See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/291824703

Technological Innovation as a Process

Chapter · January 1990

CITATIONS READS
81 20,999

2 authors, including:

Jd Eveland

37 PUBLICATIONS 968 CITATIONS

SEE PROFILE

All content following this page was uploaded by Jd Eveland on 25 January 2016.

The user has requested enhancement of the downloaded file.

3
Technological Innovation
as a Process

H
aving bounded our i n q u i r y to that of understanding the evolution of
knowledge-based technologies, w e w i l l n o w open up the l i d of that
black b o x . I n this chapter, w e w i l l describe t w o of the most i m p o r -
tant analytic problems that m a k e technological i n n o v a t i o n so fascinating as
well as so difficult to u n d e r s t a n d . T h e y are:

1. how to portray technological i n n o v a t i o n as a process of m a n y discrete

decisions and behaviors that u n f o l d s l o w l y over time
2. how to parse out the involvement of social units at m a n y different levels
of aggregation ( i n c l u d i n g i n d i v i d u a l s , groups, organizations, industries,
and economies) i n technological i n n o v a t i o n .

The Life C y c l e of T e c h n o l o g i c a l Innovation

Technological i n n o v a t i o n is neither a single event nor even a small number

of discrete events. T e c h n o l o g i c a l i n n o v a t i o n involves a rich embroidery of
events: many activities, m a n y decisions, a n d m a n y bits of behavior on the part
of individuals and social u n i t s , most of w h o m are not even self-consciously
aware of being part of such a process.
One w a y of appreciating the vast n u m b e r of discrete events that are
encompassed i n the process of technological i n n o v a t i o n is to realize h o w long
that process takes to u n f o l d . F r o m initial idea generation ( w h i c h m a y or m a y
not derive directly f r o m scientific research) to widespread use of a technologi-
cal innovation by an appropriate p o p u l a t i o n of users, the time can range f r o m
months to decades. T h e e n v i r o n m e n t a l pressure exerted by a user c o m m u n i t y
may accelerate the o v e r a l l process. T h e first A I D S drugs appeared w i t h i n a
matter of months after p u b l i c consciousness about the severity of the disorder,
while in contrast, the first c o m m e r c i a l photocopying products by X e r o x

This chapter was written by L o u i s G . T o r n a t z k y , J . D . E v e l a n d , and M i t c h e l l Fleischer.

 28 • The Processes of Technological Innovation

involved decades o f development, a n d the c o n v i n c i n g of potential users that

they needed photocopying after a l l . T h e n o r m is a l o n g , convoluted period
(see table 3 - 1 ) .
T h a t span o f time is also correlated w i t h the activities of significant n u m -
bers o f people. T h e s e m a y range f r o m m a n y thousands o f person-years i n the
development, design, a n d m a n u f a c t u r e o f a nuclear s u b m a r i n e , to a few

TE person-months that might encompass a more defined technology—such as a

n e w computer p r o g r a m . Nonetheless, these are b r o a d swaths o f behavior;
these are people's lives. I s there a w a y i n w h i c h w e c a n organize o u r t h i n k i n g
about these processes? A n appropriate conceptual metaphor is that of a life
cycle i n i n n o v a t i o n . T h a t i s , technological innovations have some intellectual
. b i r t h i n g ( s ) , go through a developmental process, m a t u r e , a n d then at some
p o i n t , after functioning for more o r less time as part o f the fabric of everyday
life, die a n d are replaced by further i n n o v a t i o n s .
A s an analogy, the life cycle n o t i o n has considerable p o w e r . F o r e x a m p l e ,
'y there is a m a j o r division i n technological i n n o v a t i o n that separates " p a r e n t "
and " c h i l d . " T h a t i s , there is one set o f processes that deal w i t h the creation,
P^^'^ n u r t u r i n g , a n d delivery o f n e w technologies; there is another set o f activities
' concerned w i t h adoption a n d use. T h e s e correspond respectively to parts I I
a n d I I I o f this v o l u m e .
on T o elaborate the concept o f life cycle, some analysts have developed
cor models defining stages o f i n n o v a t i o n . Stages i n this sense are generally defined
anc by one or more decisions and related behavior that are connected i n some logi-
P cal fashion a n d accumulate a body o f experience. Stage models tend to take
one o f t w o very distinct perspectives. T h e y either focus o n the creation/
^ p r o d u c t i o n of n e w technology, o r they focus o n h o w the technology gets used.
coi F r o m the point o f v i e w of the source or p r o d u c t i o n o f technology, stages are
ini often defined as some v a r i a n t o n the pattern:
pa
he 1. basic research
va 2. applied research
3. development

Table 3 - 1
Length of the Innovation Process
I
Technology Initial Proof-of-Concept General Use

Television 1920s 1940s

Anesthesia 1844 1870-1880
Digital electronic computer 1939 1950s
Incandescent light 1879 1900s
Jet propulsion 1939 1950s
 Technological Innovation as a Process • 29

4. testing or evaluating
5. manufacturing or p a c k a g i n g
6. marketing or dissemination

As these stages proceed, the i n n o v a t i o n becomes defined w i t h greater specif-

icity, both for w h a t it is a n d w h a t it is not. A t the conclusion of these stages,
one has a definite n e w p r o d u c t , service, or both in h a n d , although that thing
may be as much social a n d p r o c e d u r a l as it is h a r d w a r e or p h y s i c a l object.
Speaking postnatally n o w , the stages f r o m the perspective of the user of
technology generally highlight decisions a n d actions concerning the incorpo-
ration of a new tool or practice. Stages f r o m this perspective are generally a
variant on the pattern:

1. awareness-problems
2. matching-selection
3. adoption-commitment
4. implementation
5. routinization

As in the developer's perspective, the user's perspective also involves a

process of specification, of c o m i n g to understand w h a t the i n n o v a t i o n is and
is not going to be used f o r , and just w h a t the nature of its interconnections
with other parts of the system m a y be. I n fact, the stage of r o u t i n i z a t i o n is
somedmes defined by the i n a b i l i t y of people to m a k e any more decisions
about the innovation, leaving it to evolve no further.^ B o t h sets of models
implicidy reflect a significant degree of "uncertainty a b s o r p t i o n " ( M a r c h and
Simon 1958), in w h i c h m u c h idiosyncratic i n f o r m a t i o n (the differing v i e w s of
individuals) is lost or submerged as people come increasingly to share the
same definitions of the i n n o v a t i o n .
It is difficult to combine these t w o perspectives into a single sequence. I n
general, the source stages occur before the user stages a n d i n v o l v e a different
set of actors, but not i n e v i t a b l y ; some studies ( v o n H i p p e l , 1 9 7 6 , 1 9 8 8 ) have
noted crucial interactions between sources a n d users that guide the develop-
ment of particular technologies. M o r e o v e r , m a n y users are their o w n sources,
and the user/source interactions take place w i t h i n the same organizational
structure. E v e n w h e n source a n d user are apparently distinct, m u c h "source
type" behavior occurs i n the user o r g a n i z a t i o n . F u r t h e r development of tech-
nology almost a l w a y s takes place w i t h i n the user o r g a n i z a t i o n , p a r t i c u l a r l y
the crucial processes of custom-fitting the technology to the organization (Pelz
andMunson 1 9 8 0 ) . T h u s , the t w o approaches c a n be considered i n t e r t w i n e d
throughout, although they c a n be distinguished for different types of analyses
of innovadon ( H a v e l o c k a n d H a v e l o c k 1 9 7 3 ) .
One key to the distinction m a y be the element of interconnectivity. T h a t
 30 • The Processes of Technological Innovation

i s , source-centered models tend to emphasize the distinctiveness of the tech-

n o l o g y — w h a t makes it different f r o m all other things o n the m a r k e t . E a c h
stage of the model is devoted to r e m o v i n g successive bits of ambiguity about
the i n n o v a t i o n ; the end is the thing standing alone. B y contrast, user-centered
models emphasize the integration of the i n n o v a t i o n into a larger system, to
the point where the final stage is defined by its essential i n v i s i b i l i t y , by our
inability to tell w h a t is n e w or unique about the system. A t some point among
users, that w h i c h w a s once i n n o v a t i v e a n d n e w becomes embedded in a larger
number of tools a n d procedures, and " r o u t i n i z e d " ( Y i n 1 9 7 9 ) .

Nonlinearity

T
W h i l e stage models have a certain c o m f o r t i n g appeal i n that they provide an
intellectual organizing theme for m a n y events and large blocks of time, there
are distinct limitations to this a p p r o a c h . O n e is that neither life nor i n n o v a -
pri
tion are at all simple or linear. W h i l e there are some logical priorities i n the
At
sequence of stages a n d phases (one cannot disseminate something one has
OV(
never heard of; implementation generally must f o l l o w some decision to adopt
on
or p u r c h a s e ) , there are numerous v a r i a t i o n s o n the presumed sequence. T h i s
cor
is p a r t i c u l a r l y so on the creation/production side. F o r e x a m p l e , as noted
anc
elsewhere ( K l i n e 1 9 8 5 ; K l i n e a n d Rosenberg 1 9 8 6 ) , product development
pol
m a y often stimulate applied or basic research, rather t h a n derive f r o m it.
bu!
T h e r e are typically so m a n y feedback and f e e d f o r w a r d cycles of i n f o r m a t i o n
tak
exchange that are part of the i n n o v a t i o n process, a n d so m a n y shocks and u n -
cor
predictable setbacks and surprises (Schroeder et a l . 1 9 8 6 ) to almost defy the
inr
notion of stages and phases. It is better to t h i n k of it as a highly iterative
pai
process.
ho^
vai F l o w e v e r , the process is h a r d l y r a n d o m . Perhaps a better w a y of t h i n k i n g
trai about these issues is not to t h i n k of the sets of behaviors defined i n most
/ stage/phase models as steps o n a s t a i r w a y , but rather as rooms connected by
• ^ a finite number of doors. F o r e x a m p l e , i n the creation/production of

i technological i n n o v a t i o n w e can generally agree that something called

development occurs a n d is i m p o r t a n t , as w e l l as something called research,
and another set of activities called dissemination. T h e relative o r d e r i n g , or i m -
portance of these activities might v a r y depending u p o n other factors ( a n d be
subject to differing evaluation based on time or c r i t e r i a ) , but nonetheless, we
need to consider them all as p a r t of the larger puzzle of technological i n n o v a -
t i o n . E v e n t u a l l y , a participant w i l l probably w a n d e r through all of the rooms,
spending more or less time i n each. K l i n e ' s ( K l i n e 1 9 8 5 ; K l i n e and Rosenberg
1 9 8 6 ) concept of i n n o v a t i o n as a set of " c h a i n s " a n d connections represents
a p a r a l l e l metaphor.
U s i n g this m e t a p h o r , readers m a y consider the r e m a i n i n g chapters of this
 Technological Innovation as a Process •31

book as reflecting sets of b e h a v i o r s , both i n d i v i d u a l and s o c i a l , that define the

particular " r o o m s " of the i n n o v a t i o n process. E a c h r o o m has core behaviors
that take place w i t h i n i t ; movement between rooms is defined by marker
events that tell us w h e n w e are m a k i n g significant b e h a v i o r a l transitions f r o m
one kind of activity to another.^ F o r e x a m p l e , one set of core behaviors
might be basic research conducted a r o u n d some technical focus, perhaps in
the setdng of a lab i n a research-oriented u n i v e r s i t y . A m a r k e r event might
be the point in time w h e n the participants i n another set of core b e h a v i o r s —
engineers doing development i n a private company—see the potential for
using the academic research results to enrich a n already existing product
development activity, or to seed a n e w one. B o t h k i n d s of models w e have
described—those that emphasize the distinctiveness of the i n n o v a t i o n and
those that emphasize its i n t e g r a t i o n — c a n be defined i n terms of core behavior
and marker events.
There is sometimes a temptation to confuse the behavior w i t h its m a r k e r s ,
and to separate those m a r k e r s f r o m their intellectual a n d social context. F o r
example, one m a r k e r event i n the analysis of i n n o v a t i o n ( f r o m the user's
perspective) is a d o p t i o n , a point w h i c h divides the organization's not h a v i n g
the technology f r o m its h a v i n g i t , or at least deciding to have it. I n the real
world, this might be the point w h e r e a c o m p a n y cuts a purchase order for our
previously mentioned d r i l l press. It is easy to t h i n k that the c o m p a n y n o w has
the drill press, but does it? A rigid model tends to m a k e adoption s y n o n y m o u s
with the entire i n n o v a t i o n process—a dichotomous (adopt/nonadopt) deci-
sion. But, if the act of adopting becomes simply a m a r k e r event or " a u t h o r i t a -
dve commitment" ( L a m b r i g h t 1 9 8 0 ) w i t h i n a larger series of events that also
shape the impacts of the i n n o v a t i o n , attention then becomes focused on the
core behavior that m a k e s possible that m a r k e r event. A l l core behavior tends
to be exhibited in a defined social context (for e x a m p l e , the user organization)
and by a finite number of social role incumbents (for e x a m p l e , the decision
makers in that user o r g a n i z a t i o n ) .
Finally, it is w o r t h noting that the utility a n d validity of stage concepts
is in the eye of the beholder—the i n d i v i d u a l t r y i n g to m a k e sense out of the
overall process. T h e r e is nothing inherent i n a stage conception of i n n o v a t i o n
that implies that individuals actually i n v o l v e d must agree o n or even realize
just what stage they are going t h r o u g h . I n n o v a t i o n often occurs w i t h o u t m u c h
awareness on the p a r t of most participants of w h a t is going on outside their
own spheres of activity ( L a m b r i g h t 1 9 8 0 ; E v e l a n d , R o g e r s , a n d K l e p p e n
1977). W e all live i n our o w n w o r l d s .
T o summarize our comments about stages, phases, a n d related issues, w e
would hke to first of a l l emphasize that no single terminology is u n i v e r s a l l y
accepted for defining stages or phases of the process of technological i n n o v a -
don. While most research and analysis has focused o n either the production
or use of innovations, the evidence they have produced suggests that most
 J

32 • The Processes of Technological Innovation

TE(

T.
prim
A t \i
over
on t Figure 3 - 1 . Processes of Technological Innovation
con(
and
poli innovations do not f o l l o w any single, set ordering of events or stages. Never-
bus: theless, most innovations do appear to proceed through a relatively c o m m o n
take set of m a r k e r events, w h i c h in t u r n are embedded i n foci of activity (rooms
con of core b e h a v i o r s ) , that pretty m u c h correspond to t r a d i t i o n a l stages and
inni phases. T h e real p r o b l e m w i t h stage models is that they delude people into
par t h i n k i n g that one stage i n e x o r a b l y leads to the next (for e x a m p l e , that basic
ho\ research leads to development). Such is not necessarily the case.
var
I n figure 3 - 1 w e have attempted to capture m a n y of the concepts dis-
trai
cussed i n this chapter. T h e s e include the notion of i n n o v a t i o n as a
longitudinal set of processes, as w e l l as the distinction between developers of
technology a n d users o f technology. O n e r e m a i n i n g set o f issues concerns the
different levels of social aggregation that are i n v o l v e d i n those processes. T h i s
w i l l be the topic of the next several sections.

W h o o r W h a t I n n o v a t e s : T h e U n i t of A n a l y s i s P r o b l e m s

It should be apparent that there are multiple social units that participate in
any process of technological i n n o v a d o n . A s the process u n f o l d s , decisions are
 Technological Innovation as a Process • 33

continually being made by i n d i v i d u a l s , groups, and even organizations. T h i s

means that, in practice, it m a y be extremely difficult to identify h o w decisions
feed each other i n a linear or logical sequence ( W i t t e 1 9 7 2 ) . T h e question of
who or what innovates is one that bears close analysis as w e try to understand
the overall process.
One problem is simply deciding w h a t level of aggregation o n w h i c h to
focus. For example, w e often hear in casual conversation about this organiza-
tion being innovative or that c o m p a n y being i n the d a r k ages. I n fact, there
is a highly popular literature that develops lists and vignettes about the " b e s t "
companies ( K a n t e r 1 9 8 3 ; Peters a n d W a t e r m a n 1 9 8 2 ) . B u t only r a r e l y is an
entire firm innovative or even i n v o l v e d i n the i n n o v a t i o n process; i n n o v a t i o n
is usually carried out by s m a l l groups (Pressman a n d W i l d a v s k y 1 9 7 3 ) .
Moreover, different components of the f i r m are likely to be i n v o l v e d at dif-
O
ferent times ( M a r c h a n d R o m e l a e r 1 9 7 9 ; G i a c q u i n t a 1 9 7 8 ) . Some parts are
z
likely to be in early stages, other parts are i n later stages, some parts m a y not
be involved at a l l , and other parts w d l be actively resisting the changes. I n
fact, after implementation is complete m a n y , if not most, parts of the
organization may r e m a i n essentially untouched by the change.
Given that i n n o v a t i o n is r a r e l y a property of the entire system that w e
wish to examine, the question r e m a i n s , h o w to decide w h e r e to focus our
attendon? In the context of the entire social system, w e have at least a choice
of concentrating o n one or more h i e r a r c h i c a l levels:

individuals as i n d i v i d u a l s
individuals enacting social roles
social groups
organizations
interorganizational l i n k s
aggregates of organizations (for e x a m p l e , industries)
societies

Each of these categories of units has its o w n i n n o v a t i o n literature, and each

can tell us something about the process.
There is no uniquely " c o r r e c t " level of concentration.^ I n this b o o k , our
focus is generally on phenomena o f technological i n n o v a t i o n o c c u r r i n g w i t h i n
organizational settings. W e t h i n k this corresponds most closely to most peo-
ple's frame of reference. W e tend to t h i n k of innovative factories, or schools,
or laboratories.
In most cases, h o w e v e r , w h e n w e are t h i n k i n g of organizations w e are in
fact focusing on a subunit w i t h i n an organization (Jelinek 1 9 7 7 ) . B e y o n d a
certain m i n i m a l size of o r g a n i z a t i o n , the technologies being used (or
developed) w i l l v a r y w i d e l y across different organizational subunits.
Attention to the o r g a n i z a t i o n cannot be divorced f r o m attention to the
 34 • The Processes of Technological Innovation

i n d i v i d u a l s a n d groups that m a k e it u p ; nor c a n it ignore the implications of

o r g a n i z a t i o n a l i n n o v a t i o n for the larger system w i t h i n w h i c h it is embedded.
A s a result, w h i l e our p r i m a r y focus is i n t r a o r g a n i z a t i o n a l , w e also bring to
bear the influences of larger a n d smaller social aggregations, such as industries
and i n d i v i d u a l s .
1 I n the remainder of this chapter, w e discuss some o f the insights into inno-
v a t i o n processes that have been f o r m e d by research concentrated at one or
another of these levels of aggregation. T h e purpose of this e x c u r s i o n is illus-
t r a t i v e , rather t h a n definitive. T h e case for a multilevel f r a m e of a n a l y t i c refer-
ences is derived f r o m understanding t r a d i t i o n a l w a y s o f approaching inno-
vation.
E a c h level has tended to f o r m its o w n v o c a b u l a r y , often dominated by a
single discipline. H o w e v e r , each perspective has something to contribute to
our later understanding of the more d y n a m i c aspects of the process. W e w i l l
conclude w i t h some observations on interrelating levels of a n a l y s i s , an act that
P is considerably more problematical t h a n its practitioners generally a d m i t .
A
o
Characteristics of Individuals
o
C( O n e w a y of l o o k i n g at i n n o v a t i o n is to concentrate o n personal characteristics
ai of the people i n v o l v e d , regardless of their social role, group m e m b e r s h i p , and
P the l i k e . T h e p r e v a i l i n g assumption of this approach is that i f w e k n o w the
b- people, then all w i l l become clear. F o r e x a m p l e , researchers have observed
ta m a n y consistencies i n demographic characteristics, psychological traits, and
C( job skills of those i n d i v i d u a l s w h o are heavily i n v o l v e d in i n n o v a t i o n
ir activities.
P' T h e visibility of i n d i v i d u a l s i s , of course, likely to be greater i n relation
h( to some m a r k e r events (such as initiation or invention) t h a n others. F o r e x a m -
v< ple, technical gatekeepers (those w h o f o r m a l l y or i n f o r m a l l y broker the input
tr of n e w technical i n f o r m a t i o n to companies) are usually i n their t h i r t y ' s , hold
at least a master's degree, a n d have h a d some R & D w o r k experience w i t h
emphasis o n development research ( R o b e r t s 1 9 6 9 ; A l l e n 1 9 7 7 ) . U d e l l , B a k e r ,
and A l b a u m ( 1 9 7 6 ) compared inventors a n d n o n i n v e n t o r s , using a personal-
ity checklist, and concluded that inventors w e r e more emotionally adventur-
ous, achievement-oriented, independent, resourceful, creative, goal-centered,
and h a r d - w o r k i n g t h a n n o n i n v e n t o r s . A variety of personal skills have been
attributed to knowledge about the needs of the f i r m , technical competence,
political astuteness ( C h a k r a b a r t i 1 9 7 4 ) , leadership, a n d great diversity i n job
activities.
It is unclear i n m u c h of this research literature w h e t h e r these good behav-
iors are manifestations of people's personality, or of situationally specific and
role-related behaviors. I n fact, there has been considerable disagreement in
psychology about the extent to w h i c h behavior is the result of personality or
of situation-specific events (for e x a m p l e , M i s c h e l 1 9 7 3 ) .
 Technological Innovation as a Process • 35

Within the f r a m e w o r k of i n d i v i d u a l personal characteristics a n d i n n o v a -

tion, the w o r k of E v e r e t t Rogers ( 1 9 8 3 ) has been i n f l u e n t i a l . Rogers r e v i e w e d
hundreds of studies and used s u m m a t i v e techniques to identify the personality
and demographic characteristics most associated w i t h innovativeness as
expressed in three c r u c i a l activities i n the process: i n n o v a t i o n a d o p t i o n , o p i n -
ion leadership (interpersonally i n f l u e n t i a l ) , a n d change agentry.
T h i r t y - t w o generalizations associated early adoption of i n n o v a d o n w i t h
various demographic, personality, a n d other characteristics of i n d i v i d u a l s .
Eady adopters tended to h a v e , a m o n g other t r a i t s , higher social status, more
favorable attitudes t o w a r d credit, change, r i s k , education, a n d sciences,
greater intelligence, more social p a r t i c i p a t i o n , more change agent contact,
more exposure to mass media a n d interpersonal c o m m u n i c a t i o n channels,
more cosmopolitan b a c k g r o u n d s , more highly integrated l i n k s w i t h the social
system, and more i n f o r m a t i o n a n d knowledge about i n n o v a t i o n s .
For successful o p i n i o n leadership a n d change agentry, the i n d i v i d u a l char-
acteristics were s i m i l a r . F o l l o w e r s seek o p i n i o n leaders w i t h higher social
status, more e d u c a d o n , greater mass media e x p o s u r e , and more change agent
contact, and w h o are more c o s m o p o l i t a n a n d i n n o v a t i v e . C h a n g e agent suc-
cess was positively related to i n d i v i d u a l characteristics such as social status,
educadon, and c o s m o p o l i t a n i s m .
The important point is that i n d i v i d u a l characteristics do indeed seem to
make some difference. W h e t h e r this is because they genuinely cause these
behaviors, or because i n d i v i d u a l s i n certain k i n d s of positions a n d roles tend
to turn into certain k i n d s of people, is not w e l l established.
Nevertheless, given the a m o u n t of effort that has gone into cataloging the
individual characteristics of i n n o v a t o r s , there are at least t w o important
implications that can be d r a w n for our purposes. O n e is that in any given set-
ting in which innovation-related activities o c c u r , the personal attributes of
participants may be equally or m o r e i m p o r t a n t than group or organizational
factors. T h a t i s , no matter w h a t pains are t a k e n to provide the right decision-
making processes or r e w a r d systems, and no matter h o w pressing a n d persua-
sive the external economic e n v i r o n m e n t i s , if rigid a n d t i m i d people are
employed in jobs that are key to fostering a n i n n o v a t i o n process, it w i l l likely
fail.
Unfortunately, the converse is also true. T h e best and the brightest people
do not guarantee success. I n m a n y (or most) cases, the promise of desirable
individual characteristics of p a r d c i p a n t s m a y be smothered by o r g a n i z a d o n a l
or environmental context.
A second important point is that there are limits to the utility of i n d i v i d u a l
characteristics i n f o r m a d o n , i n terms of m a k i n g a difference i n i n n o v a d n g sys-
tems. Individual characteristics that are truly i n d i v i d u a l characteristics ( s e x ,
personality) tend not to be easy to change. I f w e are t r y i n g to using o u r k n o w l -
edge about the i n n o v a t i o n process to influence the course of i n n o v a t i o n activi-
ties, knowing the personal characteristics of participants is a p o o r lever. T h e
 36 • The Processes of Technological Innovation

only practical w a y to use this i n f o r m a t i o n is through the selection of p a r t i c i -

pants. T h a t i s , if w e have identified a number of p e r s o n a l , demographic, or
s k i l l characteristics that are predictive of i n n o v a t i o n success, then it should be
a relatively easy task to develop assessment/selection tests to p i c k appropriate
participants. U n f o r t u n a t e l y (or fortunately, depending o n one's p o l i t i c a l point
of v i e w ) w e rarely have a clean slate situation. W e usually have an existing
w o r k f o r c e to start w i t h and rarely have the freedom to select at w i l l . W h e n
w e d o , a n d w h e n w e c a n exercise the selection option (Schmidt et a l . 1 9 7 9 ) ,
it often w o r k s exceedingly w e l l . T h u s it m a y be of value to those p l a n n i n g
change activities.

Individuals as Role Holders

T h e next step beyond l o o k i n g at individuals as isolated units is to l o o k at indi-

viduals as they p e r f o r m roles i n a social system. A role is a pattern of expected
behaviors, attitudes, a n d decision criteria c a r r i e d out by a person (or group)
w i t h i n a specific situation ( K a t z a n d K a h n 1 9 7 8 ) . It is something that a person
does, not something that he or she is. R o l e s tend to be defined e x p l i c i t l y (for
e x a m p l e , job descriptions) or i m p l i c i t l y (by expectations of the peer group) in
a specific organizational context; i n d i v i d u a l characteristics are tendencies we
carry f r o m one context to another. T h i s m a y seem like a s m a l l d i s t i n c t i o n , but
it is c r u c i a l . T o say that someone enacts a p a r t i c u l a r role i n a p a r t i c u l a r situa-
tion or context is to emphasize that i n a different situation he or she m a y take
an entirely different role ( G o f f m a n 1 9 5 9 ) . W e all k n o w that there are perfect
managers w h o abuse their spouses, and l o v i n g spouses w h o abuse their staff
at w o r k . W e have the capability to be almost all things depending o n where
w e are.
M o s t roles have associated c r i t e r i a , that i s , w h e n w e enact a particular
r o l e , w e are expected to use certain criteria i n m a k i n g decisions w i t h i n that
role. A person enacting the role of father is expected to use the criterion child's
well-being w h e n m a k i n g decisions about the c h i l d . A manager is expected to
use the criterion c o n t r i b u t i o n of m y group to the overall o r g a n i z a t i o n . W h e n
people do not use the criteria associated w i t h a p a r t i c u l a r r o l e , great confusion
results.
R o l e s v a r y w i d e l y i n specificity a n d the degree of constraint placed upon
those w h o occupy t h e m . C e r t a i n k i n d s of r o l e s — p a r t i c u l a r l y those associated
w i t h organizational positions—have relatively permanent or institutionalized
criteria-sets associated w i t h t h e m . O t h e r roles, p a r t i c u l a r l y those associated
w i t h f a m i l y and friendship relationships, have m u c h m o r e widely v a r i e d and
diffuse c r i t e r i a . G e n e r i c roles, w h i c h exist a n a l y t i c a l l y to describe overall rela-
tionships between people in broad classes of situations ( b u y e r , manager, cid-
z e n , e t c . ) , place far fewer constraints o n p a r t i c u l a r behavior than more spe-
cific roles such as stock clerk at Payless D r u g s or manager of software systems
 Technological Innovation as a Process • 37

for Fulcourte Press. I n general, roles tied to p a r t i c u l a r positions w i t h i n organi-

zational structures tend to be relatively c i r c u m s c r i b e d — a l t h o u g h there is
always room for v a r i a t i o n in even the most constrained circumstances.
Some roles (as expected sets of b e h a v i o r ) only exist because other people
informally agree that they s h o u l d . T h e s e are roles that have no o f f i c i a l sanc-
don of legidmacy i n an o r g a n i z a d o n a l setting. T h e r e are no job descriptions
written d o w n , no explicit or overt criteria for role p e r f o r m a n c e . I n fact, m a n y
of the important roles i n the i n n o v a t i o n process are of this type.
All systems that are i n v o l v e d i n i n n o v a t i o n — e i t h e r developing things or
struggling to learn h o w to use t h e m — a r e made up of individuals enacting dif-
ferent but usually complementary roles. I n effect, there is a division of labor
(Chakrabarti and H a u s c h i l d t 1 9 8 9 ) i n the process. T h e s u m of h o w i n d i v i d -
uals function in organizations c a n be understood as a c o m b i n a t i o n of i n d i v i d -
ual characteristics or as role features. F o r e x a m p l e , one might define a n entre-
preneur as a specific type of i n d i v i d u a l , or look at entrepreneurial behavior
as a repertoire of activities c a r r i e d out by individuals occupying such posi-
tions. T h e concept o f social roles p e r m i t s — i n fact, r e q u i r e s — m u l d p l e levels
of analysis to be e x a m i n e d simultaneously. A s s u c h , it is m u c h more congru-
ent with a process model of i n n o v a d o n , a n d it a l l o w s the internal w o r k i n g s
of the social system to be studied directly, rather t h a n as atomistic units w i t h i n
systems. T o continue o u r e x a m p l e , it is m u c h easier to understand the d y n a m -
ics of entrepreneurial behavior at a n i n d i v i d u a l level, if one also understands
the economic and technological opportunities i n the environment (see chapter
5 for a more complete discussion).
Actors in the i n n o v a t i o n process p e r f o r m vital roles that are both f o r m a l
and informal. F o r m a l roles, such as those p r o v i d i n g o f f i c i a l a u t h o r i z a t i o n (for
example, the president of a c o m p a n y m a y have to approve all purchases over
a certain amount), are usually f a i r l y clear-cut. H o w e v e r , i n f o r m a l roles are
often not well understood. F o r e x a m p l e , several studies ( A l l e n 1 9 7 7 ; K e l l e r
and Holland 1 9 7 8 ; a n d C h a k r a b a r t i 1 9 7 4 ) have observed that a s m a l l number
of people in organizations are relied u p o n by others to serve as i m p o r t a n t
sources of technical i n f o r m a t i o n . T h e s e gatekeepers have the ability to absorb
complex technical i n f o r m a t i o n a n d to translate it into a more understandable
form for coworkers and top management. People w h o p e r f o r m this role have
a particularly important f u n c t i o n i n b u i l d i n g awareness of n e w products and
processes during the early stages of i n n o v a t i o n , although only r a r e l y w i l l one
see this role defined i n a f o r m a l job description.
Another important i n f o r m a l organizational role is that of the product
champion w h o l i n k s the different phases of the i n n o v a t i o n decision m a k i n g
process ( C h a k r a b a r t i 1 9 7 4 ) . T h i s role is rarely o f f i c i a l (thus i n f o r m a l ) and
involves a variety of activities such as arguing for the changes, nudging people
along i j do the things that need to be done to bring about the change, finding
needed resources, a n d so o n . I n one study ( T o r n a t z k y et a l . 1 9 8 0 ) , t w o
 38 • The Processes of Technological Innovation

administrators i n a state mental health office functioned as bureaucratic entre-

preneurs in implementing innovations i n hospitals i n their state. I n this role
they " w o r k e d the s y s t e m " i n n o v e l and creative w a y s — o b t a i n i n g resources,
shortcutting procedures, stretching the boundaries of job descriptions. Similar
findings have been reported o f innovations i n u r b a n mass transit (Rogers,
M a g i l l , a n d R i c e 1 9 7 9 ) as w e l l as i n m a n u f a c t u r i n g p r o d u c t i o n systems ( G e r -
w i n 1 9 8 2 ) , and i n industrial organizations ( K a n t e r 1 9 8 3 ) . A s an interesting
footnote, not all the important roles are on the side o f change. Some authors
have pointed out the importance of the role of defender against i n n o v a t i o n
( K l e i n 1 9 7 7 ) , the inverse of the product c h a m p i o n .
T h e r e are interesting a n d i m p o r t a n t interactions between the personal
characteristics of i n d i v i d u a l s w h o occupy social roles a n d the expectations
relating to these roles. F o r e x a m p l e , one study ( U d e l l , B a k e r , and A l b a u m
1 9 7 6 ) described problems facing independent i n v e n t o r s , problems that stem
f r o m mismatches between requirements of the inventor role and the character-
istics of the people w h o find themselves i n it (often, technically talented engi-
neers). F r e q u e n t l y , preferring to w o r k alone, the inventor is ignorant of the
needs of potential user f i r m s . A s a result, the inventor has difficulty locating
an o r g a n i z a t i o n that is interested i n his or her idea ( v o n H i p p e l 1 9 7 8 ) and
rarely is equipped to receive a n d process evaluative feedback concerning the
technical feasibility of ideas, i n f o r m a t i o n concerning appropriate modifica-
tions, and other data that could increase the probability of acceptance.
C h a k r a b a r t i ( 1 9 7 4 ) has described several pervasive problems faced by the
product c h a m p i o n w i t h i n a n o r g a n i z a t i o n . W h i l e , by personal characteristics,
the c h a m p i o n is likely to be a highly c o m m u n i c a t i v e i n d i v i d u a l , the product
champion's advocacy role m a y thrust h i m or her into situations i n w h i c h com-
m u n i c a t i o n input is severely c u r t a i l e d . F o r e x a m p l e , he or she is often seen
as a n outcast by other members of the organization a n d as an advocate of
ideas that seem unrealistic. W h i l e some organizations have n o r m s that toler-
ate or even r e w a r d such behaviors, it is not the usual case. O t h e r studies high-
light the inherent conflicts between certain roles. F o r e x a m p l e , B e a n and
Mogee ( 1 9 7 6 ) discuss clashes between p u r c h a s i n g agents and engineers over
cost a n d performance considerations i n decisions about n e w technology, and
R o b e r t s ( 1 9 6 9 ) describes the different roles of research lab personnel and
y o u n g entrepreneurs.
T h e s e scattered studies a n d observations n o t w i t h s t a n d i n g , far more
research needs to be done to understand h o w i n d i v i d u a l role expectations and
i n d i v i d u a l characterisdcs (personality, demographics, values) intersect with
one another. F o r e x a m p l e , h o w are personal f a c t o r s — s u c h as education, cos-
m o p o l i t a n i s m , and age—related i n a f u n c t i o n a l w a y to such i n n o v a t i o n activi-
ties as roles? D u c h e s n e a u and D u t t o n ( 1 9 7 7 ) suggest that m a n y i n d i v i d u a l dif-
ferences m a y actually operate as p r o x y v a r i a b l e s , that i s , measurable quand-
ties that represent aspects of role behavior. F o r e x a m p l e , one w r i t e r ( M a n s -
 Technological Innovation as a Process • 39

field 1971) has suggested that the education o f the president of a f i r m m a y

be important only because a highly educated president is better able to under-
stand the implications of i n n o v a t i o n s , to be more flexible intellectually, and
to have more extensive outside c o n t a c t s — i n effect, being more capable of
enacdng those portions of the presidential role that have implications for
innovadon. I n a s i m i l a r set of findings f r o m the area of mental health i n n o v a -
tion (Fairweather, Sanders a n d T o r n a t z k y 1 9 7 4 ; T o r n a t z k y et a l . 1 9 8 0 ) ,
some personal and demographic characteristics such as age and number of job
moves, which were f o u n d to be related to i n n o v a t i o n , m a y likely have consti-
tuted proxy variables for values about r i s k t a k i n g , a n d related roles.
Focusing on roles also does not get us out of selection issues. F o r e x a m p l e ,
it is one thing to be able to say after the fact that a product c h a m p i o n role
was critical in a technology development project i n one's o r g a n i z a t i o n . It is
quite another to replicate the experience i n another project. D o w e simply cre-
ate a formal role of c h a m p i o n , a n d give it the appropriate powers a n d respon-
sibilities, or do w e also try to identify and select a product c h a m p i o n i n d i v i d -
ual, and when w e d o , w h a t are the criteria and selection processes?
There is one additional p r o b l e m i n using social roles—to the e x c l u s i o n of
other analytic f r a m e w o r k s — t o understand technological i n n o v a t i o n . T h a t i s ,
there is a tendency to reify and concretize social roles, once w e describe them
on paper. I n real life, a social role is often a temporary p h e n o m e n o n , that only
has reality in a certain context a n d time f r a m e . I n n o v a t i o n processes are
dynamic, and as w e move f r o m r o o m to r o o m , the i m p o r t a n t roles move o n
and off the stage. F o r e x a m p l e , i n one project focused o n mental health treat-
ment innovations ( T o r n a t z k y et a l . 1 9 8 0 ) , a n attempt w a s made to enlist a n d
train individuals w h o h a d been successful implementers of a n i n n o v a t i o n i n
their own hospitals to become traveling change agents, to w o r k w i t h other
hospitals across the country w h o were considering adoption. T h e experiment
was a dismal failure; none of the i n n o v a t o r s w a n t e d to leave their home base.
The role of being a l o c a l i n n o v a t o r and hero w a s comfortable; that of being
an advocate in foreign territory w a s not.

Groups

Moving beyond i n d i v i d u a l social roles, the next levels of social aggregation

tend to be the group, and then the o r g a n i z a t i o n . G e n e r a l l y , the ad hoc social
group has not been carefully studied i n the context of i n n o v a t i o n processes.
While there is a fairly rich literature ( S h a w 1 9 7 6 ) on group p r o b l e m solving
(what factors contribute to quicker or more creative s o l u t i o n s ) , that w o r k has
tended to be divorced f r o m issues of technological i n n o v a t i o n per se. Social
groups tend to be i n f o r m a l , not legitimized (recognized, o f f i c i a l ) . T o the
extent that the m a r k e r events o f i n n o v a d o n tend to be visible a n d official (for
example, adoption decision), there has been less concern w i t h groups as
 40 • The Processes of Technological Innovation

opposed to organizational d y n a m i c s . Some exceptions include the classic

studies of i n f o r m a l w o r k groups and the adoption (or rejection) of various
p r o d u c t i o n process innovations a n d approaches ( C o c h and F r e n c h 1 9 4 8 ;
Roethlisberger and D i c k s o n 1 9 3 9 ) .
T h e r e are other exceptions. C o n s i d e r a t i o n of s m a l l group p h e n o m e n a ,
such as w o r k groups, has been important i n understanding the conduct of
basic and applied research ( C o h e n et a l . 1 9 8 5 ; A l l e n 1 9 7 0 ) . E f f e c t i v e team
interaction processes seem to be important predictors of successful outcomes
(see chapter 4 ) . B y the same t o k e n , there has been some interest i n s m a l l group
phenomena i n the context of implementation b e h a v i o r . F o r e x a m p l e , one
project ( T o r n a t z k y et a l . 1 9 8 0 ) intervened in the group d y n a m i c s of imple-
mentation teams in a number of mental hospital settings and observed effects
on implementation success. A t the same time that a w o r k group w a s receiving
technical assistance and t r a i n i n g on a treatment i n n o v a t i o n , they also received
( i n the e x p e r i m e n t a l condition) consultations on organizational structure and
c o m m u n i c a t i o n patterns of their group. T h i s significantly accelerated use and
implementation of n e w technologies.
Some of the most focused w o r k on groups and their relationship to tech-
nological i n n o v a t i o n h a s , not s u r p r i s i n g l y , been conducted i n the sociotechni-
c a l r e a l m . A t the same time that technology is being designed a n d imple-
mented, a p a r a l l e l a n d , it is hoped, integrated process o f w o r k group design
is being undertaken ( w e discuss this i n chapter 9 ) . M u c h of the w o r k i n the
sociotechnical t r a d i t i o n has been focused on process ( p r o d u c d o n ) technolo-
gies being used by some organizational subunit ( a n engine l i n e , a coal-mining
crew).
O t h e r group-oriented research has focused o n a p a r t i c u l a r group process,
or o n h o w group processes facilitate larger systems of technological change.
F o r e x a m p l e , recent w o r k i n the area of computer-supported cooperative
w o r k has concentrated o n the properties o f i n f o r m a t i o n technology that
enhance the abdity of w o r k groups to interact effectively, recognizing that
i n f o r m a t i o n w o r k is inherently a group effort ( G r e i f and S a r i n 1 9 8 6 ) . There
has been a g r o w i n g interest (albeit little research) o n the use of temporary sys-
tems (generally ad h o c , i n f o r m a l teams) being used to facilitate the deploy-
ment of c o m p l e x p r o d u c t i o n systems. T h e term temporary systems a c k n o w l -
edges that m a j o r implementation effects are one-time or infrequent events.
Despite this interest, proportionately more attention ( a n d research) has been
devoted over the years to defining innovative organizations.

Organizations

So f a r , w e have not encountered a great deal of d i s c i p l i n a r y divergence.

Studies of individuals and small groups tend to be generally dominated by
v a r i o u s types of psychologists, a n d the language of such studies tends to be
 Technological Innovation as a Process • 41

within the same o v e r a l l type of discourse. H o w e v e r , w h e n w e move to studies

of organizations and aggregates of organizations, some significant differences
in focus begin to appear. M o s t p a r t i c u l a r l y , differences c a n be detected
between the sociological/social psychological perspective a n d the economic
perspective. E c o n o m i c studies generally l o o k at aggregate behavior of units
over time; social/organizational studies, by contrast, tend to l o o k for struc-
tural features w i t h i n systems at p a r t i c u l a r points i n time. E c o n o m i c s generally
imputes rationality and relatively consistent motives to organizational actors;
social/organizational studies, based on role m u l t i p l i c i t y , generally argue for
a much less r a d o n a l picture of organizational life.
Ultimately, both perspectives ( a n d others that might be considered) are
talking about the same b e h a v i o r , by the same people, i n the same places. T h e
point is not that one perspective is correct, it is simply that the use of different
analytic vocabularies a n d sets of guiding assumptions about people and
behavior can lead to v e r y different pictures of the same thing. T h e analyst
selects the perspecdve that appears most useful i n terms of the questions he
or she is raising at any given p o i n t . It is c r u c i a l for the reader to understand
how that perspective affects the a n a l y s i s . I n the remainder of this b o o k , w e
will employ a w i d e range of disciplinary perspectives, doing our best to ensure
an adequate cross-disciplinary understanding.
One of the first sociologically focused studies of i n n o v a t i o n w a s that of
Burns and Stalker ( 1 9 6 1 ) . I n a multisite c o m p a r i s o n of i n d u s t r i a l organiza-
tions, they found that those that w e r e relatively nonbureaucratic ("organic")
in structure were more amenable to technological i n n o v a t i o n t h a n those that
were more bureaucratic ("mechanistic") i n structure. U n f o r t u n a t e l y , w h a t
constituted technological i n n o v a t i o n tended to be f a i r l y f l e x i b l e , including
product and process technologies, incremental and r a d i c a l changes, and pro-
cesses of creating as w e l l as adopting/implementing technologies. M o r e o v e r ,
the specific divergent properties of bureaucratic or n o n b u r e a u c r a t i c organiza-
tions that contributed most to i n n o v a t i o n w e r e not w e l l identified by the
research. T h i s study illustrates a w e a k n e s s i n m u c h o f the early w o r k — a
single variable ( i n this case, organic/mechanistic structure) w a s presumed to
describe an entire organization a n d the m a n n e r i n w h i c h it carried out a c o m -
plex process. T h a t early w o r k also failed to specify the functional relationship
between the organization a n d some p a r t i c u l a r focus of innovative a c t i v i t y —
was the whole organization the user? the locus of invention? both? or an inno-
vation itself? ( K i m b e r l y 1 9 8 6 ) .
Following B u r n s a n d S t a l k e r , there have been a variety of organizational
structure and i n n o v a t i o n studies at s o m e w h a t more detailed levels of analysis.
These studies have usually focused on three d o m a i n s : c o m p l e x i t y , f o r m a l i z a -
tion, and centralization.'' W o r k i n g at the o r g a n i z a t i o n a l level, such studies
tend to employ relatively aggregated measurements. B o t h the properties of the
organizational system a n d the i n n o v a t i o n behavior i n question tend to be rela-
 4 2 • The Processes of Technological Innovation

tively abstracted, both in d o m a i n (over lots of individuals) and in time (over

more or less long i n t e r v a l s ) . T h u s , the observations w e summarize here should
i not be taken as telling us m u c h about i n n o v a t i o n processes, but rather as
J reflecting s u m m a r y judgments l o o k i n g back at defined points i n time and
space. W e w d l r e t u r n later to this question of h o w to deal w i t h aggregation
1 issues, but w i l l note that aggregation of events over time m a y do as m u c h vio-
lence to understanding those events as does aggregating over levels of analysis.
— Nonetheless, let us consider some of the results. O r g a n i z a t i o n a l complex-
ity has been positively l i n k e d to i n n o v a t i o n . U n f o r t u n a t e l y , c o m p l e x i t y has
' been measured i n a number of nonequivalent w a y s , most frequently using
~ either the degree of professionalization (number of professional groups i n the
— organization) or the diversity of specialists as the p r i n c i p a l index ( A i k e n and
H a g e 1 9 6 8 ; H a g e 1 9 8 0 ; H e y d e b r a n d 1 9 7 3 ; D u c h e s n e a u , C o h n , a n d Dutton
1-1 1 9 7 9 ) . A l t h o u g h a relationship exists, the discrete processes i n v o l v e d are
again left unclear, a n d the level at w h i c h the effects are felt has generally been
pj less than adequately specified. T h i s is just one e x a m p l e .
A I n s u m m a r y , issues of organizational structure a n d process do m a k e a dif-
o\. H o w e v e r , the nature of the relationship between organizational fac-
Oi tors and i n n o v a t i o n processes seems to depend highly o n w h i c h particular
cc processes w e are t a l k i n g about a n d other c o n t e x t u a l constraints. I n subse-
ar quent chapters w e w i l l revisit the question of h o w organizational structure
p( affects i n n o v a t i o n i n more detail,
bi

Innovation from the Perspective of the

^° Economic Environment
in
pa A higher, or more aggregated, level of analysis is that of the environment in
he w h i c h the f i r m or organization operates. W h i l e social/organizational analysts
va have been struggling to come to terms w i t h i n n o v a t i o n f r o m their perspective,
tn economists have l i k e w i s e been t r y i n g to define h o w technological change fits
into their accustomed models of f i r m behavior. T h a t i s , economists tend to
assume some hyper-rationality o n the part of firms ( a n d other economic
u n i t s ) , as they attempt to m a x i m i z e advantage v i a v a r i o u s input a n d output
options. F o r e x a m p l e , Rosenberg ( 1 9 7 6 ) and N e l s o n a n d W i n t e r ( 1 9 7 7 ) both
suggest that f i r m s w i l l continue to produce a given product until factor prices
change (induced by changing e n v i r o n m e n t a l conditions) i n such a w a y as to
reduce profit significantly. F i r m s w i l l then actively begin to search for ways
of saving inputs or i m p r o v i n g output q u a l i t y , a process that is riddled with
uncertainty about future input prices, about the outcome of the search pro-
cess, and about the m a r k e t a b i l i t y of the resulting technical products.
T h e fundamental question for economists is whether traditional equilib-
rium-oriented models should be discarded for more evolutionary models of
 Technological Innovation as a Process • 43

technological i n n o v a t i o n that e x p l i c i t l y consider r i s k , uncertainty, and f i r m

strategy. T h i s issue is discussed i n greater detail i n chapter 7.
Economic analysis has also been p a r t i c u l a r l y attentive to the relative per-
formance of firms w i t h i n a m a r k e t . T h a t i s , w h a t is the distribution of firms
of various sizes, by i n d u s t r y , a n d does it m a k e a difference o n the pace of
technological i n n o v a t i o n w i t h i n i n d i v i d u a l f i r m s and/or across a w h o l e i n -
dustry? For e x a m p l e , is it better ( i n n o v a t i v e l y speaking) to have lots of small
firms, or a few big ones, or w h a t ?
T w o conflicting hypotheses have been typically represented in size-
distribudon research. O n e hypothesis is that a n increasing concentration of
firms enables those f i r m s to restrict entry by others and reduces the incentives
of the dominant firms to undertake R & D a n d i n n o v a t i o n . A contrasting argu-
ment is that some c r i t i c a l mass of f i r m resources is necessary both for efficient
operadons and effective innovative activities such as R & D , since certain
development projects m a y require large outlays that only large firms can
afford.
Neither of these hypotheses has been strongly c o n f i r m e d through em-
pidcal research, although the latter appears to characterize some k i n d s of
development in some industries, such as petroleum r e f i n i n g , steel m a k i n g
(Gold, Pierce, and Rosegger 1 9 7 0 ; B o y l a n 1 9 7 7 ) , a n d some chemical process
areas (Kamien a n d S c h w a r t z 1 9 7 5 ) . T h e incentives are greatest w h e r e the
innovadon can be applied to products or processes c o m m a n d i n g large m a r k e t
shares. Such m a y be the case w i t h incremental product improvements and
process technology ( A b e r n a t h y a n d U t t e r b a c k 1 9 7 8 ) .
T w o points are w o r t h remembering as w e continue discussion of these
issues in later chapters: F i r s t , the larger economic and political environment
in which the f i r m operates is an important determinant of i n t r a f i r m behavior.
Second, often the effects of those larger e n v i r o n m e n t a l variables are highly
dependent upon the idiosyncracies of specific industries or technical areas
(Garud and V a n de V e n 1 9 8 9 ; L u c a s 1 9 8 6 ; a n d M i l l e r 1987).

Interorganizational and Environmental Interactions

Depending upon w h i c h aspects of the i n n o v a t i o n process one is i n v o l v e d w i t h ,

different social units are likely to be the focus of our concern at any given time
(for example, the research team d u r i n g early development, o r , later o n , the
implementation task force in the user o r g a n i z a t i o n . G i v e n that different
groups or social units are a l w a y s phasing i n or out of the total process, the
problem of interorganizational (or intergroup) relations is critical and
chronic. These units are t y p i c a l l y i n v o l v e d i n "trade r e l a t i o n s " w i t h each
other, often about matters h a v i n g little or nothing to do w i t h i n n o v a t i o n as
such. Moreover, these exchange relationships take place w i t h i n m a n y dif-
 44 • The Processes of Technological Innovation

ferent c o n t e x t s — s o c i a l , economic, p o l i t i c a l , t e c h n i c a l — a n d , as at the

organizational level, no single a n a l y t i c a l v o c a b u l a r y w i l l suffice to describe
them.
U n d e r s t a n d i n g o f interorganizational (or i n t r a o r g a n i z a d o n a l ) behavioral
transactions has been l i m i t e d , p a r t i c u l a r l y i n the context of studying i n n o v a -
t i o n . T h e r e are m a n y elegant concepts such as " d y n a m i c loosely coupled
s y s t e m s , " " b o u n d a r y s p a n n i n g , " " e q u i l i b r i u m , " a n d " i n t e r o r g a n i z a t i o n a l net-
w o r k s , " but they tend to be i l l defined a n d p o o r l y grounded i n everyday prac-
tice. T h e y are generally aimed at expressing the d y n a m i c s of exchanges. T h e
exchange relationship is seen as " v o l u n t a r y activity between t w o or more
organizations w h i c h has consequences, actual or anticipated for the realiza-
tion of the respective goals and objectives" ( L e v i n e and W h i t e 1 9 6 1 ) . I f one
organization needs certain resources held by another, it w i l l try to enter into
an exchange relationship w i t h the resource-holding o r g a n i z a t i o n . T h i s con-
cept f o r m s the basis of the c o m m o n resource dependence v i e w of interorgani-
zational relations. Since organizadons are not self-sustaining endties, they are
forced into relationships w i t h other organizations to obtain resources includ-
ing m o n e y , s k i l l s , a n d access to markets ( A i k e n and H a g e 1 9 6 8 ) .
U n f o r t u n a t e l y , the weaknesses and limitations of this w o r k i n relation to
technology a n d i n n o v a t i o n are apparent w h e n one l o o k s at some examples.
Studies of the exchange of innovative i n f o r m a t i o n have typically dealt w i t h
the analysis of n e t w o r k s of individuals w i t h i n a scientific specialty (Allen
1 9 7 0 ; G r i f f i t h and M u l l i n s 1 9 7 2 ) and overlooked interorganizational transac-
tion mechanisms. I m p o r t a n t relations between f u n c t i o n a l units (or suborga-
nizations) w i t h i n a larger organization (such as between R & D and marketing)
have not been w i d e l y studied ( a n exception is Souder 1 9 7 7 ) . Interorganiza-
tional studies have not generally dealt explicitly w i t h i n n o v a t i o n , but rather
w i t h h o w best to achieve " c o o r d i n a t i o n " among fragmented organizations
such as health or social service agencies ( Z e i t z 1 9 7 4 ) .
C e r t a i n l y c o o r d i n a t i o n a n d c o m m u n i c a t i o n do matter. T h e early
technical success of N A S A ' s space p r o g r a m , for e x a m p l e , has been attributed
as m u c h to its authority and ability to manage its o w n laboratories and the
thousands of private sector contractors w o r k i n g for it as to its internal
technical vision per se ( D o c t o r s 1 9 6 9 ) . C o r r e s p o n d i n g l y , the more recent
tragic failures of the same organization have been attributed largely to its cur-
rent inability to manage the interorganizational relations of that same net-
w o r k of suppliers.
T h e interorganizational context p r o v i d e d by economic analysis is perhaps
more appropriately described as a n environmental perspective, i n that atten-
tion is generally focused on h o w large-scale influences—market forces and
social policies—affect either f i r m decisions or m a r k e t structures. Included in
this a r r a y are t a x policies, government regulatory practices (including stan-
dards activities), patent l a w and policies, the general issue of intellectual prop-
 Technological Innovation as a Process • 45

erty, and various government policies that affect i n n o v a t i o n indirectly by their

influence on various inputs to the process (for e x a m p l e , government policies
and spending on education can have serious implications for the supply of per-
sonnel necessary for technical activities).
In addidon to the effects o f b r o a d b r u s h policies, government m a y also
mount specific programs to actively intervene i n some part of the i n n o v a t i o n
process. Such programs tend to have little effect o n the nature of the overall
innovadon process, but rather are intended to intensify or accelerate the w a y
in which particular subprocesses play themselves out. F o r e x a m p l e , the
Federal Small Business I n n o v a t i o n Research P r o g r a m is designed to enhance
the role played by s m a l l entrepreneurial f i r m s i n the creation of n e w product
technology by i m p r o v i n g their access to research and development funds.
Whether such targeted programs have their intended effects is open to some
dispute (Eveland 1 9 8 6 ) .
The interorganizational a n d e n v i r o n m e n t a l level has been s o m e w h a t less
attended to in the analysis of technological change t h a n might be desired. P a r t
of the difficulty, in a l l contexts, has been i n finding the right k i n d of things
to look at. E v e n more than groups a n d o r g a n i z a d o n s , interorganizational
systems are abstractions, existing only to the degree that people t h i n k they d o .
What one needs, therefore, is not so m u c h a v o c a b u l a r y to describe the larger
system itself, as a w a y to characterize the different w a y s that different people
see subsystems—and to reflect h o w that diversity of perception affects deci-
sion making and criteria at other levels.
In the previous several pages w e have described m a n y different points of
view and levels of analysis of w h a t , to some readers, m a y once have seemed
to be a fairly straightforward p h e n o m e n o n . W e have also p r o v i d e d a few illus-
trations of these perspectives. H o w e v e r , m a n y readers, at this p o i n t , m a y feel
a sense of incompleteness or fragmentation. T h i s is appropriate. It should
become clear over the course of this b o o k that no single level or unit of
analysis is sufficient to comprehend the totality of technological i n n o v a t i o n ,
and that often various processes are o c c u r r i n g simultaneously at m a n y dif-
ferent levels.
One way to cope w i t h this c o m p l e x i t y m a y be to emphasize the idea of
contexts. Contexts are sets of related goals, roles, r u l e s , assumptions, a n d
expectations about behavior a n d outcomes. It is the f r a m e w o r k w i t h i n w h i c h
a pardcular situation takes place. O p e r a t i o n a l l y , w e define a context every
time we say, " T h i s is a s i t u a t i o n , " w h e r e the b l a n k is some
descripdve, nonevaluative statement about w h a t w e are engaged i n . ^ T h e
notion of context parallels the concept of role a n d serves m u c h the same
purpose—providing a f r a m e w o r k of predictability and understanding to
social relations.
Within this construction, i n n o v a t i v e behavior is c a r r i e d out by i n d i v i d -
uals—but individuals w o r k i n g through roles, w i t h i n contexts that guide and
 46 • The Processes of Technological Innovation

shape h o w that behavior is executed. It is possible to learn a great deal about

these contexts and h o w they shape b e h a v i o r . A sociological analysis of cen-
tralization is a comment about a set of c o n t e x t u a l bounds a n d influences on
organization p a r t i c i p a n t s ; so is a n economic analysis of the use of factor prices
in m a r k e t i n g decision m a k i n g . D e p e n d i n g o n the relevant context, people in
organizations use a w i d e range of v o c a b u l a r i e s — s o c i a l , economic, technical,
p o l i t i c a l — t o describe w h a t is going o n .
A s w e proceed through the rest of this b o o k , readers are cautioned to
remember that the w a y technological i n n o v a t i o n is described a n d analyzed is
critically dependent on both the contexts available to the participants and the
contexts available to the person doing the describing. T o a significant degree,
" a l l is c o n t e x t " i n understanding the processes of technological innovation.
T h e c o n t e x t u a l frame of reference that w e w d l use is heavily dependent on a
modified stage/phase model of i n n o v a t i o n . T h a t i s , w e w i l l visit each of the
several r o o m s of the i n n o v a t i o n process a n d understand each, as w e l l as how
they interconnect to another. H o w e v e r , before w e e m b a r k on that journey,
let us first provide some a d d i d o n a l methodological/conceptual comments on
the interconnectivity of different levels of analyses.

Mixing and Matching Levels of Analysis

T h e theme of this chapter has been that technological i n n o v a t i o n is a process

that involves h u m a n systems at m a n y different levels and through m a n y dif-
ferent contexts. A c t i o n s undertaken by one i n d i v i d u a l or group of individuals
at one point in a system are likely to have a w i d e range of consequences, not
only for those people, but for others, more or less far removed i n time and
space. O r g a n i z a t i o n s set contexts for people w o r k i n g w i t h i n t h e m , a n d those
contexts are i n t u r n affected by action. T h e r e is no single level at w h i c h these
interactions are appropriately studied; rather, w e need w a y s of designing and
conducting research that take advantage of the m u l t i p l e x i t y of innovation
processes w i t h o u t being o v e r w h e l m e d by t h e m .
It is a basic premise of models of social systems that behavior can be
studied at v a r i o u s levels of aggregation. E a c h level of aggregation has certain
properties that are not possessed by the components of the level acting inde-
pendently. F o r e x a m p l e , a group is not s i m p l y the s u m of i n d i v i d u a l group
members' behaviors. T h e r e are distinct group behaviors. Aggregation is not
simply a process of pooling l o w e r levels. Simon's ( 1 9 7 3 ) analysis of the
" d e c o m p o s a b i l i t y " of systems a n d W e i c k ' s ( 1 9 7 6 ) discussion of "loose coupl-
i n g " w i t h i n systems both emphasize that behavior at any one level is a com-
plex function of the w a y s in w h i c h subunits come together to f o r m larger
units. T h i s m a y sound like a t r u i s m — h o w e v e r , it is one of the most commonly
neglected truisms in analysis of i n n o v a t i o n .
 Technological Innovation as a Process • 47

Most problems in grappling w i t h multiple levels of analysis stem f r o m

conceptualization, belief systems, or method. T h e s e problems are p r i m a r i l y
of three types:

1. failures to acknowledge the v a l i d i t y of different levels a n d units of

analysis
2. problems w i t h research design and measurement
3. confusion in relating data derived f r o m different methods.

The first error, failure to acknowledge the v a l i d i t y of different levels and

units, is often a function of basic d i s c i p l i n a r y c h a u v i n i s m . Some examples
include macroeconomists' failure to acknowledge the existence or a n a l y z a b i l -
ity of organizational behavior, psychologists' dismissal of organizational
norms and procedures as they affect i n d i v i d u a l behavior, and policy m a k e r s '
lack of understanding that a l l policy stands or fails at the level of the specific
industry, organization, or w o r k g r o u p . Perhaps the most persuasive and criti-
cal set of disciplinary blinders, is that w h i c h prevents technical people (physi-
cal scientists and engineers w h o " d o " technological i n n o v a t i o n for a living)
from understanding that the system of technological i n n o v a t i o n i n w h i c h they
are enmeshed is one i n w h i c h the c r i t i c a l ingredients of success lie i n nontech-
nical domains. T h a t i s , the economics of technological i n n o v a t i o n , or the
management of technological i n n o v a t i o n , m a y be more important than the
technology itself.
There seem to be only t w o w a y s i n w h i c h the p r o b l e m of disciplinary
rigidity can be addressed. O n e solution w o u l d be for researchers on i n n o v a -
tion phenomena to become conceptually a n d methodologically proficient in
several fields and become, i n effect, interdisciplinary people. W h i l e this might
be a worthy goal, it is difficult to fulfill given current academic mores. I n fact,
there is a significant literature on the issues a n d difficulties of interdisciplinary
sdence ( E p t o n , P a y n e , a n d Pearson 1 9 8 5 ; T e i c h and Pace 1 9 8 5 ; Porter and
Rossini 1985). A second solution is to use a research team approach i n v o l v i n g
pooling of different specialties. T h i s is the m u l t i d i s c i p l i n a r y a p p r o a c h , and
one that is achieving greater credibility and use i n v a r i o u s settings.
Epistemological conflicts n o t w i t h s t a n d i n g , difficulties i n dealing w i t h
muldple levels of analysis m a y simply be the result of inadequate research
design and inappropriate research methods. Research o n organizational phe-
nomena often requires the use of statistical inference: procedures for d r a w i n g
conclusions about a set of hypotheses on the basis of a relatively s m a l l number
of observations. H o w m u c h i n f o r m a t i o n one c a n isolate about the relation-
ships of interest f r o m " n o i s e " produced by uncontrolled factors depends to a
great extent upon the design of the research. E l e m e n t s o f the research design
indude; operationalization of the v a r i a b l e s , development of a sampling p l a n .
 48 • The Processes of Technological Innovation

determination o f sample size, selection of research instruments a n d instru-

ment items, development o f the data collection p l a n , a n d selection of analyti-
cal techniques.
Nonetheless, these rules of research design a n d statistics are ruthless and
u n y i e l d i n g , a n d i n real-life settings especially difficult to u p h o l d . A s Festinger
a n d K a t z ( 1 9 6 6 , 1 7 3 ) note about sampling considerations:

M a n y investigators protest that they do not make these decisions—these

decisions are made for them. O n e does research in industries into w h i c h he
can get entree . . . But such situations frequent though they may be do not
obviate the necessity of the nature of a sample and its characteristics.

T h i s suggests that the limited p o w e r a n d robustness o f the existing research

o n technological i n n o v a t i o n m a y be as m u c h a function o f politics a n d logis-
tics as o f inadequate methodological tools. F o r e x a m p l e , m a n y o f the issues
of w e a k inference a n d multiple levels o f influence nested w i t h i n one another
could be addressed i f experimental methods were applied to understanding
technological i n n o v a t i o n ( F a i r w e a t h e r and T o r n a t z k y 1 9 7 7 ) . O f the hundreds
of citations i n this v o l u m e , no more t h a n a h a n d f u l involve true experiments.
T o t u r n o u r attention to the second p r o b l e m , the field is beset w i t h basic
problems o f w h a t is a variable a n d h o w to measure i t . Simple aggregation of
i n d i v i d u a l measures to f o r m group descriptors m a y produce i n v a l i d compo-
sites. F o r e x a m p l e , i n some analyst's eyes, the R&cD spending of a f i r m mea-
sures its ability to generate n e w products to compete w i t h its competitors; in
contrast, average R & D spending i n a n industry presumably measures mainly
the dependence of that industry o n n e w technology. D o e s either " m e a s u r e " the
u n d e r l y i n g factor.' Burstein ( 1 9 7 8 ) a n d R o b e r t s , H u l i n , a n d R o u s s e a u (1978)
comment o n this p r o b l e m at length. I n addition to aggregation problems there
are also disaggregation problems. T h e r e are certain " g l o b a l " features of larger
units (Burstein 1 9 7 8 ) that are incorrectly attributed to component units—for
e x a m p l e , the tendency to assume that the p r o d u c t i v i t y of a nation such as
J a p a n is a n index o f the values o f the i n d i v i d u a l people i n v o l v e d . M u c h of this
p r o b l e m could be alleviated if the conceptual understanding o f different social
units, f r o m different levels of social aggregadon, matched the precision of the
data themselves.
T h e d r a w i n g of inferences about behavior at one level f r o m data mea-
sured at another level is not automatically a n e r r o r , but it frequently is. Unfor-
tunately, w e often tenant to d r a w inferences about f i r m s f r o m the behavior of
i n d i v i d u a l s . T h i s m a y yield inferences that are either w r o n g o r w e a k (Fry
1 9 8 2 ) . A t other times w e m a y face the inverse p r o b l e m , as w h e n w e try to
d r a w inferences about the productive behavior o f individuals f r o m data about
w o r k groups' o v e r a l l p r o d u c t i v i t y . T h u s , technique a n d purpose are often at
odds ( M o s s h o l d e r a n d Bedeian 1 9 8 3 ) .
 Technological Innovation as a Process • 49

Aggregation problems exist not only across structure or space but also
across time ( K i m b e r l y 1 9 7 6 a ) . P o o l i n g data describing events that occur at
different dmes is a c o m m o n technique; it is seldom done w i t h explicit atten-
tion to the inferential consequences of this a s s u m p t i o n . A t its e x t r e m e , this
pooling results in r e m o v i n g time f r o m analyses entirely a n d c a n lead to a n a -
chronisms such as " c o r r e l a t i n g yesterday's innovativeness w i t h today's inde-
pendent variables" (Rogers a n d E v e l a n d 1 9 7 8 ) in a reversal of the plausible
causal ordering. S i m i l a r l y , the aggregation of observations f r o m different time
points is frequently done s i m p l y i n order to increase sample sizes.
The third p r o b l e m , methodological choice, is p a r t i c u l a r l y difficult to
resolve. Although the a i m of all varieties of research is to describe the same
set of organizational actions, there are serious epistemological questions
regarding the techniques best adapted to identifying the critical features of
that behavior. H o w to balance off a finding derived f r o m experimental
methods (Boruch, M c S w e e n e y , and Soderstrom 1 9 7 7 ) against one derived
from a case study analysis ( Y i n 1 9 8 1 ; Greene and D a v i d 1 9 8 1 ) , or a highly
quantitative finding against a more qualitative one ( M i l e s 1 9 7 9 ; V a n M a a n e n
1982), is not intuitively apparent except to ideological partisans of a p a r t i c u -
lar point of view.
Much attention has been given to the question of w h a t is the " r e a l " level
of analysis for particular issues, given the difficulty of multilevel inference.
However, restricting analyses to one level does not seem to be either helpful
or necessary. Phenomena do not belong neatly to one level alone of a hierar-
chical system, and analyses that attempt to restrict their inferences to one level
will miss many of the most interesting interactions that affect their d a t a .
Thus, instead of spending time attempting to find the proper type or level of
analysis, we encourage using all types of research strategies and data sources,
with due attention to the limitations of each. T h e r e is a tradition of, a n d argu-
ment for, using data f r o m different research methodologies i n a triangulation
approach (Denzin 1 9 7 0 ; K i r k a n d M i l l e r 1 9 8 6 ) . G i v e n the c o m p l e x i t y and
mcssiness of innovation p h e n o m e n a , it is p r o b a b l y a w i s e strategy, and one
that students of the field are increasingly r e c o m m e n d i n g ( U t t e r b a c k 1 9 8 6 ) .

Summary

In this chapter we have stressed the extensiveness a n d pervasiveness of techno-

logical innovation processes i n h u m a n systems. T h e s e processes operate over
long periods of time and spread their consequences liberally across all parts
of society. T h e y can be understood in terms of m a n y different a n a l y t i c a l
vocabularies, and at m a n y different levels of aggregation. E a c h line of
inquiry—into different stages, into different levels, into different c o n t e x t s —
 r

50 • The Processes of Technological Innovation

helps us u n d e r s t a n d p a r t of the issue, but n o one line of a t t a c k is either neces-

sary o r s u f f i c i e n t .
T h e themes s o u n d e d here c a r r y into subsequent chapters: i n n o v a t i o n as
both thing a n d action; technology as inseparable f r o m h u m a n v a l u e s a n d pur-
poses; a n d technological i n n o v a t i o n as a c o m p l e x i n t e r a c t i o n of people, scien-
TI t i f i c concepts, a s p i r a t i o n s , a n d consequences. T h e m o r e w e l e a r n about the
c o m p o n e n t s of these processes, the m o r e w e realize that the w h o l e remains
b e y o n d o u r g r a s p . B u t after a l l , h u m a n efficacy is a l w a y s at the m a r g i n ; w h a t -
ever w e c a n f i n d o u t t h a t helps us a v o i d being w h i p s a w e d by o u r o w n systems
is better t h a n n o t h i n g .

Notes

T 1. A s we will argue in later chapters, there is some doubt that this stage is often
pri: reached, particularly in so-called "high technology" areas.
At 2. T h e r e is an intentional relationship here to Y i n ' s (1979) concept of "passages"
in innovation.
on 3. T h i s is not totally accurate. A s Fry (1982) has s h o w n , relationships {in his
cor case, between structure and technology) may appear attenuated if one uses an inappro-
priate level of analysis.
an(
4. Generally excluded from this review is a detailed discussion of questions rela-
po:
ting to organizational resources, because of considerable ambiguity in how such vari-
bui
ables should be interpreted ( M o h r 1 9 6 9 ; Bourgeois 1 9 8 1 ; R o w e and Boise 1981). See
tak
the section relating to organizational size issues.
CO]
5. T h e definition of context is obviously dependent to some degree on material
inr circumstances, but not a direct function of them. F o r example, one person faced with
pa the prospect of heavy snow may define the situation as " f u n , " while for another it con-
ho stitutes "shear torture."
va:
tra

View publication stats