, S. (eds.) Mass Communication Review
Yearbook , Beverly Hills, CA, Sage, pp. 8µ–µµ.
Katz, E. and Lazarsfeld, P. F. (+µjj): Personal Influence: The Part Played by People in
the Flow of Mass Communications, Glencoe, IL, Free Press.
Kennamer, J. D. (+µµo): ‘Self-serving biases in perceiving the opinions of others:
Implications for the spiral of silence’, Communication Research, 1y, ¡µ¡–¡o¡.
King, C. L. (+µ:8): ‘Public opinion in government’. Introduction to Graves, W. B.
(ed.) Readings in Public Opinion: Its Formation and Control, New York, Appleton, pp.
Klapper, J. T. (+µôo): The Effects of Mass Communication, New York, Free Press.
Koffka, K. (+µ::): ‘Perception’, Psychological Bulletin, 1q, jôô–yo.
Lasorsa, D. L. (+µµ+): ‘Political outspokenness: Factors working against the spiral of
silence’, Journalism Quarterly, 68, +¡+–¡o.
Latane´, B. and Darley, J. M. (+µô8): ‘Group inhibition of bystander intervention’,
Journal of Personality and Social Psychology, 1o, :+j–:+.
Lavrakas, P. J., Holley, J. K., and Miller, P. V. (+µµ+): ‘Public reactions to polling news
during the +µ88 presidential election campaign’. In Lavrakas, P. J. and Holley, J. K.
(eds.) Polling and Presidential Election Coverage, Newbury Park, CA, Sage, pp. +j+–8¡.
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr :j
McAllister, I. and Studlar, D. T. (+µµ+): ‘Bandwagon, underdog, or projection? Opinion
polls and electoral choice in Britain, +µyµ–+µ8y’, Journal of Politics, ¡¡, y:o–¡+.
McLeod, J. M. and Blumler, J. G. (+µ8y): ‘The macrosocial level of communication
science’. In Chaffee, S. and Berger, C. (eds.) Handbook of Communication Science,
Beverly Hills, CA, Sage, pp. :y+–¡::.
Merton, R. K. (+µô8): Social Theory and Social Structure, enlarged edn, New York,
The Free Press.
Meeus, W. H. and Raaijmakers, Q. A. W. (+µ8µ): ‘Theorie und Methode: Au-
torita¨tsgehorsam in den Experimenten des Milgram-Typs: Eine Forschungsu¨bersicht’
[An overview of theory and method: Obedience in the Milgram experiments],
Zeitschrift fu¨r Sozialpsychologie, zo, yo–8j.
Milgram, S. (+µô+): ‘Nationality and conformity’, Scientific American, zo¡, ¡j–j+.
Miller, D. T. and Prentice, D. A. (+µµ¡): ‘Collective errors and errors about the
collective’, Personality and Social Psychology Bulletin, zo, j¡+–jo.
Moscovici, S. (+µ8j): ‘Social influence and conformity’. In Lindzey, G. and Aronson,
E. (eds.) Handbook of Social Psychology. Volume II. Special Fields and Applications,
¡rd edn, New York, Random House, pp. ¡¡y–¡+:.
Moy, P., Domke, D. S., and Stamm, K. (+µµµ): The Spiral of Silence and Affirmative
Action: The Case of Washington’s Initiative :oo. Paper presented at the annual
meeting of the American Association for Public Opinion Research, St. Petersburg,
FL, May.
Mutz, D. (+µ8µ): ‘The influence of perceptions of media influence: Third person effects
and the public expression of opinions’, International Journal of Public Opinion Research,
1, ¡–:¡.
Mutz, D. C. (+µµ¡): ‘The political effects of perceptions of mass opinion’, Research in
Micropolitics, ¡, +¡¡–ôy.
Navazio, R. (+µyy): ‘An experimental approach to bandwagon research’, Public Opinion
Quarterly, ¡1, :+y–:j.
Neuwirth, K. J. (+µµj): ‘Testing the ‘‘spiral of silence’’ model: The case of Mexico’.
Doctoral Dissertation, University of Wisconsin-Madison. UMI Microform µj¡+ô8:.
Neuwirth, K. and Ilundain, C. S. (+µ8¡): Family Communication Patterns and the
Spiral of Silence. Paper presented at the annual meeting of the Midwest Association
for Public Opinion Research, Chicago, IL, November.
Newcomb, T. M. (+µ¡8): ‘Attitude development as a function of reference groups: The
Bennington study’. In Sherif, M. (ed.) An Outline of Social Psychology, New York,
Harper & Brothers, pp. +¡µ–jj.
Nisbett, R. E. and Kunda, Z. (+µ8j): ‘Perceptions of social distributions’, Journal of
Personality and Social Psychology, ¡8, :µy–¡++.
Noelle-Neumann, E. (+µy¡): ‘The spiral of silence: A theory of public opinion’, Journal
of Communication, z¡, ¡¡–j+.
Noelle-Neumann, E. (+µyô): ‘Menschen unter Konformita¨tsdruck. Eine Theorie der
o¨ffentlichen Meinung’ [Human beings and the pressure to conform: A theory of
public opinion], Frankfurter Allgemeine Zeitung, ¡1, February +8, +o–++.
Noelle-Neumann, E. (+µyµ): ‘Die Schweigespirale, u¨ber die Entstehung der o¨ffentlichen
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn :ô
Meinung’ [The spiral of silence: The evolution of public opinion]. In Wilke, J. (ed.)
ffentlichkeit als Bedrohung, Beitra¨ge zur empirischen Kommunikationsforschung, :nd,
rev edn, Freiburg i.Br., Alber Broschur, pp. +ôµ–:o¡.
Noelle-Neumann. E. (+µ8¡): ‘Neue Forschungen im Zusammenhang mit der Schwei-
gespiralen-Theorie’ [More recent research on the spiral of silence theory]. In Saxer,
U. (ed.) Politik und Kommunikation, Neue Forschungsansa¨tze, Mu¨nchen, O
pp. +¡¡–+¡¡.
Noelle-Neumann, E. (+µ8j): ‘The spiral of silence: A response’. In Sanders, K. R.
Kaid, L. L., and Nimmo, D. (eds.) Political Communication Yearbook , Carbondale
and Edwardsville, Southern Illinois University Press, pp. ôô–µ¡.
Noelle-Neumann, E. (+µµ+): ‘The theory of public opinion: The concept of the spiral
of silence’. In Anderson, J. A. (ed.) Communication Yearbook, , Newbury Park,
CA, Sage, pp. :jô–8y.
Noelle-Neumann, E. (+µµ:): ‘Manifeste und latente Funktion o¨ffentlicher Meinung’
[Manifest and latent functions of public opinion], Publizistik, ¡y, :8¡–µy.
Noelle-Neumann, E. (+µµ¡): The Spiral of Silence. Public Opinion—Our Social Skin,
:nd edn, Chicago, IL, University of Chicago Press.
Noelle-Neumann, E. (+µµ¡): ‘Are we asking the right questions? Developing measure-
ment from theory: The influence of the spiral of silence on media effects research’.
In Hamelink, C. J. and Linne´, O. (eds.) Mass Communication Research: On Problems
and Policies. The Art of Asking the Right Questions. In Honor of James D. Halloran,
Norwood, NJ, Ablex, pp. µy–+:o.
Noelle-Neumann, E. (+µµj): ‘Public opinion and rationality’. In Glasser, T. L. and
Salmon, C. T. (eds.) Public Opinion and the Communication of Consent, New York,
Guilford, pp. ¡¡–j¡.
O’Gorman, H. J. and Garry S. L. (+µyô): ‘Pluralistic ignorance: A replication and
extension’, Public Opinion Quarterly, ¡o, ¡¡µ–j8.
Oshagan, H. (+µµô): ‘Reference group influence on opinion expression’, International
Journal of Public Opinion Research, 8, ¡¡j–j¡.
Palmer, P. A. (+µ¡ô): ‘The concept of public opinion in political theory’. In Essays in
History and Political Theory: In Honor of Charles Howard McIlwain, Cambridge, MA,
Harvard University Press, pp. :¡o–jy.
Pollis, N. P. and Cammalleri, J. A. (+µô8): ‘Social conditions and differential resistance
to majority pressure’, Journal of Psychology, yo, ôµ–yô.
Price, V. (+µ88): ‘On the public aspects of opinion: Linking levels of analysis in public
opinion research’, Communication Research, 1¡, ôjµ–yµ.
Price, C. and Allen, S. (+µµo): ‘Opinion spirals, silent and otherwise: Applying
small-group research to public opinion phenomena’, Communication Research, 1y,
Putnam, L. L. and Wilson, C. E. (+µ8:): ‘Communicative strategies in organizational
conflicts: Reliability and validity of a measurement scale’. In M. Burgoon (ed.)
Communication Yearbook , Beverly Hills, CA, Sage, pp. ô:µ–j:.
Rahim, M. A. (+µ8¡): ‘A measure of styles of handling interpersonal conflict’, Academy
of Management Journal, z6, ¡ô8–yô.
+wr ×+v- rr vr vr+as or +nr s rr a+r or s r rr×cr :y
Ross, L., Bierbrauer, G., and Hoffman, S. (+µyô): ‘The role of attribution processes in
conformity and dissent’, American Psychologist, ¡1, +¡8–jy.
Rucinski, D. and Salmon, C. T. (+µµo): ‘The ‘‘other’’ as the vulnerable voter: A study
of the third-person effect in the +µ88 U.S. presidential campaign’, International
Journal of Public Opinion Research, z, ¡¡j–ô8.
Salmon, C. T. and Kline, F. G. (+µ8j): ‘The spiral of silence ten years later: An
examination and evaluation’. In Sanders, K. R., Kaid, L. L., and Nimmo, D. (eds.)
Political Communication Yearbook , Carbondale and Edwardsville, Southern
Illinois University Press, pp. ¡–¡o.
Salmon, C. T. and Neuwirth, K. (+µµo): ‘Perceptions of opinion ‘‘climates’’ and
willingness to discuss the issue of abortion’, Journalism Quarterly, 6y, jôy–yy.
Salmon, C. T. and Rucinski, D. M. (+µ88): Fear of Isolation from Whom? Environmental
Cues and Willingness to Express Opinions on Controversial Issues. Paper presented
at the annual meeting of the International Communication Association, New Orleans,
LA, May.
Salwen, M. B., Lin, C., and Matera, F. (+µµ¡): ‘Willingness to discuss ‘‘official English’’:
A test of three communities’, Journalism Quarterly, y1, :8:–µo.
Sanders, K. R., Kaid, L.L., and Nimmo, D. (+µ8j): ‘Introduction’. In Sanders, K. R.,
Kaid, L.L., and Nimmo, D. (eds.) Political Communication Yearbook , Carbondale
and Edwardsville, Southern Illinois University Press, pp. xiii–xxi.
Scheufele, D. A. (+µµµ): ‘Discussion or dispute? An exploratory study examining
dimensions of public opinion expression’, International Journal of Public Opinion
Research, 11(+), :¡–j8.
Sherif, M. (+µôy): Social Interaction: Process and Products, Chicago, IL, Aldine.
Siegel, A. E. and Siegel, S. (+µj¡): ‘Reference groups, membership groups, and attitude
change’. In Cartwright, D. and Zander, A. (eds.) Group Dynamics: Research and
Theory, :nd edn, New York, Harper & Row, pp. :¡:–¡o.
Skalaban, A. (+µ88): ‘Do the polls affect elections? Some +µ8o evidence’, Political
Behavior, 1o, +¡ô–jo.
Straffin, P. D., Jr. (+µyy): ‘The bandwagon curve’, American Journal of Political Science,
z1, ôµj–yoµ.
Sudman, S. (+µ8ô): ‘Do exit polls influence voting behavior?’ Public Opinion Quarterly,
¡o, ¡¡+–µ.
Taylor, D. G. (+µ8:): ‘Pluralistic ignorance and the spiral of silence: A formal analysis’,
Public Opinion Quarterly, ¡6, ¡++–¡j.
Ting-Toomey, S. (+µ88): ‘Intercultural conflict styles: A face-negotiation theory’. In
Kim, Y. Y. and Gudykunst, W. B. (eds.) Theories in Intercultural Communication,
Newbury Park, CA, Sage, pp. :+¡–¡j.
Tokinoya, H. (+µ8µ): ‘Testing the spiral of silence in East Asia’, KEIO Communication
Review, 1o, ¡j–¡µ.
Tokinoya, H. (+µµô): ‘A study on the spiral of silence theory in Japan’, KEIO
Communication Review, 18, ¡¡–¡j.
To¨nnies, F. (+µ::): Kritik der o¨ffentlichen Meinung [A critique of public opinion], Berlin,
r ×+r a×++r o×+r J ota×+r or rturr c orr ×r o× ar s r +acn :8
Triandis, H. C., Bontempo, R., Villareal, M. J., Asai, M., and Lucca, N. (+µ88):
‘Individualism and collectivism: Cross-cultural perspectives on self-ingroup re-
lationships’, Journal of Personality and Social Psychology, ¡¡, ¡:¡–¡8.
Trubinsky, P., Ting-Toomey, S., and Lin, S. (+µµ+): ‘The influence of individualism-
collectivismand self-monitoring on conflict styles’, International Journal of Intercultural
Relations, 1¡, ôj–8¡.
Tyson, H. L. and Kaplowitz, S. A. (+µyy): ‘Attitudinal conformity and anonymity’,
Public Opinion Quarterly, ¡1, ::ô–¡¡.
Weimann, G. (+µµ+): ‘The influentials: Back to the concept of opinion leaders?’ Public
Opinion Quarterly, ¡¡, :ôy–yµ.
Weimann, G. (+µµ¡): The Influentials: People who Influence People, Albany, NY, State
University of New York Press.
Wei-ming, T. (+µ8j): ‘Selfhood and otherness in Confucian thought’. In Marsella, A.
J. DeVos, G., and Hsu, F. L. K. (eds.) Culture and Self: Asian and Western Perspectives,
New York, Tavistock, pp.:¡+–j+.
Wertheimer, M. (+µ:j): Drei Abhandlungen zur Gestalttheorie [Three essays about Gestalt
theory], Berlin, Philosophische Akademie.
West, D. M. (+µµ+): ‘Polling effects in election campaigns’, Political Behavior, 1¡,
Wilson, F. G. (+µ¡¡): ‘Concepts of public opinion’, The American Political Science
Review, zy, ¡y+–µ+.
Zech, C. E. (+µyj): ‘Leibenstein’s bandwagon effect as applied to voting’, Public Choice,
z¡, ++y–:¡.
Yang, S. (+µµy): ‘Factors influencing willingness to speak out in the public sphere: A
test of the spiral of silence in Korea’, Unpublished Doctoral Dissertation, University
of Wisconsin-Madison.
Dietram A. Scheufele is Assistant Professor in the Department of Com-
munication at Cornell University. His current research focuses on the role of
public opinion and its expression in democratic societies and the political effects
of mass communication.
Patricia Moy is Assistant Professor in the School of Communications at the
University of Washington. Her research interests lie in the areas of public
opinion, political commmunication, and research methodology.
Correspondence regarding this paper should be addressed to Dietram A.
Scheufele, Department of Communication, Cornell University, ¡+j Kennedy
Hall, Ithaca, NY +¡8j¡, Phone: ôoy.:jj.:ôo¡, Fax: ôoy.:j¡.+¡::, Email: