1
Interpersonal influence and pro- societal consumption:
A review of diffusion, conformity, dissemination, translation and reflexivity
Jonn Axsen a* and Kenneth S. Kurani a
Affiliations:
a Institute of Transportation Studies, University of California at Davis, 2028 Academic
Surge, One Shields Avenue, Davis, CA, 95616, U. S. A.
* Corresponding author. Tel.: 1 530 574 2150, Fax: 1 530 752 6572, E- mail address:
jaxsen@ ucdavis. edu
i
Abstract
Understanding consumption behavior will facilitate the successful deployment of new
pro- societal products and behaviors. Interpersonal influence is thought to play an
important role in the successive adoption of pro- societal products and behaviors within
groups across time, but processes of influence are poorly understood. This paper begins
by presenting a typology for new products according functional, symbolic and pro-societal
attributes. We then identify and summarize five perspectives on interpersonal
influence and pro- societal consumption behavior. Contagion focuses on the effect of
functional information flowing among people or groups. Conformity focuses on
individual thresholds and motivations to mimic others. Dissemination is the intentional
diffusion of information by a core group of motivated individuals. Translation is the
tendency for various social groups to negotiate interpretations and valuations of the
technology. Finally, reflexivity is a theoretical framework that explains the motivations
of individuals seeking to establish lifestyle practices consistent with their self concept in
an uncertain modern world. We describe strengths and weaknesses of each perspective
and suggest complementarities.
1
1. Introduction
Understanding consumer purchase behavior will facilitate the successful
deployment of new pro- societal products and consumption behaviors— practices that are
at least in part valued according to the collective benefits they offer. Social influence is
known to play an important role in consumer perceptions, values and behavior, and
ultimately in the impacts of environmental policy ( Norton, et al., 1998). However, such
processes are poorly understood, at least in part due to complexities in separating
confounding factors in behavioral outcome data, as well as widespread confusion in
terminology, concepts and theory across relevant literatures ( Manski, 2000). While the
concept of social influence can include interactions and relationships among and between
consumers and institutions, here we focus on interpersonal influence as a sub- category
that occurs among consumers. This review paper attempts to illuminate the confusion and
point towards a framework that integrates ideas from several of these perspectives on
interpersonal influence.
Processes of interpersonal influence are typically absent in behavioral models and
research ( Jackson, 2005). When they are addressed, these processes are most often
conceptualized according to diffusion of innovations ( DOI) which posits influence as the
effect of the flow of functional information among consumers statically categorized
according to their “ innovativeness” ( Rogers, 2003). The DOI perspective is often utilized
to add social factors to economic choice models; several recent studies have empirically
estimated social parameters representing aggregated preference changes resulting from
increasing technology adoption ( Axsen, et al., 2009; Mau, et al., 2008), word- of mouth
effects ( Struben and Sterman, 2008), and information search channels ( van Rijnsoever, et
2
al., 2009). However, we argue that the DOI perspective and its applications can
oversimplify and mischaracterize processes of interpersonal influence, particularly in
cases of products offering pro- societal benefits.
This paper offers two contributions to research on interpersonal influence and
pro- societal consumption behavior. First is a two- dimensional typology to conceptualize
how a new ideas, behaviors and products may be perceived as new by consumers. After
reviewing various literatures, we argue that pro- societal products can be complex. They
are not only innovative technologically and functionally, but may also present a radical
shift in symbolic benefits and pro- societal values— such shifts will be shown to be
integral to interpersonal influence processes. Second, we review literatures on five
perspectives on the role of interpersonal influence in consumption behavior, focusing on
automotive purchases. Beginning with DOI as a specific and widely used form of
contagion model, we review contagion, conformity, dissemination, translation and
reflexivity as alternative, and as we will argue, in some ways complimentary perspectives
on interpersonal influence. We review each perspective, in turn drawing out concepts and
language that can supplement or replace those provided by DOI. A companion paper
( Axsen and Kurani, 2010) combines insights gleaned from this literature review with
empirical observation to propose a new, integrative framework.
2. Conceptualizing pro- societal product attributes
Here, we focus on a particularly complex type of pro- societal product: electric-drive
vehicles. These include: pure electric vehicles ( EVs) powered only by electricity
from the grid; hybrid- electric vehicles ( HEVs) fueled only by gasoline, but using a small
3
electric motor and battery to improve fuel economy; and plug- in hybrid electric vehicles
( PHEVs) that combines aspects of the EV and HEV, potentially operating like an EV for
a limited distance, with the addition of a gasoline engine to extend driving range and
increase power. Note that the previous sentence briefly defines electric- drive vehicles
according to their functional, technological components: what they do and how they
differ from the incumbent technology, that is, conventional internal combustion vehicles.
Analysts often attempt to assess the likelihood of market success based purely on these
technology ( and cost) components, such as how continuous or discontinuous a transition
is required for the innovation to be adopted ( e. g. Ehrnberg, 1995; Robertson, 1971).
However, studying consumer perceptions, we caution against assuming
consumers have only a purely technical, functional conceptualization of new products.
Consumer valuation of a new product can also relate to a “ paradigm shift in beliefs,
attitudes, and use” ( Adamson, 2003). From a behavioral perspective, this consideration of
beliefs and attitudes is an essential addition: what matters is how any technological or
functional change is perceived by consumers. To better understand which attributes are
important to vehicle buyers, we conceptualize the attributes of a new product according
to two dimensions: functional/ symbolic and private/ societal ( Table 1).
Table 1: Conceptualization of PHEV attributes ( hypothetical example)
Functional Symbolic
Private Save money
Reliable
Fun to drive ( experiential)
Expression of self- identity
Convey personal status to others
Attain group membership
Societal Reduce air pollution
Reduce global warming
Reduce oil use
Inspire other consumers
Send message to automakers,
government, oil companies
4
2.1 The functional/ symbolic dimension: What does it do and represent?
The functional/ symbolic dimension distinguishes between what the product can
do and what it represents. Hirschman ( 1981) explains that functional attributes are
tangible, while symbolic attributes are intangible, where the innovation communicates a
new social meaning such as sexiness or status, and may be valued according to
fundamentally different processes than functional attributes. Hirschman highlights
functional and symbolic attributes as both being important for goods such as automobiles,
though perhaps in different ways.
Empirical research supports the notion that motor vehicles are associated with
intangible, symbolic motives— often more so than functional motives ( e. g. Steg, 2005;
Steg, et al., 2001). Gartman ( 2004) explains that this tendency has existed since
automobiles were introduced in the late 19 th century, where vehicles were “ used not for
practical purpose but for leisure activities and public ostentation… the automobile quickly
became defined in American culture as an instrument of freedom and leisure, and a
symbol of wealth.” Symbolic value has also continued with the introduction of modern
day electric- drive vehicles; Heffner et al. ( 2007) interviewed dozens of HEV owners,
discovering that in contrast to the predominately functional and technical
characterizations of HEVs in several market, energy, and environmental assessments,
symbolism played an important role in buyers’ purchase decisions. Participants described
symbols representing widely shared meanings such as “ preserving the environment” and
“ embracing new technology,” as well as more personal meanings such as ethics and
individuality. In Table 1, function denotes attributes of functional or instrumental
importance, including the basic services of accessibility and mobility provided by an
5
automobile, or the incremental fuel savings provided by electric- drive; the symbolic
dimension includes less tangible attributes, such as the owner’s desire to express a certain
value or meaning.
2.2. The private/ societal dimension: Who is it good for?
The second dimension in Table 1, private/ societal, provides a clear distinction
between electric- drive and conventional vehicles. Green ( 1992) describes a private good
as being characterized by “ exclusive and personal consumption and individual payment.”
A public ( or pro- societal) good is characterized by “ nonexclusive consumption and
collective payment” such as “ clean air” and “ saving endangered species.” Canzler ( 1999)
asserts that motor vehicles have been perceived as primarily private goods, dating back to
the original “ race- travel- limousine” vision, where initial demand was driven by goals of
luxury and prestigious racing. Electric- drive vehicles may diverge from the private good
vision, having the potential to produce pro- societal benefits such as reducing air
pollution, greenhouse gas emissions, and foreign oil dependence. Thus, electric- drive
vehicles can be associated with public welfare, leading Brown ( 2001) to classify electric
vehicles as a “ mixed” good with aspects of a private and pro- societal good. In other
words, the emergence of electric- drive vehicles may not just be an extension to the race-travel-
limousine concept, but could represent a new vision of motor vehicles as
benefiting drivers and society: the pro- societal car.
6
2.3. Attribute dynamics: How might perceptions change?
Behavioral researchers must also account for dynamics in consumer perceptions
of functional, symbolic and pro- societal attributes. For an emerging technology, attributes
in any quadrant of Table 1 may change, and may do so quickly. Functional attributes of
electric drive vehicles change with advances in battery, motor, electronics and materials
technology. Current symbolic associations may evolve, solidify or disappear and new
meanings may be added ( Heffner, et al., 2007). Also, perceptions of pro- societal benefits
will be debated and negotiated along with health and emissions research, government
regulation, and mobilization by interest groups ( Hess, 2007; Smith, 2005)— as observed
with electric- vehicles in the 1990s ( Calef and Goble, 2007) and HEVs in the last decade
( Gleick, 2007). Thus, a behavioral perspective on product adoption should represent how
such dynamics affect, and are affected by, the consumption behavior. In the following
review we consider five perspectives on the role of interpersonal influence in the
development of consumer values.
3. Reviewing perspectives of interpersonal influence and purchase behavior
Interpersonal influence literatures are rife with confusing terminology. Manski
( 2000) explains how much of this terminology is typically “ borrowed” from sociology
and social psychology and often loosely defined as “ peer influence,” “ imitation,”
“ epidemics,” and “ herd behavior.” In this review we attempt to sort through such
confusion by precisely defining and comparing five perspectives: contagion, conformity,
dissemination, translation and reflexivity. Each perspective summary is structured around
questions inspired by Bruun and Hukkinen ( 2003):
7
1. What is the new product, and what attributes are important?
2. What are the system boundaries?
3. Who purchases earlier, and why?
4. Who purchases later, and why?
5. What process drives purchase from earlier to later adopters?
A summary of these approaches and questions is presented in Table 2.
( Table 2)
3.1 Contagion: Interpersonal communication
The contagion perspective describes interpersonal influence according to the flow
of information among individuals. The term is borrowed from epidemiological studies of
how diseases are spread through populations. Contagion can alternatively be called
diffusion, a term borrowed from physics that refers to the movements of a substance from
higher to lower concentration areas. While contagion includes any interpersonal influence
approach that focuses on the effect of information flow, here we focus on DOI and social
network analysis as prevalent approaches to the spread of new products and behaviors.
8
Table 2: Comparing alternative perspectives on interpersonal influence and adoption behavior
Contagion Conformity Dissemination Translation Reflexivity
Diffusion of
innovations
Social
networks
Thresholds Critical mass Social
construction
Actor- network
theory
Modernity and
self identity
1. What is the new thing?
( Static or dynamic?)
Innovation
( static)
Innovation
( static)
Behavior
( static)
Collective good
( static)
Artifact
( dynamic)
Actor
( highly dynamic)
Lifestyle practice
( highly dynamic)
2. System boundaries?
( Static or dynamic?)
Social system of
potential
adopters, the
market
( static)
Social network,
typically a
“ bounded”
community
( static)
Relevant social
group
( static)
Social system, and
critical mass
( static)
Relevant social
groups:
consumers,
organizations,
government, etc.
( dynamic)
All actors: adopters,
social groups,
organizations,
technology, etc.
( highly dynamic)
Social system,
lifestyle sectors, in
context of
modernity
( highly dynamic)
3. Who buys first? Innovators and
early adopters
The most
“ connected”
individuals,
opinion leaders
Instigators Organizers Social groups
who perceive
artifact as a
solution to
problem
Actors who view
adoption as
consistent with their
“ actor world”
Those who find
practice is
compatible with
their self- concept
Why? Higher
“ innovativeness”
More likely to
receive info
Low threshold High interest and
resources
Interpretation of
solution
Not addressed Narrative of self
4. Who buys later? Imitators, early
to late majority,
laggards
Less
“ connected”
individuals
Conservatives Non- organizers Social groups
that later
reinterpret
problems or
solutions
Actors who are
pulled into the
“ actor worlds” of
others
Those who find
practice is
compatible with self
concept
Why? Lower
“ innovativeness”
Less likely to
receive info
High threshold,
social norms,
social learning
Efforts of
organizers and
accelerating
production
function
Closure Alignment Narrative of self
5. What drives influence? Contagion:
interpersonal
communication
Contagion:
interpersonal
communication
Conformity:
motivation to
mimic, learn from,
or join others
Dissemination:
willingness of
organizers to
achieve social
good
Interpretation:
perceived ability
of innovation to
solve a problem
Translation:
perceptions of other
actors and behaviors
according to actor
world
Reflexivity: creating
and sustaining self-concept
Best applied to what types
of attributes? ( Table 1)
Private-functional
Private-functional
Symbolic ( private
and societal)
Societal
( functional and
symbolic)
All All All
9
3.1.1. Diffusion of innovations ( DOI)
In DOI, diffusion is “ the process in which an innovation is communicated through
certain channels over time among the members of a social system… a special type of
communication, in that the messages are concerned with new ideas” ( Rogers, 2003). The
innovation is “ an idea, practice or object that is perceived as new by an individual or
other unit of adoption” ( Rogers, 2003). The likelihood of an innovation successfully
diffusing is hypothesized to depend on five main characteristics of the innovation: 1)
relative advantage over the product it replaces, 2) compatibility with existing values,
experiences and needs, 3) complexity, 4) trialability, and 5) observability of outcomes
( Rogers, 2003).
The setting for diffusion is the social system, “ a set of interrelated units that are
engaged in joint problem solving to accomplish a common goal” ( Rogers, 2003). Rather
than social systems, Moore’s ( 1999) business- oriented approach speaks of markets:
potential customers with similar needs who reference each other.
Whether they are members of social systems or markets, potential buyers are
divided into adopter categories based on the empirical observation of adoption rates
following a bell- curve over time. The first to adopt are the innovators, a sub- group
characterized as obsessively venturesome, progressive, cosmopolite individuals, usually
with a love of technology, and above average education and socioeconomic status
( Rogers, 2003). Next are the early adopters who are characterized as visionaries who use
extensive social networks to spread information about the innovation to the masses.
Following are the early majority, late majority and finally the laggards. Earlier adopters
are more interested in the functions of the innovations, and resistant later adopters are
10
eventually influenced by peer pressure and economic necessity. Individuals are placed in
adopter categories according to innovativeness: “ the degree to which an individual or
other unit of adoption is relatively earlier in adopting new ideas than other members of
the social system” ( Rogers, 2003).
DOI hypothesizes interpersonal influence is driven by communication including
word- of- mouth and mass media. Rogers ( 2003) classifies two types of information: 1)
knowledge includes basic awareness of the innovation’s existence, how it can be used,
and potentially its underlying principles, and 2) persuasion which the persuaded
individual uses to form attitudes about the overall value of adoption. Both types of
information are thought to flow from individuals in the innovator category to those in
later adopter categories, with early adopters described as playing a particularly important
role as the gatekeepers between technology loving innovators and the functionally
oriented majority ( Moore, 1999).
Thus within DOI the adoption of an electric- drive vehicle, say an HEV, would
primarily be explored from a technological, functional point of view. The targeted social
system is the entire new automobile market, though perhaps limited to buyers of vehicle
body styles in which an HEV is available. To anticipate the diffusion of HEVs,
researchers following the DOI perspective look to new car buyers with higher education
and socioeconomic status ( e. g. de Haan, et al., 2006; Santini and Vyas, 2005), perhaps
with a history of being the first to buy new products. The motives of the first HEV
buyers, defined to be innovators, are explained by their general love of technology, along
with their willingness and resources to pay a premium to be the first to own and try the
new product. After gaining experience with this technology, these innovators provide
11
feedback within their social networks that diffuses to early adopters. Early adopters may
then envision the HEV as having mass market appeal, and through their channels of
influence, accelerate the diffusion of positive HEV information to stimulate demand.
We outline several of the criticisms of DOI. First, as acknowledged by Rogers
( 2003), the concepts and language of DOI are subject to “ pro- innovation bias” because
the approach is deeply rooted in retrospective analyses of successful innovations ( e. g.
Coleman, et al., 1957; Ryan and Gross, 1943). DOI provides little insights into
innovations that fail, and is not well- suited for predictive applications— where the size
and characteristics of the social system are very difficult to ascertain a priori. Second,
DOI ignores the broader context for adoption and is vulnerable to errors of “ dependent
diffusion” where the diffusion of one product depends on another but is misread as
independent, and “ phantom diffusion” where diffusion is thought to have occurred but
did not ( Blaut, 1987). Third, DOI does not explain the underlying motivations of
consumer behavior. Instead, it relies on the vague, tautological notion of innovativeness
as both an explanatory variable and observed outcome ( Hirschman, 1980). Further,
because innovativeness is likely to be a dynamic trait, it may be inappropriate to assign
consumers to static categories ( Hirschman, 1980)
Referring to Table 1, Hirschman ( 1981) argues that DOI focuses on the private-functional
attributes of a new product and may be inappropriate for describing new
products with symbolic attributes— a limitation also noted by Rogers himself ( 2003).
Further, DOI infers that these private- functional attributes remain static throughout the
relevant time frame— neglecting the potential for substantial change over time, such as
improvements to vehicle performance, fuel savings, and variety of available designs.
12
While DOI does allow for the notion of “ re- invention” ( Rogers, 2003), this concept is not
meant to be applied to the complexities of technological development. A final criticism
relates to the process of diffusion itself, which asserts that uptake of an innovation starts
within the innovative core of a social system, and diffuses uni- directionally to the
periphery of less innovative individuals and communities ( Blaut, 1987). Blaut argues
instead for the notion of “ crisscross diffusion” in which new products, ideas and uses can
be “ generated, transmitted and received” multi- directionally across all parts of a social
system.
3.1.2 Social networks analysis ( SNA)
Instead of focusing on the individual, social network analysis ( SNA) explores
how the structure of linkages ( or ties or relationships) between individuals influences
diffusion processes ( Degenne and Forse, 1994). In SNA, the timing of adoption is
primarily determined by the “ connectedeness” of individuals to other individuals in the
network. Individuals with many social ties are more likely to adopt earlier ( Rogers and
Kincaid, 1981), corresponding with the DOI notion of early adopters being more
cosmopolitan.
Social networks are detailed representations of social systems as patterns or
structures of interpersonal communication determined by factors such as “ who talks to
whom” ( Valente, 1995). Fig. 1 illustrates the potential influence of social structure on
diffusion patterns ( Degenne and Forse, 1994). Although both networks have the same
linkage density ( the number of effective links divided by the number of potential links), if
all else is held constant including the quality of the links, the diffusion of information is
13
likely to be more rapid and complete in Network B. Also notice the particular importance
of the connection between individuals 4 and 5 in Network A; if this linkage is blocked,
further diffusion of information is impossible. Granovetter ( 1973) discusses the
importance of this phenomenon in social networks, where the existence of weak ties—
interpersonal connections scoring low in time, intimacy and reciprocity— are more likely
to serve as bridge linkages between social clusters, and can ultimately do more to create
cohesion in an extended social network than a prevalence of strong ties.
Fig. 1: Comparing network structures
Source: Degenne and Forse ( 1994, p4)
SNA is subject to many of the same general limitations as DOI: retrospective bias,
exclusion of external factors, lack of understanding consumers’ underlying motives, and
focus on static, private- functional information and attributes. Further, identifying the
appropriate social network for analysis a priori proves challenging because no single
network serves all purposes ( Bandura, 2006). Due to challenges of data collection, SNA
tends to be best suited for studying small and isolated or bounded communities, such as
small- town farmers ( Valente, 2005). A final challenge of SNA is its vulnerability to
spurious associations between adoption behavior and social proximity to other adopters.
For example, Valente ( 2005) explains that although original analysis of medical
14
innovation data from Coleman et al. ( 1957) suggests the importance of social network
influence, more recent re- analyses finds no evidence of interpersonal influence playing a
role once publicity, aggressive marketing and other external events are accounted for
( e. g. Valente, 1995; Van den Bulte and Lilien, 2001).
3.2. Conformity: Thresholds, social learning and social norms
Conformity represents interpersonal influence as occurring through an
individual’s perceptions of what others are doing or expecting. Conformity includes
applications of threshold modeling, social learning theory, and social norm research. The
conformity perspective is not as unified as DOI, and does not discuss specific processes
of communication. However, conformity provides concepts and language to help
understand adoption decisions in social contexts and the dynamics of symbolism.
Strang and Soule ( 1998) describe threshold models as breaking “ with the notion
of direct contagion to view potential adopters as responsive to the distribution of present
adopters in the population.” Granovetter ( 1978) provides a classic threshold model of
collective action or behavior which he illustrates with rioting behavior. The system
boundary for this example is a crowd, which is the relevant social group. The adoption of
rioting behavior is determined by each individual’s threshold, defined as the proportion
of fellow crowd members that must engage in the rioting behavior before the individual
will join. The first rioters are instigators with relatively low thresholds, while
conservatives riot later ( or not at all) due to their higher thresholds. Granovetter explores
how different distributions of thresholds among the crowd can significantly influence the
overall outcome, often in counterintuitive ways, concluding that the “ most important
15
causal influence on outcomes is the variation of norms and preferences within the
interacting group” ( Granovetter, 1978).
Granovetter also demonstrates how thresholds vary according to the relationships
between individuals, where a friend’s behavior may be weighted more heavily than that
of a stranger. To Granovetter, the main strength of threshold models is to understand
outcomes that appear to be inconsistent with underlying individual preferences or beliefs.
He cites the case of a group of delinquent boys that steal cars to maintain status within
their group, where each individual boy acts contrary to the norms they hold in other
contexts, e. g. “ stealing is bad,” in order to impress others in the group.
The threshold approach has also been used to explore patterns of adoption within
SNA. Valente ( 2005) describes a simple model where an individual’s adoption is
determined by their personal network exposure— the percentage of individuals in their
personal network that have already adopted the innovation. All else held constant, an
individual is most likely to adopt with higher personal network exposure, i. e., more
people in the relevant network having already adopted the new thing or behavior. An
individual with a lower threshold requires less exposure than one with a higher threshold.
More sophisticated models weigh the influence of various individuals by physical or
social proximity, similarity of social environment, or other factors.
Similar to DOI’s representation of innovativeness, threshold models typically do
not consider the origin of individual preferences ( Granovetter, 1978). Plausible
mechanisms can be drawn from two related research areas. First, social learning is based
on the same premise as threshold models, where the likelihood of adoption “ varies in
response to how common the behavior is in a relevant social group” ( Efferson, et al.,
16
2008). The driving force behind this premise is not just mimicry, but a process of
learning from the outcomes of others in order to increase personal benefits ( Efferson, et
al., 2008). In other words, the individual’s threshold preferences are based on their ability
to glean useful information from the behavior of the group. Second is social norm theory.
Cialdini ( 2003) describes two classes of norms: descriptive norms as “ perceptions of
which behaviors are typically performed”, and injunctive norms as “ perceptions of which
behavior are typically approved or disapproved”. Cialdini ( 2003) asserts that both types
of norms can pressure individuals to adopt or not adopt certain behaviors.
Conformity could influence the purchase of electric- drive vehicles in several
ways. For instance, a car buyer may want to see a certain number of HEVs on the road, or
purchased among friends and acquaintances, before they are willing to buy. A social
learning explanation could be that thresholds serve as a cue to better functional
performance of HEVs, where higher incidence of other buyers is evidence of superior
performance, reliability, or realized fuel savings. Threshold effects could also work
within more specialized groups, where an individual wants to see a certain percentage of
“ fellow environmentalists” adopt before they are convinced of the societal benefits of the
technology. On the other hand, the social norms perspective suggests that frequency
information could be used to infer a trend to HEV adoption ( a descriptive norm), or to
interpret that HEV adoption is becoming socially desirable ( an injunctive norm).
Threshold processes may help explain the symbolic attributes of pro- societal cars.
To establish a particular meaning— whether social prestige, technological advancement,
environmentalism or some other message— a certain threshold of prior adopters is
required for the meaning to be successfully conveyed. In other words, even if an
17
individual believes an HEV is environmentally- friendly, to assure this symbol will be
successfully communicated to others in their social group they might only adopt after this
belief is widely, or sufficiently, shared among their social group as demonstrated by HEV
purchase.
Although conformity helps conceptualize the influence of trends on adopters, it
does not explain how trends start. Like DOI’s failure to explain why innovators act
earlier, conformity perspectives lack explanation as to why instigators have lower
thresholds and thus act before the mechanisms of social learning or social norms can play
a role. The conformity approach does not explain the emergence of new behavior or
innovations, nor does it explain where social norms come from, or how they can change.
3.3. Dissemination: Collective action and critical mass
Rogers ( 2003) defines dissemination as “ diffusion that is directed and managed.”
We relate dissemination to processes of collective action and critical mass, which apply
specifically to issues of pro- societal goods. As noted above, pro- societal attributes are
insufficiently accounted for in DOI. In an individual- centric world, we expect pro-societal
goods to be under- provided and pro- societal cars to be “ under- adopted.” In other
words, why would an individual pay extra for an HEV to reduce environmental pollution
when the next buyer can purchase an SUV and still benefit from the HEV buyers’
contribution? However, the idea of collective action states that “ the assumption that
individuals act in isolation is usually wrong,” where in most decisions “ people are at least
generally aware of what others are doing, and often they have social relations that make
influence, or even sanctions, possible” ( Marwell, et al., 1988). In other words, motivated
18
individuals can interact and collaborate to provide pro- societal goods that would not have
been provided otherwise.
Oliver et al. ( 1985) categorize societal goods according to the shape of the
production function of marginal returns, which can be, for example, decelerating or
accelerating. Societal goods with a decelerating production function offer the biggest pro-societal
effect with the first few units of resources contributed, while those with an
accelerating production function offer larger pro- societal impacts with the addition of
later resources after larger initial start- up costs ( Oliver, et al., 1985). It is not clear which
production function shape best represents the case of electric- drive vehicles. Purely
functional- societal benefits, e. g. the contributions of each HEV sold to reducing
emissions and petroleum use, are relatively linear while private financial costs may
decrease with the development of battery technology. Symbolic- societal benefits may
follow an even more complex pattern— for example, the first few HEVs sold may have
more influence in inspiring other consumers to “ buy green” than later purchases, or these
later buyers may complete the critical mass that facilitates the shift of the balance of the
population.
The challenge of collective action is to get “ some relatively small subset of a
group interested in the provision of a public [ or pro- societal] good to make contributions
of time, money, or other resources toward the production of that good” ( Oliver, et al.,
1985). Oliver et al. ( 1985) state that cases with accelerating production functions are the
most problematic, where resolution “ depends on the rare circumstance of there being a
critical mass of persons whose combination of interests and resources is high enough to
overcome the feasibility problem.” This critical mass is defined as a “ pool of highly
19
interested and resourceful individuals willing to contribute in the initial regions of low
returns” in order to set up the conditions to sustain more widespread action ( Oliver, et al.,
1985). Where DOI labels the first adopters of private goods as innovators, in
dissemination the first to adopt pro- societal goods are members of the critical mass. Like
innovators, critical mass members tend to have extraordinarily high interest in the
product and above average access to resources ( Oliver, et al., 1985).
Critical mass theory can be applied to private goods with societal attributes.
Focusing on the diffusion of interactive media ( e. g. telephone, email, etc.), Markus
( 1987) describes the accelerating production functions associated with high start- up
network costs and susceptibility for free- ridership later on, a problem that is not
sufficiently addressed by the DOI approach. Markus ( 1987) highlights the importance of
reciprocal interdependence, where one user’s inputs depend on another users’ outputs and
vice versa. For electric- drive vehicles, potential buyers that are truly interested in
functional- societal attributes may face similar barriers; success in reducing pollution,
greenhouse gases or oil use cannot be achieved by the individual alone, but also relies on
previous and subsequent decisions by others to adopt ( and not just vehicle choice, but in
other energy- using actions also). A potential buyer might not just look to previous buyers
for information, but may also assess the likelihood of further adoption. Where earlier
buyers generally face higher private costs than later buyers, success of further purchase is
improved by the intentional coordination among some critical mass of dedicated,
resourceful pro- societal car buyers. This group acts not only through purchase of the
particular vehicle technology, but also by testing, promoting and assigning value to the
20
vehicles. These groups may be formalized in some cases, as seen with HEV driver
groups, but in most cases are less formal networks of loosely connected social groups.
The dissemination approach to diffusion can help conceptualize efforts to
coordinate the purchase of products with pro- societal attributes. These concepts and
language may be useful for further investigation of innovations involving the patterns of
production functions and reciprocal interdependence among buyers.
3.4. Translation: Social construction and interpretation
The translation perspective includes social construction of technology ( SCOT)
and actor- network theory ( ANT)— though only ANT actually uses the term translation.
Taken together, these approaches provide a rich set of concepts and language to explore
the development and adoption of new products as dynamic, socially defined artifacts.
3.4.1. Social construction of technology ( SCOT)
SCOT looks beyond the notion of adoption being driven by the diffusion of a
functionally advantageous technology ( Bruun and Hukkinen, 2003). Instead,
technological change is described as “ the culmination of a social process of interactions
that ( lead) to changed attitudes towards the ( technology) and its use” ( Bruun and
Hukkinen, 2003). The development of a technology follows a multi- directional process,
where success in a given direction is determined by the changing problems and
interpretations of relevant social groups ( Pinch and Bijker, 1984).
A newly introduced artifact— a term intended to emphasize the role of social
processes in shaping the innovation of interest— has a high degree of interpretive
21
flexibility, where social groups have differing interpretations of its meaning and content
which influence further technological development ( Pinch and Bijker, 1984). These
interpretations are “ socially and culturally embedded” where individuals in a particular
social group tend to have a common perception of a given artifact, known as a
technological frame ( Bruun and Hukkinen, 2003). In some cases the shared frame is what
defines the social group. Eventually the stages of interpretive flexibility ( or controversy)
reach a state of closure and stabilization where the perspectives of various social groups
converge with the streamlining of interpretations among them ( Bruun and Hukkinen,
2003).
SCOT was originally developed to describe the design stages of a technology
including engineering and manufacturing decisions ( e. g. Pinch and Bijker, 1984). Kline
and Pinch ( 1996) extend SCOT to analyze the use stage of innovations, exploring the
“ reciprocal relationships between artifacts and social groups… how the identities of social
groups are reconstituted in the process.” Focusing on early automobile use among rural
Americans, the authors illustrate how the “ anti- car crusade’s” initially negative
interpretations of automobiles were gradually overcome by positive interpretations that
were both functional, e. g. providing stationary assistance for farm tasks, and symbolic,
e. g. reinforcing gender roles. In addition to demonstrating how different social groups
can shape the development of a technology, Kline and Pinch ( 1996) highlight how the
development of the automobile also transformed the rural social groups, increasing the
connectivity of communities and allowing new methods of saving labor.
The social construction process can be represented visually using a conceptual
diagram to depict relationships between social groups and their problems on the one hand
22
and artifacts and their solutions on the other. In Fig. 2, we adapt Pinch and Bijker’s
( 1984) bicycle diagram to the case of electric- drive vehicles and pro- societal cars. Part A
presents one conceptualization of how electric- drive vehicles might have developed in
the U. S. auto market. In the early 1990s, full- performance EVs such as General Motor’s
EV- 1 proved unsuccessful. Further development could have followed multiple paths,
including lower- range, lower- power neighborhood or regional EVs, or the development
and deployment of HEVs. Ultimately, HEVs proved successful with the release of “ full”
HEVs, such as the Honda Insight and the Toyota Prius. As these vehicles gained
popularity, other manufacturers released alternative models utilizing drivetrains with less
fuel efficient hybridization, e. g. the Honda Accord Hybrid, or applying hybridization to
the use of power tools in “ contractor” applications, e. g. the Chevy Silverado Hybrid.
Both alternate pathways have so far failed to achieve market success. In recent years,
PHEV technology has received substantial attention as the next potential stage of
development. This potential is noted by a question mark, though if consumers do not
eventually form positive interprets of PHEVs the technology would also fail ( Fig. 2)
Part B of Fig. 2 conceptualizes the development of pro- societal cars on a symbolic
level. Starting with the dominant race- travel- limo vision of automobiles described in
Section 2.2, the practical notion of fuel economy emerged as a potential alternative
pathway in the aftermath of the energy crises of the 1970s, moving towards smaller,
cheaper, more fuel efficient cars. However, this pathway gradually disappeared with the
advent of less efficient minivans and SUVs and cheaper energy in the 1980s— regressing
back to the race- travel- limo vision. More recently, with the growing popularity of HEVs,
the pro- societal car concept has emerged as a potential contender.
23
Fig. 2: Illustrative depiction of quasi- linear development of innovations
Source: Adapted from Pinch and Bijker’s ( 1984, p413) depiction of the Penny- Farthing
Bicycle
While the diagrams in Fig. 2 help to conceptualize the trajectory of different
technologies and symbols, SCOT’s greater strength is illustrating how the interpretations
of various social groups guide this development process. Fig. 3 is another adaptation
from Pinch and Bijker ( 1984), illustrating the potential drivers of the symbolic
development in Part B of Fig. 2. The dominant vision of the race- travel- limo is
represented in the gray hexagon, surrounded by several social groups with differing
interpretations of what problems need to be solved by technological developments. For
instance, consumers may be concerned about high gas prices while environmentalists and
governments may focus on reducing greenhouse gas ( GHG) emissions.
24
Fig. 3: Illustration of relevant social groups, problems, and solutions in the
development of the pro- social car
Source: Adapted from Pinch and Bijker’s ( 1984, p413) depiction of the Penny- Farthing
Bicycle
The fuel economy technology is presented as a potential direction of
development, consisting of smaller, more fuel- efficient vehicles using conventional
gasoline engines. This efficiency solution solves one problem of consumers ( fuel costs)
and one problem of environmentalists and governments ( GHG emissions). However,
25
consumers have the additional problem of using automobiles as a form of self- expression:
they may interpret this as being unachievable by the smaller, cheaper, low- power fuel
economy vehicles. In contrast, pro- societal cars emerge as a technology that not only
reduces fuel costs and GHG emissions through increased efficiency, but also provide a
visible, higher price, technologically advanced symbol that meets the symbolic needs of
many consumers. If other social groups eventually yield similarly positive interpretations
of pro- societal cars, the overall interpretation of the pro- societal car as successor to the
race- travel- limo could reach a state of interpretive closure.
Brown ( 2001) provides a further addendum to the original SCOT approach,
asserting that certain social groups within the SCOT framework can have particularly
powerful influence over the interpretations of other social groups. For instance, when the
California Air Resources Board ( CARB), a government agency, established the zero-emissions
vehicle ( ZEV) program in the early 1990s, clean air benefits were highlighted
as important criteria in the technological development of vehicles. Brown ( 2001) argues
that CARB’s statements and actions served to reopen the interpretive flexibility of the
race- travel- limo vision, prompting consumers to consider social attributes as an important
concern for buyers of motor vehicles. Thus, social groups that are not actual buyers can
also play an important role in influencing the interpretations of consumer groups.
Overall, SCOT is useful for conceptualizing the dynamics of an innovation, the
social groups that can guide development ( including non- consumers), and the interplay
between competing problems and interpretations. However, SCOT alone is not sufficient
to explore the diffusion and purchase process; its origins in design stage applications
make it less appropriate for more complex problems of user groups and symbolic
26
interpretations. For instance, Hannemyr ( 2003) notes that the SCOT concept of closure is
overly definitive, where interpretations may streamline at times, but may not fully
converge, and any convergence may only be temporary. Moreover, SCOT does not
account for processes of social action in technological controversies, or explain how the
overall structure of social groups may be heavily influenced or even defined by the
technology ( Bruun and Hukkinen, 2003).
3.4.2. Actor- network theory ( ANT)
Due to such limitations of SCOT, researchers often draw from its less tangible
cousin: actor- network theory ( ANT) ( e. g. Bruun and Hukkinen, 2003; Hannemyr, 2003).
The abstract nature of ANT is its strength and weakness. Unlike SCOT, which
conceptualizes a structure of roles and relationships among social groups and
technologies, ANT provides a level playing field on which everything is an actor—
people, groups, ideas, objects, and infrastructure. All actors “ take their form and acquire
their attributes as a result of their relations with other entities” ( Law and Hassard, 1999).
The only differences among actors are the “ methods and materials that they deploy to
generate themselves” ( Law, 1992), where relationships and social structures are
extremely dynamic. Similar to later applications of SCOT, ANT asserts that while social
structure can influence technological change, the reverse is also true. The difference is
that ANT steps further in stating that a social group can only be defined by its
relationship with the technology and other actors; every actor is defined by its
interactions with other actors. Taken in another light, while SCOT states that
27
technologies are socially defined with malleable interpretations, ANT states that the
entire network, social and otherwise, is just as fluid.
The concept of translation is perhaps the greatest contribution of ANT to a
discussion of interpersonal influence. Where diffusion represents the propagation of a
static idea or object, translation emphasizes that these ideas and objects change as a result
of context and interactions among actors ( Pentland and Feldman, 2007). Bruun and
Hukkinen ( 2003) define translation as “ the mechanism through which actors can
transform themselves, displacing their own identity as well as that of others.” For some
applications, translation may be a more accurate representation of how complex ideas
spread among actors and social groups, similar to Blaut’s ( 1987) concept of crisscross
diffusion where reinvention is a continuous aspect of the communication process. For
instance, information regarding a product’s symbolic and pro- societal attributes may be
better described as translation, where interpretations of meaning are continually refined
and negotiated among users and observers.
ANT also provides a less definitive concept of closure than SCOT, where an
innovation may have multiple scripts which are “ mediated, translated, and even changed
as time passes, being the product of domination, negotiation, and mutual adjustment”
( Hannemyr, 2003). Instead of a finalization of interpretations, ANT presents the concepts
of “ alignment,” measuring the degree of agreement for a certain translation, and
“ coordination,” the restriction of interpretive flexibility by rules or conventions ( Callon,
1991). Thus, while translations may occasionally streamline through negotiations among
social groups, they are always open to revision.
28
Callon ( 1986) applies ANT to the case of electric vehicles in France during the
1970s. He describes how Electricité de France’s ( EDF) presentation of a plan for the
deployment of EVs could be represented as an actor- world— a vision of the roles required
by all relevant actors. For EDF’s vision to play out, consumers would have to be
interested in buying the EVs, automakers would have to be willing to shift manufacturing
efforts, governments would have to enact pro- EV legislation, and electrochemical
batteries— another actor and network of further components— would have to perform
adequately. In this sense, Callon describe the EDF as a “ spokesman” for the actor
network, translating the roles and relationships among actors. In turn, any of these actors
can reject such assigned roles and thus prevent EDF’s envisioned actor- world from
manifesting. Callon ( 1986, p32) summarizes that the actor network is “ distinguished from
a simple network because its elements are both heterogeneous and are mutually defined
in the course of their association.”
Fig. 4 presents a potential actor- network perspective on electric- drive vehicles,
adapted from Callon’s ( 1986) portrayal of EVs. The relevant actor- world is made of
many different actors, each of which contains its own network of sorts. The structure of
each network may differ: car buyers may include different social networks, governments
include a hierarchical order of bodies, and advanced batteries are made up of specific
components which are made up by still more specific components and organized by
researchers.
29
Fig. 4: Illustration of the pro- societal car actor- world
Source: adapted from Callon ( 1986).
In application to pro- societal cars, ANT provides flexibility to the relatively rigid
structure of SCOT. For instance, ANT allows for every item in Fig. 4 to experience
transformation with development and adoption, including the technologies, social groups,
problems and solutions. For instance, the social groups are in part defined by their
interpretations of motor vehicle technologies, and some interpretations of problems, e. g.,
GHGs, may not have emerged until after the presentation of their solution, e. g., pro-societal
cars. The translations of other social groups may stimulate an individual to
reinterpret their own problems. The individual may then become part of a new social
30
group of pro- societal car buyers whose existence results from the emergence of pro-societal
cars, rather than preceding it.
However, the flexibility of ANT is also a weakness for the present application.
ANT includes no causal theory of action ( Bruun and Hukkinen, 2003), and can not serve
as a predictive model of adoption. However, taken as a supplement to the more rigid
structure of SCOT, the concepts and language of ANT enhance the ability of the
translation perspective to facilitate the discussion of adoption processes, including the
roles of and relationships among multiple heterogeneous social groups, and the dynamic,
reciprocal relationships between all the actors involved, including the innovation itself.
The explicit incorporation of individual and group- based interpretations and translation
can aid the investigation of innovations with attributes in all categories of Table 1.
3.5 Reflexivity: Modernity and the project of the self
The fifth perspective explored in this paper is reflexivity as drawn from Giddens’
( 1991) approach to self development in the social world. Although this perspective does
not specifically focus on consumption or purchase behavior, Giddens theorizes about the
underlying driving forces of human behavior ( including consumption) neglected in the
four perspectives presented thus far.
Reflexivity starts with Giddens’ description of modern life “ propel[ ling] social
life away from the hold of pre- established precept or practices” ( 1991), which we present
here as the overall context for the generation and spread of new behaviors. In traditional
society an individual had a defined role with expected set of behaviors and interactions
with others. In contrast, modernity provides little such direction. Instead, modernity is
31
characterized by uncertainty in which individuals must actively seek out and create their
self- identity. Without the guidance of traditional roles, “ the self becomes a reflexive
project” ( Giddens, 1991), where reflexivity is defined as the dynamic, continuous process
of defining and expressing oneself. The self is understood through a reflexively defined
biography, or narrative, linking an individual’s past, present and future into a cohesive
trajectory of development. This narrative must continually sort events from the outside
world into an ongoing story about the self. The ultimate goal of this process of self-exploration
is self- actualization in the sense of authenticity or being true to oneself
( Giddens, 1991). In short, to cope with the uncertainty endemic to modernity, our
behavior is guided by efforts to establish a sense of order, direction, and development for
our self- concepts.
Given the vast selection of possible choices an individual faces on a daily basis,
Giddens’ ( 1991) describes the importance of lifestyles as “ a more or less integrated sense
of practices which an individual embraces, not only because such practices fulfill
utilitarian needs, but because they give material form to a particular narrative of self
identity.” Instead of agonizing over every choice in effort to create and extend an
authentic self- narrative, individuals seek a lifestyle as a package of practices that are
associated with their particular trajectory. These practices include fashion, eating, and
any other form of symbolic display, such as vehicle purchase and ownership. In essence,
lifestyle is the “ core” of self identity in the context of modernity ( Giddens, 1991).
However, a lifestyle is by no means a static package, but also follows a reflexive process,
continually open to re- evaluation and negotiation.
32
Giddens describes lifestyle sectors which may be divided according to locales as
settings for specific sets of practices, e. g. home and workplace, relationships, e. g.
marriage and friendship, or activities, e. g. work and recreation. Lifestyle sectors are a
dynamic and more theoretically elaborate version of the relevant social group concept
discussed in other adoption perspectives. Similar to ANT, Giddens’ description of
reflexivity highlights the indefinite and relationally defined nature of lifestyle sectors. But
Giddens adds to ANT a theoretical interpretation of why and how the process of
translation occurs in human actors.
Lifestyle and lifestyle sectors can explain aspects of other perspectives reviewed
above. DOI’s concept of innovativeness may relate to a certain type of lifestyle adopted
by individuals who define themselves as cutting edge, technologically advanced
individuals, and the practices of this lifestyle include the purchase and use of new
technologies. The critical mass groups described in the dissemination perspective may
represent another lifestyle adopted by people who want to develop and portray
themselves as environmentally aware, socially active individuals, including the practice
of driving pro- societal cars and helping to establish positive interpretations of their pro-societal
benefits.
In Giddens’ ( 1991) framework, increased ecological concern among individuals
relates to the “ recognition that reversing the degradation of the environment depends
upon adopting new lifestyle patterns.” Thus, the adoption of pro- societal cars is not just
driven by a motivation for advantageous functional or symbolic attributes, but may
instead be one component, or trial, of a more fundamental shift towards an environmental
or socially- conscious lifestyle. The visible nature of the pro- societal car can facilitate
33
reflexivity by prompting users and observers to share and negotiate interpretations.
Observers may speculate as to the motivations and lifestyle choices of the driver,
assessing if such a practice might fit into their own self trajectory. After adoption, a user
may solidify initial interpretations of the vehicle, or modify interpretations based on their
experiences and feedback from personal contact and the media. Thus, similar to ANT, the
context of modernity is uncertain and interpretations of other actors are subject to
constant revision.
An individual’s self- concept and commitment to lifestyle practices may be more
or less stable subject to different conditions. Turner ( 1969) describes the concept of
liminality as a state in which an individual is “ betwixt and between the positions and
assigned by law, custom, [ and] convention.” A liminal state is characterized by
“ ambiguous and indeterminate attributes,” which contrast with the clearly defined
attributes of a stable “ status system.” Liminality may be associated with an individual’s
life stage, where, for example, they experience higher liminality as a college student or
new retiree, and less liminality as a parent with stable child care responsibilities.
Liminality may also relate to the structure of an individual’s social network, including the
diversity of individuals and social groups that they interact with. In this sense, reflexivity
is associated with the stability or liminality of self- concept and lifestyle practices.
Ultimately, the reflexivity perspective is not meant as a stand- alone approach to
interpersonal influence and consumption behavior. However, it does provide a theoretical
backdrop to the other four perspectives reviewed in this paper. Most notably, Giddens
supplements the translation approach, particularly ANT, by explaining how and why the
multi- directional relationships among actors occur.
34
4. Discussion and conclusion
Section 2 described the complex nature of electric- drive vehicles in their potential
to offer important functional, symbolic and societal benefits. In Section 3 we reviewed
how the dominant perspective on interpersonal influence and consumption— contagion or
diffusion— focuses on the communication of information concerning the functional
attributes of an innovation among adopter categories. We suggest that contagion
approaches such as DOI are not conceptually equipped to explain the additional
complexity of symbolic and pro- social behaviors or artifacts.
The exploration of four additional perspectives, which we label conformity,
dissemination, translation and reflexivity, yields insights into the complexities of the
consumption behaviors for different aspects of an innovation. Conformity highlights how
the behaviors of others can influence an individual’s interpretations of an innovation, as
well as tendencies to observe or oppose existing social norms. Dissemination explores the
adoption of goods with pro- societal attributes, addressing the important role of an initial
critical mass willing to accept high start up costs. Translation describes the socially
dynamic nature of innovations, such that interpretations are continuously redefined and
renegotiated among social groups who are themselves being redefined and renegotiated.
Finally, reflexivity provides a theoretical backdrop to the underlying motivations of
adoption processes, describing the individual as a work in progress, continually searching
for self development and expression through lifestyle practices.
Table 2 presents a summary of how the five perspectives address the five
questions posed at the beginning of Section 3. The bottom row highlights the types of
product attributes a given approach may be designed or particularly well suited for— as
35
conceptualized in Table 1. Note that these perspectives are in some sense
complementary; viewed together, each of these perspectives may yield insights into how
interpersonal interactions influence the adoption of electric drive vehicles. DOI was
designed, and is suited, primarily for private- functional attributes and the flow of
functional information; conformity holds potential for symbolic attributes; dissemination
describes intentional efforts to promote the adoption of products with pro- societal
attributes; and both the translation and reflexivity perspectives address all types of
attributes through the interpretations of individuals and social groups. However, no single
approach reviewed here seems alone adequate for the study of interpersonal influence and
pro- societal consumption— not even the last two; translation does not discuss specific
mechanisms of communication and adoption behavior, while reflexivity only provides a
theoretical backdrop to processes of adoption.
The intent of this review is to identity alternatives to the diffusion perspective, to
illuminate key differences among these perspectives, and help mitigate confusion among
literatures of interpersonal influence and consumption. We advise that researchers
investigating this complex topic be explicit in identifying the perspective or perspectives
they are employing, and explaining strengths and weaknesses of their approach. We also
point to many potential directions for future research. In addition to improving contagion
models, research could further explore the potential use of conformity models to explain
symbolic valuation, as well as the role of dissemination processes that may be less formal
than the organized social movements addressed by collective action and critical mass
approaches. Further, while models of translation and reflexivity can effectively represent
36
social dynamics over time, further research can investigate their potential for predictive
applications, and implications for environmental policy.
Although Table 2 matches each perspective with the different types of attributes
of a new technology, we do not necessarily suggest that each perspective should be
independently applied to these different attributes. Such a patchwork of concepts and
language might not be particularly useful or interesting. Instead, this discussion could be
used as fodder for the construction of an integrative model of interpersonal influence and
consumption that can address all the relevant attributes of pro- societal goods and
consumption practices. A major lesson from the popularity of DOI is that a simple,
common language can enormously enhance the communicability and longevity of a
research approach. In a companion paper, we make our own attempt: we propose such an
integrative model using concepts explored in this review, apply it to empirical data, and
consider implications for environmental policy ( Axsen and Kurani, 2010). Further
research should explore the validity and usefulness of each perspective in various
technological, regional and cultural contexts, and perhaps develop alternative integrative
models.
Acknowledgements:
The authors would like to thank the California Air Resources Board and the Social
Sciences and Humanities Research Council of Canada. Thanks also to Dr. Thomas S.
Turrentine for his thoughtful input.
37
References
Adamson, K. A., 2003. An examination of consumer demand in the secondary niche
market for fuel cell vehicles in Europe. International Journal of Hydrogen Energy 28( 7),
771- 780.
Axsen, J., Mountain, D. C. and Jaccard, M., 2009. Combining stated and revealed choice
research to simulate the neighbor effect: The case of hybrid- electric vehicles. Resource
and Energy Economics 31( 3), 221- 238.
Axsen, J. and Kurani, K., 2010. Reflexive Layers of Influence ( RLI): An integrative
framework on interpersonal influence and pro- societal consumption, University of
California, Davis, Working Paper, UCD- ITS- WP- 10- 03.
Bandura, A., 2006. On integrating social cognitive and social diffusion theories. In: A.
Singhal and J. Dearing ( Eds.), Communication of Innovations: A Journey with Ev
Rogers, SAGE, Thousand Oaks, CA, 111- 135.
Blaut, J. M., 1987. Diffusionism - A uniformitarian critique. Annals of the Association of
American Geographers 77( 1), 30- 47.
Brown, M. B., 2001. The civic shaping of technology: California's electric vehicle
program. Science Technology & Human Values 26( 1), 56- 81.
Bruun, H. and Hukkinen, A., 2003. Crossing boundaries: An integrative framework for
studying technological change. Social Studies of Science 33( 1), 95- 116.
Calef, D. and Goble, R., 2007. The allure of technology: How France and California
promoted electric and hybrid vehicles to reduce urban air pollution. Policy Sciences
40( 1), 1- 34.
Callon, M., 1986. The Sociology of an Actor- Network: The Case of the Electric Vehicle.
In: M. Callon, J. Law and A. Rip ( Eds.), Mapping the Dynamics of Science and
Technology: Sociology of Science in the Real World, The Macmillan Press, London, 19-
34.
Callon, M., 1991. Techno- economic networks and irreversibility. In: J. Law ( Eds.), A
Sociology of Monsters: Essays on Power, Technology and Domination, Routledge:
London, 132- 161.
Canzler, W., 1999. Changing speed? From the private car to CashCar Sharing. SPEED -
A workshop on space, time and mobility, Copenhagen, 23- 31.
Cialdini, R. B., 2003. Crafting normative messages to protect the environment. Current
Directions in Psychological Science 12( 4), 105- 109.
38
Coleman, J., Menzel, H. and Katz, E., 1957. The diffusion of an innovation among
physicians. Sociometry 20, 253- 270.
de Haan, P., Peters, A. and Mueller, M., 2006. Comparison of Buyers of Hybrid and
Conventional Internal Combustion Engine Automobiles. Transportation Research
Record: Journal of the Transportation Research Board 1983, 106- 113.
Degenne, A. and Forse, M., 1994. Introducing Social Networks. SAGE Publications,
London.
Efferson, C., Lalive, R., Richerson, P. J., McElreath, R. and Lubell, M., 2008.
Conformists and mavericks: the empirics of frequency- dependent cultural transmission.
Evolution and Human Behavior 29( 1), 56- 64.
Ehrnberg, E., 1995. On the definition and measurement of technological discontinuities.
Technovation 15( 7), 437- 452.
Gartman, D., 2004. Three ages of the automobile - The cultural logics of the car. Theory
Culture & Society 21( 4- 5), 169.
Giddens, A., 1991. Modernity and Self- Identity: Self and Society in the Late Modern
Age. Stanford University Press, Stanford, CA.
Gleick, P., 2007. Hummer versus Prius: " Dust to Dust" report misleads the media and the
public with bad science. Pacific Institute, Oakland, CA.
Granovetter, M., 1973. Strength of weak ties. American Journal of Sociology 78( 6),
1360- 1380.
Granovetter, M., 1978. Threshold models of collective behavior. American Journal of
Sociology 83( 6), 1420- 1443.
Green, D. P., 1992. The price elasticity of mass preferences. American Political Science
Review 86( 1), 128- 148.
Hannemyr, G., 2003. The Internet as hyperbole: A critical examination of adoption rates.
Information Society 19( 2), 111- 121.
Heffner, R. R., Kurani, K. S. and Turrentine, T. S., 2007. Symbolism in California's early
market for hybrid electric vehicles. Transportation Research Part D- Transport and
Environment 12( 6), 396- 413.
Hess, D., 2007. What is a clean bus? Object conflicts in the greening of urban transit.
Sustainability: Science, Practice and Policy 3( 1), 45- 58.
39
Hirschman, E. C., 1980. Innovativeness, novelty seeking, and consumer creativity
Journal of Consumer Research 7( 3), 283- 295.
Hirschman, E. C., 1981. Symbolism and Technology as Sources for the Generation of
Innovations. Advances in Consumer Research 9( 537.
Jackson, T., 2005. Motivating Sustainable Consumption: A Review of Evidence on
Consumer Behaviour and Behavioural Change. Sustainable Development Research
Network, University of Surrey, Guildford, Surrey, UK.
Kline, R. and Pinch, T., 1996. Users as agents of technological change: The social
construction of the automobile in the rural United States. Technology and Culture 37( 4),
763- 795.
Law, J., 1992. Notes on the theory of the actor network: Ordering, strategy, and
heterogeneity Systems Practice 5( 4), 379- 393.
Law, J. and Hassard, J., 1999. Actor Network Theory and After. Blackwell, Oxford, UK
Manski, C., 2000. Economic analysis of social interactions. The Journal of Economic
Perspectives 14( 3), 115- 136.
Markus, M. L., 1987. Toward a critical mass theory of interactive media: Universal
access, interdependence and diffusion Communication Research 14( 5), 491- 511.
Marwell, G., Oliver, P. E. and Prahl, R., 1988. Social networks and collective action: A
theory of the critical mass 3. American Journal of Sociology 94( 3), 502- 534.
Mau, P., Eyzaguirre, J., Jaccard, M., Collins- Dodd, C. and Tiedemann, K., 2008. The
` neighbor effect': Simulating dynamics in consumer preferences for new vehicle
technologies. Ecological Economics 68( 1- 2), 504- 516.
Moore, G., 1999. Crossing the Chasm: Marketing and Selling Disruptive Products to
Mainstream Customers. Collins Business Essentials, New York, New York.
Norton, B., Costanza, R. and Bishop, R., 1998. The evolution of preferences: Why
' sovereign' preferences may not lead to sustainable policies and what to do about it.
Ecological Economics 24, 193- 211.
Oliver, P. and Marwell, G., 2001. Whatever happened to critical mass theory? A
retrospective and assessment. Sociological Theory 19( 3), 292- 311.
Oliver, P., Marwell, G. and Teixeira, R., 1985. A theory of the critical mass 1.
Interdependence, group heterogeneity, and the production of collective action American
Journal of Sociology 91( 3), 522- 556.
40
Pentland, B. T. and Feldman, M. S., 2007. Narrative networks: Patterns of technology
and organization. Organization Science 18( 5), 781- 795.
Pinch, T. J. and Bijker, W. E., 1984. The social construction of facts and artifacts - Or
how the sociology of science and the sociology of technology might benefit each other.
Social Studies of Science 14( 3), 399- 441.
Robertson, T., 1971. Innovative Behavior and Communication. Free Press, New York.
Rogers, E., 2003. Diffusion of Innovations ( 5 th Ed.). Free Press, New York.
Rogers, E. and Kincaid, D., 1981. Communication Networks: A New Paradigm for
Research. Free Press, New York.
Ryan, R. and Gross, N., 1943. The diffusion of hybrid seed corn in two Iowa
communities. Rural Sociology 8( 1), 15- 25.
Santini, D. J. and Vyas, A. D., 2005. Introduction of hybrid and diesel vehicles - Status
within the life cycle of technology adoption. In: ( Eds.), Energy and Environmental
Concerns 2005, 18- 25.
Smith, A., 2005. The alternative technology movement: An analysis of its framing and
negotiation of technology development. Human Ecology Review 12( 2), 106- 119.
Steg, L., 2005. Car use: lust and must. Instrumental, symbolic and affective motives for
car use. Transportation Research Part a- Policy and Practice 39( 2- 3), 147- 162.
Steg, L., Vlek, C. and Slotegraaf, G., 2001. Instrumental- reasoned and symbolic- affective
motives for using a motor car. Transportation Research Part F 4, 151- 169.
Strang, D. and Soule, S. A., 1998. Diffusion in organizations and social movements:
From hybrid corn to poison pills. Annual Review of Sociology 24, 265- 290.
Struben, J. and Sterman, J., 2008. Transition challenges for alternative fuel vehicles and
transportation systems. Environment and Planning B: Planning and Design 35( 6), 1070-
1097.
Turner, V., 1969. The Ritual Process: Structure and Anti- Structure. Aldine Publishing,
Chicago, Il.
Valente, T., 1995. Network Models of the Diffusion of Innovations. Hampton Press,
Cresskill, New Jersey.
Valente, T., 2005. Network Models and Methods for Studying the Diffusion of
Innovations. In: P. Carrington, J. Scott and S. Wasserman ( Eds.), Models and Methods in
Social Network Analysis, Cambridge University Press, New York, New York, 98- 116.
41
Van den Bulte, C. and Lilien, G., 2001. Medical innovation revisited: Social contagion
versus marketing efforts. American Journal of Sociology 106, 1409- 1435.
van Rijnsoever, F., Farla, J. and Dijst, M. J., 2009. Consumer car preferences and
information search channels. Transportation Research Part D: Transport and Environment
14( 5), 334- 342.