Basic Social Processes
By Barney G. Glaser, Ph.D., Hon.Ph.D.
with the assistance of Judith Holton
Abstract
The goal of grounded theory is to generate a theory that accounts for a
pattern of behavior that is relevant and problematic for those involved.
The goal is not voluminous description, nor clever verification. As with
all grounded theory, the generation of a basic social process (BSP)
theory occurs around a core category. While a core category is always
present in a grounded research study, a BSP may not be.
BSPs are ideally suited to generation by grounded theory from
qualitative research because qualitative research can pick up process
through fieldwork that continues over a period of time. BSPs are a
delight to discover and formulate since they give so much movement
and scope to the analyst’s perception of the data. BSPs such as
cultivating, defaulting, centering, highlighting or becoming, give the
feeling of process, change and movement over time. They also have
clear, amazing general implications; so much so, that it is hard to
contain them within the confines of a single substantive study. The
tendency is to refer to them as a formal theory without the necessary
comparative development of formal theory. They are labeled by
a “gerund”(“ing”) which both stimulates their generation and the
tendency to over-generalize them.
In this paper, we shall first discuss the search for, and criteria of, core
variables (categories) and how they relate to BSPs. Then we go on to
a section on several central characteristics of basic social processes.
Lastly, we discuss the relative merits of unit vs. process sociology.
Core Category and Basic Social Process (BSP)
While grounded theory can use any theoretical codes, the basic social
process (BSP) is a popular one. As with all grounded theory, the
generation of a BSP theory occurs around a core category. While a
core category is always present in a grounded research study, a BSP
may not be. BSPs are just one type of core category—thus all BSPs
are core variables (categories), but not all core variables are BSPs.
The primary distinction between the two is that BSPs are processural
or, as we say, they “process out.” They have two or more clear
emergent stages. Other core categories may not have stages, but can
use other theoretical codes.
Without a core category, an effort at grounded theory will drift in
relevancy and workability. Since a core category accounts for most
of the variation in a pattern of behavior, it has several important
functions for generating theory. It is relevant and works. Most other
categories and their properties are related to it, rendering the core
category subject to much qualification and modification because it is
so dependent on what is going on in the action. In addition, through
these relations between categories and their properties, the core has
the prime function of integrating the theory and rendering the theory
dense and saturated as the relationships increase.
These functions then lead to theoretical completeness—accounting
for as much variation in a pattern of behavior with as few concepts
as possible, thereby maximizing parsimony and scope. Clearly
integrating a theory around a core variable delimits the theory and
thereby the research project.
Upon choosing a core category, the first delimiting analytic rule of
grounded theory comes into play. Only variables that are related to
the core will be included in the theory. Another delimiting function
of the core category occurs in its necessary relation to resolving
the problematic nature of the pattern of behavior to be accounted
for. Without a focus on how the core category resolves, solves or
processes the problem, the analysis can drift to accounting for
irrelevancies in the pattern, instead of being forced to conceptually
integrate the relevant categories around the main concern.
Yet another delimiting function of a core category is its requirement that
the analyst focus on one core at a time. Thus, if two core categories are
discovered—or one worked on before another emerges—the analyst
can choose one, being sure of its relevance. S/he then demotes the
other by filtering it into the theory as a relevant “near core”—but not
core—variable. Thus, in Time for Dying (Glaser & Strauss, 1968), we
included ideas about awareness, but only insofar as they affected
time. And in Awareness of Dying (Glaser & Strauss), 1967, we did the
reverse. By this method, the analyst can be sure that the other core
does not disappear. It can still take a central focus in another writing.
Many studies yield two or (sometimes) three core variables. To try to
write about them all at once with no relative emphasis is to denude
each of its powerful theoretical functions.
Discovering the core category is our grounded answer to the perennial
research problem of “which focus.” This focus cannot fail, since it is
systematically generated, by a sentence-by-sentence grounding in its
capacity to be relevant and to work. In contrast, to core a study and
its theory around a “pet” sociological interest or a logically elaborated
interest from scholarly writings can easily miss on the many functions
mentioned above. Since it is not grounded, there is no assurance that
it will integrate any other categories or properties or account for any
or sufficient variation in a behavioral pattern. Nothing—or not much—
may emerge as related. Plus, it derails the analyst from discovering the
true core. Thus the analyst cannot start a grounded theory study with
preconceived notions, from whatever source—even grounded—about
what will work in a specific project. The focus must emerge on its own
to do justice to the data, while accounting for significant variation in
problematic behavior.
Discovering Core Categories
Looking: First, the analyst should consciously look for a core variable
when coding his data. As s/he constantly compares incidents and
concepts s/he will generate many codes, while being alert to the one
or two that are core. S/he is constantly looking for the “main theme,”
for what—in his or her view—is the main concern or problem for the
people in the setting; for that which sums up, in a pattern of behavior,
the substance of what is going on in the data, for what is the essence
of relevance reflected in the data, for categories (gerunds) which bring
out process and change (two properties of BSPs).
As the analyst asks these questions while coding, analyzing and
theoretically sampling, s/he becomes sensitized to the potential
answers. Possible core categories should be given a “best fit”
conceptual label as soon as possible so the analyst has a handle
for thinking of them. The analyst may have a feel for what the core
variable is, but be unable to formulate a concept that fits well. It is OK
to use a label, which is a poor fit until a better fit eventually comes.
As the analyst develops several workable coded categories, s/he
should begin early to saturate as much as possible those that seem
to have explanatory power. This way s/he will see which category is
related to as many other categories and their properties as possible.
S/he theoretically samples to maximize differences in the data to help
saturate the categories. This is relatively easy with quantitative data.
The analyst need only run possible core categories against all other
variables to see how much each relates to others. With qualitative
data, it is more difficult since these relations must be kept track of in
memos, which get spread out until sorted. The core category must
be proven over and over again by its prevalent relationship to other
categories thereby integrating them into a whole.
When the analyst starts coding, categories tend to emerge quickly,
giving the appearance of finding core categories. But the analyst
should be suspect of these as core. It takes time and much coding
and analysis to verify a core category through saturation, relevance
and workability. It always happens that a category will emerge from
among many and “core out”—but it happens “eventually”! And, even
then the analyst may still feel s/he is taking a chance on selecting
what the core variable is, until it is finally proven by sorting data into
a theory that works. The more data, the more sure the analyst can
become of saturation, relevance, workability and integratability of the
chosen core. Time and data can be expensive; in smaller studies
an analyst often has to take chances. Certainly, deciding on a core
category tests the analyst’s skill and abilities. If s/he acts too quickly
on a thin amount of data, the analyst risks ending up with a large array
of loosely integrated categories, and a thin, undeveloped theory with
little explanatory power.
Criteria: It is helpful to sum up the criteria by which an analyst can
make judgments as to the core category.
1. It must be central; that is, related to as many other categories
and their properties as possible and more than other
candidates for the core category. This criterion of centrality
is a necessary condition to making it core. It indicates that
it accounts for a large portion of the variation in a pattern of
behavior.
2. It must reoccur frequently in the data. By its frequent
reoccurrence, it comes to be seen as a stable pattern and
becomes increasingly related to other variables. If it does not
reoccur a lot, it does not mean the category is uninteresting. It
may be quite interesting in its own right, but it just means it is
not core.
3. By being related to many other categories and reoccurring
frequently, it takes more time to saturate the core category
than other categories.
4. It relates meaningfully and easily with other categories. These
connections need not be forced; rather, their realization
comes quickly and richly.
5. A core category in a substantive study has clear and grabbing
implication for formal theory. The analyst can talk of hospital
shifts and immediately realize the implications of shifts as a
basic social condition in any twenty-four-hour-a-day work
operation and start to conceive of generating a formal theory
of work shifts.
6. Based on the above criteria, the core category has
considerable carry-through. By this, we mean it does not lead
to dead ends in the theory nor leave the analyst high and dry;
rather, it gets him/her through the analyses of the processes
s/he is working on by its relevance and explanatory power.
S/he literally carries through his analysis based on the core’s
use.
7. It is completely variable. Its frequent relations to other
categories make it highly dependently variable in degree,
dimension and type. Conditions vary it easily. It is readily
modifiable through these dependent variations.
8. While accounting for variation in the problematic behavior, a
core category is also a dimension of the problem. Thus, in
part, it explains itself and its own variation. While “becoming”
a nurse explains the process that student nurses go through
in relation to their training and their interaction with nursing
faculty, it also in part explains why a nurse becomes a nurse.
They engage in becoming to become, while becoming
also explains how they handle those largely responsible
for formalizing their entrance to the profession (Olesen &
Whittaker, 1968).
9. The criteria above generate such a rich core category that, in
turn, they tend to prevent two other sources of establishing
a core which are not grounded but, without grounding, could
easily occur: (1) sociological interest and (2) deductive,
logical elaboration. These two sources can easily lead to
core categories that do not fit the data and are not sufficiently
relevant or workable.
10. The above criteria also generate a false criterion. Because it
has so much grab and explanatory power, the analyst begins
to see the core category in all relations, whether grounded
or not in the data. While serving as a positive indicator of
the core, this logical switch must be guarded against so
that relationships among categories are earned through
emergence and not forced upon the data through deductive
logic.
11. The core category can be any kind of theoretical code: a
process, a condition, two dimensions, a consequence and so
forth. When it is a process, additional criteria also apply.
The “Process Out” Requirement of BSPs
Once the analyst becomes theoretically sensitized to the search
for core categories and those that process out, discovering core
categories—and BSPs in particular—becomes natural. Indeed, we
have found that analysts must be careful about tacking a gerund on
to any core variable and treating it like a process when, in fact, it does
not process out. For example, in one study, “shifting” was seen as
a BSP. After review, we found no stages and reconceptualized it as
“shifts”—a basic social structural condition confronting people and
organizations that have a twenty-four-hour-a-day operation.
The “process out” requirement of—at minimum—two clear, emergent
stages requires that the stages should differentiate and account
for variations in the problematic pattern of behavior. If not, the
stages collapse conceptually and there is no BSP. For example,
in information-gaining processes, the stages of playing completely
naive, playing mildly informed but needing correction, and finally,
playing knowledgeable, each results in a different interaction pattern
in bidding subcontractors. In this sense, a BSP processes a social
or social psychological problem from the point of view of continuing
social organization. Irrespective of whether it solves the problem, to
some degree, it processes it.
A process is something that occurs over time and involves change over
time. These changes over time ordinarily have discernable breaking
points—discernable to the extent that stages can be perceived, so
they can be treated as theoretical units in themselves, with conditions,
consequences (which may be another stage), other properties, and
so forth which are unique in form to each particular stage. Stages are
perceivable, because they sequence with one another within certain
temporal limits. Sets of codes related to these stages may “carry
forward” into one or more stages further on in the process.
Stages may be in vivo (generally perceivable by those persons
involved), or purely heuristic (generally not perceivable by the persons
involved, but demarcated by the sociologist for theoretical reasons), or
some shade in between. If the stages are built into the social structure,
they and their transition points will likely be clearly perceived by social
actors (e.g. receiving a diploma, passing a course of study, getting a
promotion from “worker” to “supervisor”, and so forth). Conversely,
stages that are perceivable before one goes through them would likely
be built into a social structure (Glaser & Strauss, 1971). However,
stages not determined by social structure can also be perceived by
social actors (“When they started joking with me I knew I was in”).
In some instances, stages may be perceivable by social actors only
after they have been through them. This would likely be the case with
stages that are marked by common sense indicators and such.
Some stages may be learned as persons go through them. For
example, milkmen, when learning to “cultivate,” learn from their coworkers
that a particular stage in cultivating a relationship is reached
when the customer routinely offers the milkman a cup of coffee (Bigus,
1972). This is, the novice learns, a “coffee stop” and is considered
the last and most successful stage of a relationship, if the customer
is worth it in monetary return. The novice is informed in one way or
another that when this occurs, he no longer need worry about the
relationship to the extent that he does others, and that “coffee stops”
will perform certain functions for him—a place to go to the bathroom,
a place to get a payment when one is needed, and so forth.
Stages, if perceivable by social actors, may be brought about by
their conscious intentions. Again, the milkman: once he learns about
the “coffee stop” stage, he consciously sets about cultivating to get
particular customers (the large ones) to that stage. Other stages,
particularly those demarcated by institutionalization, begin and end
without conscious effort on the part of participants.
A person may perceive the events that make up stages of a process
he is going through without perceiving the overall process or any
particular stages. These events may be perceived as idiosyncratic—
events that are unique to his own experience—rather than as stages
of a social process which many persons go through. A sociologist,
however, can perceive the stages because he studies large numbers
of individual histories and sees as social what individuals may see as
personal.
The development into stages prevents a BSP theory from being
static—a condition ordinarily found in most types of theory. It allows
one to follow changes over time, yet remain in grasp of a theoretically
“whole” process—which has a beginning and an end. When the stages
and their properties, conditions, consequences, and so forth are
integrated into the “whole” process, when each stage’s relationship
to the process and to the other stages—how they affect it, shape it,
and so forth—are integrated, then the process can be conceptually
followed from stage to stage, the change over time being theoretically
accounted for, without the imagery of the overall process being lost.
This allows a reader to momentarily focus on the dense codes without
losing grasp of the larger scope of the BSP theory.
Stages, then, function as an integrating scheme with which to tie
together various sets of conditions, properties, etc. in a manner that
allows for a high amount of densification and integration. At the same
time, stages allow for conceptual grab and tractability as well as the
theoretical tracing of and accounting for change over time.
Stages have a time dimension; that is, they have a perceivable
beginning and end. The length of time between these points may
or may not be fixed. In one instance, a stage may always be of
fixed duration. In another, it may last several days or weeks. This
will depend upon what brings about the transition from one stage to
another. If the length of a stage is determined by institutional timing,
for instance, it could always be of the same duration. The length of
time a stage lasts could also be determined by events that do not
occur according to a time schedule. A stage in a “residential career,”
for instance, could be determined by the move from renting to buying
a home. Thus, the renting stage (if such a stage were developed)
could last several months or many years.
The transition from one stage to another is ordinarily contingent
upon one or more things happening (e.g. the decision to purchase a
house—as above). This contingency may be in the form of a critical
juncture (Strauss, 1969) – a period of time between stages when
the occurrence or non-occurrence of a particular critical event (or
whatever) will determine whether a new stage is entered (a stage
is skipped, one of several possible stages is entered, etc.) or the
previous stage is maintained. For example, exploratory surgery in
search of cancer could be such a critical juncture. If cancer is found,
the beginning stage of a dying trajectory or a recovery trajectory
(depending upon the severity of the cancer) may be entered. If cancer
is not found, a diagnosing stage may be returned to.
The transition from one stage to another may not be as clear as it is
when a contingency or a critical juncture marks it. It may, instead, be
marked by a general set of indicators in such a way that the transition
point is somewhat blurry. For example, an “acceptance” stage may
be entered around the general time that insiders begin to allow a
newcomer to joke about the group, let him attend insider affairs,
disclose “secrets” to him, and so forth. An exact time of transition
may be impossible (or arbitrary) to pin down, but the transition may
be obvious later after a short period of time, through the gradual
occurrence and clarity of a set of indicators.
We now turn to a discussion of further characteristics of BSPs. Much
of what we shall say in the next section applies in general to all core
categories, except when the property specifically refers to process.
More about the Basic Social Process
Stages, as we have just seen, are the prime property of BSPs,
however there are several other defining properties: pervasiveness,
full variability and change over-time. BSPs are pervasive since
they are fundamental, patterned processes in the organization of
social behaviors which occur over time and go on irrespective of the
conditional variation of place.
The pervasiveness of such core processes gives rise to the word
basic in BSP. BSPs, then, are more than just heuristic devices that
allow sociologists to conceptually order the social world. BSPs
are theoretical reflections and summarizations of the patterned,
systematic uniformity flows of social life that people go through, and
which can be conceptually “captured” and further understood through
the construction of BSP theories.
No matter what the sociologist does, s/he cannot alter the basic
substantive patterns of the process. S/he can only apply whichever
theoretical codes best illuminate variations in what is going on. Not all
persons go through a process in the same manner; that is to say, there
is much variation. But, a BSP theory can uncover what condition or
variables give rise to particular variation and can therefore theoretically
account for them. For example, “becoming” is basic, occurs over time,
and is still becoming no matter where it occurs, and irrespective of
how it is varied by current conditions. So, for instance, there’s a basic
pattern or process to becoming a nurse, regardless of variation in
individual experiences.
The pervasiveness of BSPs, due to their fundamentality to social
organization makes them necessary, unavoidable processes,
irrespective of variations. However, social organization itself being
sets of infinitely variable conditions makes BSPs fully variable. By
this, we mean that although BSPs are activated through the units of
social organization, they are abstract of any specific unit’s structure
and can vary sufficiently to go on in other, very different units. Thus,
recruitment processes go on no matter what the social unit; people
are continually brought into units or eventually the units disappear.
As such, their full variability makes BSPs independent of structural
units: that is, free of their time and place and the perspective of their
participants and fully generalizable as abstract processes to be found
anywhere they may emerge.
As an analytic unit, BSPs receive relative emphasis over the structural
unit in which they are analyzed. The essential point is that, for
example, we focus on becoming processes when talking of nursing
education, not on the structured unit—the school—in which the study
took place. The school is merely a set of varying conditions of a
becoming process.
The full variability and generality of BSPs transcend the nature of
any structural unit and hence, unit-focused theories. They transcend
the boundaries of unit analyses as we understand the general, basic
processes that shape people’s lives instead of solely their particular
units of participation. (We shall discuss these properties of BSPs in
relation to unit analysis more fully in the next section of this paper).
BSPs are not only durable and stable over time but they can account
for change over time with considerable ease of meaning, fit and
workability. Since process connotes a temporal dimension, focus
is on patterned lines of conduct as they occur over time under
different conditions that generate change. Thus, change is fully as
much an inherent feature of BSPs as their stability and variability.
This characteristic contributes toward solving a perennial problem
in sociology—accounting for change. The notion of change is not at
all built into many other generic concepts in sociology such as social
class, role, social structure, social system, functionalism and so forth.
These categories can often be rejected when it comes to analyzing
change since they become obsolete or clumsy in reflecting the
realities of change.
When things change because of full variability, new conditions,
stages, and transitions can be added to the BSP in order to handle
the change. Take for example, locating “progress in a class” as a
process. Students are able to locate themselves by comparing grades
with one another. But, suppose a particular school eliminates grading.
New methods of locating may be found, such as noting how often one
is called upon in class, or other such subtle forms of “feedback.” At
any rate, the theory of locating can be modified to handle the change.
Whatever changes and adjustments take place can simply be added
as conditions or consequences of the process. The theory has not
been “disproved” or made obsolete in any way. A process of locating
still exists—it has merely been modified slightly in form, densified and
made more general.
BSPs can also handle change over much longer spans of time by
merely adjusting for the changes in conditions in the same general
way that adjustments could be made for changes encountered in
going from one substantive area to another. What would be accounted
for theoretically would be the absence of some conditions and the
presence of new or different conditions. The basic theory, however,
would remain intact. The “size” of temporal scale is included.
Basic Social Psychological Process (BSPP) and Basic
Social Structural Process (BSSP)
There are two types of BSPs—basic social psychological process
(BSPP) and basic social structural process (BSSP). A BSPP refers
to social psychological processes such as becoming, highlighting,
personalizing, health optimizing, awe inspiring and so forth. A BSSP
refers to social structure in process—usually growth or deterioration—
such as bureaucratization or debureaucratization, routinization,
centralization or decentralization, organizational growth, admitting
or recruiting procedures, succession, and so forth. A BSSP abets,
facilitates or serves as the social structure within which the BSPP
processes. Thus the growth of free clinics facilitates the prescribing
process of birth control and family planning (Lindemann, 1974). The
growth of spiritualizing of health food stores was necessary to “hippie”
health optimizing (Hanson, 1976). Consolidating a revolution is
accomplished by bureaucratization of charisma (Weber, 1947).
Most sociology these days focuses on social psychological process
and assumes social structural process—or simply treats it as a
changing set of structural conditions—without formulating it clearly
as a process. The question remains is the latter all that necessary?
Perhaps the BSPP is more prevalent and relevant to understanding
behavior, since one does not need the BSSP to understand it, but
usually one needs a BSPP to understand the focus on a BSSP. This
question is, of course, to be answered empirically for any particular
study. But given this prevalence, BSP implies a BSPP and when the
analyst is generating a social structural process theory, he states it
clearly as such and uses BSSP.
Society swings on the relevance of its interest, sometimes focusing
on social psychological problems (getting poor people to upgrade)
or sometimes focusing on social structural problems (providing
opportunities for work, health distribution systems, government
programs). Sociologists follow both foci. The most sophisticated
sociological renditions include both processes, however; perhaps
most will focus on the social psychological. It takes skill and clarity of
purpose to mix both with full development, as opposed to focusing on
one and using variables from the other.
Two general kinds of mix occur. One is that a BSP includes both
BSSP and BSPP. Examples are admitting, screening or recruitment
processes to an organization. The recruitment to a fraternity in college
is a clear mix of social psychological and structural in the screening
and initiation ceremonies. The other type is that the BSPP and BSSP
are clearly separate. For example, building housing tracts with better
homes and on better terrain is a process growing builders go through.
At the same time people are upgrading their housing circumstances
when they choose new neighborhoods with better homes, schools,
roads, parks and so forth. The new neighborhood can easily include
new homes or old homes or both. As another example, developing
health food stores was clearly separate from spiritualized, health
optimizing.
When the BSSP follows and facilitates the BSPP, it takes on
properties of the latter. Thus, the growth in health food stores
occurred by taking on properties of the health optimizing process
that it services; e.g. they sold natural vitamins with rhetoric. And vice
versa, when the BSSP comes first, the BSPP takes on properties of
it. Thus, in the beginning, birth control prescriptions took on the rules
of family planning agencies. Women had to be married at one time to
get a prescription for birth control. When the disjunction is great, as
in this case, the social psychological may either exert a change over
the social structural or may be purged. Thus, BSPPs can become
structural conditions that affect the nature of BSSPs, and vice versa.
In this way, a theoretical link is made between the two general levels.
The theoretical links that relate the two are many and emergent. Being
analytically clear about their separateness allows for a well formulated
analytic mix of the two. Otherwise, an analysis tends to become
confused or unclear as to the referent process. For example, how does
one analyze job transfers in an occupational career as related to time
for personalizing rental housing, without a notion of how to develop
both processes? Or how does one analyze upgrading life styles in
housing related to unavailability of new and better housing, without a
clear picture on the disjuncture of the stages of each process?
An analysis can emphasize the BSPP or BSSP, or some mix of the two,
depending on which process or which mix emerges as more relevant
in the situation under study. In studying a process that optimizes
change, fluidity, and unfreezing of behavioral patterns, it is likely that
the emergent mix would emphasize the BSPP. In studying a structural
phenomenon as it is growing, such as behavior in new communes or
people engaging in a new health practice, one would also bring in the
new BSSP that supports the BSPP. In studying a phenomenon that
requires little change in existing support systems, structural process
might not be as important, for instance, as a process occurring in
a bureaucratic setting where the actors have little control over the
structural support. Even in such a situation, however, there may be
informal modifications of the formal support structure.
Beside the above defining properties, a BSP has other important
characteristics. For instance, a BSP applies a theoretically useful
approach to deviance. It is, as well, systematically tied to a
methodology. Both characteristics are further elaborated below.
BSP and Deviance
It seems that most sociological theories are unable to explain with
ease “negative or deviant cases” of whatever it is they are supposed
to explain. So, they must resort to the use of additional theories—
ordinarily some sort of deviance theory. Since deviant events could
easily be explained as an integral part of a normal basic social process
that takes place under certain conditions, there is no need to see the
events as deviant or extraordinary. As the idea of basic social process
becomes commonly used, the notion of “negative case” disappears.
What were once considered negative cases merely highlight further
conditions under which behavior varies according to the pertinent
basic social process.
It is an error for sociologists to preconceive certain behaviors as
fundamentally deviant, but even more an error for them to assume
from the start that the most relevant thing about a particular behavior
is its deviant dimension (regardless of how “deviant” is defined). Even
if it is a behavior that is unquestionably far from general societal
norms, values, etc., there is no reason, before it emerges, to take that
as a starting point for analysis of the behavior. Such a consensual
label may, in reality, have little to do with the motivation, organization,
etc. of the behavior. Whether or not it does is a matter for empirical
inquiry. The starting point is to discover the BSP.
If the analyst were to begin with the preconception that a particular
behavior, organization, or whatever, was deviant and that was the
most important thing about the study, the chance is very high that
s/he would miss the core and relevance of what is actually happening.
To use an example: If s/he were to study brothels (which one can
safely say are generally considered deviant) from the point of view
that the fact of their deviance is the most important thing about them
sociologically, s/he would likely miss the more general relevant fact
that sociologically—in terms of structure, function, organization, and
process—they are similar to barber shops, beauty salons, garages,
and so forth. All are servicing operations.
All of these organizations service persons or their belongings. All have
steady as well as casual clients. All encourage their clients to remain
on the premises only while they are being serviced. After servicing,
they are “spent” and are no longer useful until they require servicing
again, and so forth. These seemingly different organizations have
much in common sociologically, regardless of how they are seen and
defined in common sense terms, and regardless of whether or not
they are defined as deviant. Servicing need not be seen as deviant or
non-deviant sociologically unless it is discovered that the deviant label
has consequences for the servicing operation and those persons who
are a part of it. In the case of the brothel, the deviant label would likely
result in its being more isolated, less obtrusive, and so forth, than
many other types of service operations.
In other words, from a BSP view, the deviant label (i.e. the fact that
other persons see the activity and the organization as deviant) is
merely one of many conditions that affect the servicing operations.
Anyone who questioned the women would soon discover that their
main concern is about servicing efficiently not about being “deviant.”
In this fashion, deviance is put in integrative perspective as part of
a BSP, rather than being developed as a separate body of theory.
As such, its part in the development of theory would be reduced
in importance in terms of the amount of time and effort spent, but
increased in terms of its contribution to an integrated theory of what
makes a part of society work.
If the analyst is interested in accounting for how particular persons
engage in an act or series of acts which happened to get labeled
deviant or have great potential for such a thing happening, a BSP
approach would look different from other approaches, primarily
because the grounded explanation for the behavior would be
contextualized and multivariate.
It would be contextualized in that it would not seek to explain too
much (as most other theories do), but rather would seek to explain
the sources (i.e. the conditions, properties, and so forth) of “deviance”
within a particular context such as a servicing operation. Once enough
grounded data has been gathered, presumably through several
studies and through the use of theoretical sampling, it may be possible
to lift the theory out of particular contexts and elevate it to a more
formal level. This could be accomplished if a number of dimensions,
properties, etc., were discovered which were cross-contextual enough
to form a foundation for a formal theory. However, this would not be
taken as the starting point (as it is in functionalist theory, for instance)
but rather as the advancing of a substantive theory to a formal one,
abstract of time and place.
A BSP view would be multivariate in that it would seek to discover all
of the many relevant variables (conditions, consequences, properties,
etc.) that constitute the process leading up to a particular form of
“deviant” behavior as covariant among other behaviors. In contrast
to this, the ordinary approach is to preconceive several variables
and then go out and try to verify their existence (overlooking all
the other possible variables which come into play). In addition, a
grounded BSP would pick up and integrate structural as well as social
psychological variables. The relationship between these various
levels of variables could be shown; how they interact and affect one
another in a systematic way. This has not been accomplished by the
multivariate theories that exist presently. They have merely admitted
that different levels of variables are involved in the explanation of
deviant behavior.
BSP and Methodology
As BSPs are densified and integrated, they may become multivariate
to the point of including variables from other disciplines, such as
psychology, political science, medicine and so forth. They easily
become stages in process, consequences or conditions. Thus, as
an isolating BSP, mental depression can cause social isolation that
can cause physical illness that results in hospitalization, with further
isolating in an isolating BSP. One handles emergence with whatever
categories (from whatever discipline) that fit and work and that the
analyst is trained to understand.
Since basic social processes are fundamental patterns in the
organization of social behavior as it occurs over time, the BSP
conception is a generic theoretical construct of the same genre as
Max Weber’s “ideal type” and Alfred Schutz’s “homunculus.” However,
unlike these conceptions, the idea of BSP (and core variable) was
developed within and is systematically tied to a specific methodology
for generating theory. The conception is not a presupposition of the
methodology, but rather is a product of its operations. The theoretical
construct—BSP—was conceived as a by-product emergent in the
process of doing and developing the methodology of grounded theory
research. In contrast to ideal types and homunculi, BSPs are more
than post hoc honorary labels. The BSP is fully “operational” at every
step of the grounded research process. This is not the case, so far as
I know, with any other type of theory construction. Weber and Schutz,
for instance, leave the operationalization of their theoretical type up to
one’s imagination. This may allow for flexibility, but it also allows for
deductive speculation and floundering before a research method and
effort is applied.
Grounded theory methodology does not rely solely on “cleverness,”
“ingenuity,” “insight,” and so forth, yet it is not so rigid and specific that
it can be learned and carried out by mere “technicians.” It requires
theoretical sensitivity as well as technical skills, and some persons will,
of course, be better at it than others. It also requires a specific course
of training (by teaching or reading) because it is a system that must be
used in whole. If it is used in part, or if parts are used incorrectly, it will
work less than properly. We have learned that analysts who use it only
partially are not likely to realize this, because many of its advantages
are not evident until it is used as a whole (e.g. the advantages of
writing memos, coding, sorting and so forth—both individually and
combined—become evident primarily through experience in doing
these things). This is not to say, however, that one should use it as
a whole or not at all. Every step used will improve one’s ability to
construct theory, regardless of what kind. The methodology provides
a perpetual development of skill as one uses each part.
BSPs can be developed by this methodology at various levels of
conceptual abstraction ranging from substantive theory (theory about
a specific substantive area—e.g. Karate) (BEESON, 1973) through
general substantive theory (theories about several similar substantive
areas—e.g. kinds of physical self-defense) to formal theory (theory
abstract of specific, substantive times and places areas—e.g. self-
defending). Thus BSPs can be conceptually ordered according to
abstraction, but each level is always theoretically and methodologically
linked with a less abstract level and with systematically collected data
of the empirical world. They never become operationally distant or
remote from reality. We might add that BSPs are not theories of the
middle range.
Finding a BSP
There are two basic models for finding a BSP; by discovery and by
emergent fit. By discovery, the analyst goes to a fairly contained social
unit attempting by observation and interviewing to see as much as
possible and find out the most salient social problem of the people
there. Then s/he discovers the core variable—hopefully a BSP—that
accounts for most of the variation in the behavior about the problem.
S/he then switches focus from studying the unit to studying the
process and proceeds to generate a substantive theory of the process
by constant comparisons of incidents within different comparative
groups in the same substantive class.
By emergent fit, the analyst has a BSP—discovered elsewhere—and
wishes to extend it or to do a grounded formal theory of it. S/he then
proceeds to find groups within which to study the BSP and, as in the
first model, starts comparing incidents and groups within or between
classes of units to achieve a level of generality, whether general
substantive or formal.
Of course, we favor the first model, but since many BSPs are known
already, some analysts may prefer the second model. It has, however,
various pitfalls. In discovering the emergent fit, the analyst should be
cautious about assuming that if the BSP fits, it is the core variable of
that unit. It very likely is not; the BSP is being imposed for the purpose
of generating a theory of it, not of explaining the variation of behavior
in the unit studied. Thus one can study temporal pacing in just about
any social unit, but it is seldom, if ever, the core variable of the unit.
Since it is not the core variable, the BSP will usually be less than
densely developed in the study unit. It will very likely become overshadowed
by a more salient core variable or BSP. Thus using the
second model, the analyst skips between many chosen units looking
for grounded densifications of properties and does not overwork any
one group and incidents in a unit for what is not their BSP as it would
be for a discovered BSP.
Furthermore, the second model is somewhat contradictory to the first
and to the main theme of this paper, but it has a place in grounded
theory if done carefully—since there are many grounded BSPs
already discovered that need further development within and between
substantive areas. The second model looks a bit like deductive, logical
elaboration, but it is not, providing the analyst follows the grounded
approach. S/he does not start “empty” or “non-preconceived” as in
the first model. S/he engages in pre-emergent analytic thinking, and
sampling before approaching the field. But once in the field, s/he starts
correcting early thoughts and follows the grounding in subsequent
theoretical sampling. And s/he ends up as s/he would in the first
model, searching for comparison groups, as it becomes clearer and
clearer where to go for fit as the theory develops.
There seems to have arisen a tacit rule in naming BSPs. It is turning
a substantive noun or verb into a gerund. Thus we have “friending”
and “becoming” respectively. While most BSPs are labeled with a
gerund, not all are; thus, career, alarm system or recruitment system.
As we said above, caution should be applied in over-use of gerunds.
They may mask a basic social structural condition, such as “security
system” or “shift”(as in our earlier example). As in all grounded theory
work, there is an area for theoretical creativity in labeling and rendering
the BSP or core variable.
As the analyst becomes practiced in spotting and conceptualizing
BSPs, s/he should avoid a probable occurrence. In reading others’
works, a BSP may become evident, which the author did not know s/he
had in the data. The analyst should say as much in his/her own work,
and not attribute the idea to the author. The analyst should distinguish
his/her good idea from the author’s “good data but conceptual miss”.
In fact, most BSPs are implicit and taken for granted in data, both by
sociologists and participants alike. Only with training does the analyst
see the strong contribution of a BSP to the on-going activity in the area
under study, and only then can a theory be consciously generated for
a BSP.
BSPs Compared to Units
Most sociology is focused on a rendition of a social structural unit.
That is, no matter what the substantive issues or concepts, or
whether the study is description, verification or theory building, we
read about properties of a unit; persons, groups, organizations,
aggregates, statuses, nations, and so forth. In contrast, in this paper
we have placed a relative emphasis on social process as the focus of
analysis. We generate properties of process. It is important and useful
to develop here the distinction between unit analysis and process
analysis, so that their relative use and merits for sociology can begin
to be clearly understood and used accordingly.
In itself, the focus on either unit or process sociology is not intrinsically
meritorious. The test of relative worth lies in how well each may
contribute to the knowledge of sociology and the purpose at hand. We,
of course, are biased toward process, as we see many comparative
advantages in the transcending nature of BSPs. The reader must
make his/her own calculations for each project. These distinctions
listed below are opening ideas, not final dicta. Some items do not have
to occur, but empirically, they do.
UNIT
PROCESS
1. Relative Focus
Process is one property of the
unit. Analysis focuses on unit
itself.
A unit is a place where a process
goes on and it provides a set
of conditions for its operation.
Analysis uses properties of unit,
not unit itself.
Focus is on process as it explains
or processes a problem or behavior
pattern.
2. Freedom From Time and
Place
Unit bound. Rendition of unit is
always bound by its time and
place during period of study.
Process is free of unit’s time and
place. These properties of unit are
only varying conditions. Another
unit varies process differently.
3. Generalizing
Finite to unit; analyst can
only generalize a study to a
similar, usually larger unit.
Generalizing is difficult and
slow as must study large unit
to analyze differences or use
random sampling of smaller
unit. Number of units to
generalize to is limited.
Fully generalizable quite easily, as
a BSP transcends the boundaries
on any one unit by just varying it for
another unit’s properties. Thus, the
analyst generalizes a substantive
BSP to a generic BSP. BSP is more
general as it may apply to all units.
4. Action
Provides the conditions
that more or less allow the
action. Units rely on BSPs to
run. Units are where BSSPs
and BSPPs intersect. Units
themselves may be a BSSP
that processes very slowly,
compared to BSPP, and is
actuated by BSPP. A static unit
is a frozen BSPP.
The action of life is always in the
process rather than of the unit itself.
The unit is actuated by process as
it bounds and locates it. The action
process is a BSPP.
5. Freedom from Perspective
Study of unit is always from
perspective of analyst and/or
participants. Bias is part of
analysis as it is built (the
establishment view of a
corporation, for example).
BSPs are a separate perspective,
irrespective of the perspective of
participant or analyst. BSPs go
on irrespective of bias of analyst.
“Purging”is always purging,
becoming is always becoming,
no matter how perspectived the
rendition. Bias is just one more
variable in a multivariate analysis.
6. Durability
Time and place change so
studies of a unit becomes
obsolete, whether unit
description, unit theory, or unit
formulations of change.
BSPs are quite durable. They
transcend the fallibility of units
and, while keeping up with unit
changes, as units change, BSPs
get modified.
7. Transferability
Once out of generalizing
range, it is difficult and
hazardous to transfer ideas or
findings of one unit to another
unit. Transferring ideas about
a nursing school to an Air
Force academy probably does
not apply.
Since BSPs are fully general, they
transfer easily with modification.
Becoming applies to both a nursing
school and an air force academy.
8. Consultation Based on
Transferability
An expert on a unit is restricted
to that type of unit, and he
requires much knowledge.
An expert on a process can consult
on any unit where process is
occurring by just knowing general
process and applying it to new
conditions.
9. Misattribution of Source
To describe a process as a
property of a unit implies that
it is uniquely the result of the
people in the unit. This is
inaccurate. The unit simply
uses a general process. Thus,
“women in karate are trying to
neutralize sex status” implies
they produced this process,
which is inaccurate.
A BSP implies that it is being
used by the unit, not a source of
it, and the use varies within it. For
example, it is accurate to say that
women in karate use one mode
of neutralization of an otherwise
differentiating sex status.
10. Learning
Typical unit studies can be
boring unless on a deviant or
other particularly interesting
group. It is hard to remember
the plethora of facts, and
understanding the unit is
often bereft of intrinsic scope
of meaning, because of low
generality.
BSPs have much “grab”(they catch
interest quickly), because they
have high impact in meaning, are
easily understandable, and have
general ideas that are easiest to
remember.
11. Research Sampling
Random sampling of unit itself
is used so the analyst can
generalize to a large unit.
Theoretical sampling of properties
is used to generate to the theoretical
completeness of process.
12. Research Coverage
Full range of representative
factual coverage needed to
describe the unit accurately,
whether for description or
verification.
Theoretical coverage requires
only theoretical sampling of that
segment of all behavior needed
to generate an explanatory theory
of a process. The analyst does not
need representative coverage of all
behavior.
13. Research Accuracy
Units tend to require
accuracy so the descriptions
will be considered correct.
Statements are facts to be
believed, and subject to slight
correction.
Not crucial with a BSP, since
successive comparisons correct
categories and hypotheses.
Statements are hypotheses, thus
claimed as suggestions to be
checked out; they are not claimed
as facts.
14. Research Reading
Read as accurate description.
Unfortunately BSP theory is
still read by many as factual
description, not as hypothetical
generalizations.
15. Historiocity
Unit studies are fixed in time.
They are static. They are
cross-sectional; picking up a
moment in time, as if forever,
but it becomes outdated, thus
temporal scope is severely
limited.
A BSP, since it deals with on-going
movement, implies both a past
and a future that can almost be
extrapolated. A BSP has change
built into it, as it is modified to
incorporate new data. A BSP
considers categories as part of
larger ongoing process, historical
scope. A BSP is in motion, not
restricted to time.
16. Theoretical Impact
Based on the above
differences, unit analysis has
limited impact and scope.
Based on above differences,
a BSP allows for an expansive
amount of grounded theorizing
about every facet of social life. It
has high impact.
17. New Data
Typically refutes part of unit
study.
Generates more BSP theory by
comparing it and modifying theory
by extension and densification.
18. Relationability
Units are seen as separate
entities with definite
boundaries. Theory related to a
unit is not theoretically related
significantly to other units,
except perhaps to a larger
similar unit to which it may be
generalized. Thus unit studies
are non-integrative to social
organization, they make units,
which are similar on underlying
dimensions, seem separate,
which is only arbitrarily so;
e.g., normal and deviant
studies appear different, not
as two dimensions of the
same general process. More
fundamental patterns are
obscured.
BSPs, by cutting across and
transcending the boundaries of
separate units, provide ways of
relating units to each other through
the same process; e.g., cultivating
clientele, is a way of relating
milkmen to lawyers. Thus BSPs tie
social organization together. They
are integrating. BSPs also relate to
each other within units.
Sociology along Process Lines
The above comparisons clearly indicate the quite different appearance
and import that sociological renderings of the world will take in
generating grounded BSPs. Our effort is to show that focusing on
process, as well as on units, will facilitate theoretical development in
sociology. Process analysis will partly alter the conceptual appearance
of sociology by cutting across the transcending traditional concerns,
topics and boundaries, such as check forgers, political parties,
adolescents, homosexuality, prisons, patient care and so forth.
Much of unit sociology is delineated along lines that are not
theoretically contiguous, although they are treated as such. As we
indicated above, if a unit sociologist were to begin a study of brothels,
s/he would probably place the study in the traditional category of
“deviant behavior” or possibly “social problem.” In doing so, the
presumption is that the essence or at least a primary property of the
behavior to be studied is deviant or socially problematic. Concomitant
results will explain the motivations, attitudes, or other social
characteristics of persons who engage in such practices as distinct
from non-practitioners; i.e., “normals”. However, in categorizing
brothel activities as merely another instance of deviant behavior,
other—perhaps more central characteristics of the phenomenon—are
denied serious consideration by the researcher.
If we hold in abeyance the deviance assumption, we note that the area
to be studied is an organized activity, established for the expressed
purpose of exchanging a “service” for remuneration. Viewed in
terms of process, it would be found that the structural properties
of the brothel are akin to servicing operations in general—a basic
social process in American society. Quite simply, the brothel exists
to provide a service(s), which happens to be sex. One property of a
servicing process in this particular context is that the service being
provided is generally considered deviant in the everyday world. The
“fact” that it is so conceived may have some consequences for the
organization of some of its publicly visible activities, such as making it
necessary to maintain a low profile, putting limits on public advertising,
necessitating payoffs to the police, etc.
However, the deviant conception of brothel activities is only one among
many conditions and properties in this and other servicing contexts.
Compared to other possible characteristics of the general process of
“servicing” such as power symmetry, role of expertise, specialized
knowledge, right of grievance, duration on premises, malpractice
problems, waiting properties, etc., the primacy afforded the role of
deviance in a unit analysis seems more reflective of common-sense
considerations than theoretical fit. Conceptualized from a process
orientation, the behavior of prostitutes and their customers has more
in common theoretically with behavior found in garages and beauty
parlors than it does with check forgery, alcoholism, and the vast array
of other instances ordinarily conceptualized as deviant behavior.
One further observation seems warranted. From our example
of brothel activities, it might be concluded that we have merely
transposed a hypothetical social psychological study into one
focusing on organization. We would answer that this is again a priority
characterization that is not reflected in the empirical world. Instead, in
our ongoing work with BSPs we have found one of its strengths to be
an ability to conjointly render both structural and social psychological
variables in terms of social process. It may be the case that either
structural or social psychological variation has primacy in a given
area, but that is a data-related question.
Regardless of the usual sociological interests, whether it be deviance,
religion, collective behavior, etc.; and, regardless of the usual primary
focus as either organizational or social psychological, the referent
for BSP theory is always the process itself and not the particular
substantive or conceptual unit involved. This does not mean that the
analyst will be unable to explain how the particular substantive unit
functions. Quite the contrary! BSP accounts of the world contribute
substantial insight into the practical realities of the day-to-day world by
explaining its variation (Glaser, 1969). However, as mentioned earlier,
the analytic focus seeks theoretical coverage and not descriptive
completeness, which is seen as impossible. As such, no claim is
being made that “servicing” is the only aspect of brothels of theoretical
importance. The only claim being advanced is that “servicing”
explains much of the variation to be found in the actions, interactions,
and perceptions found in the collected data from that research site.
The process illuminates organizational features about the brothel,
interactional patterns between prostitute and customer, prostitutes’
conceptions of their roles, and a wide variety of less obvious variables.
As such, “servicing” is not to be taken as a “theory” about brothels (or
deviance), but rather as a theoretical statement about processes that
occur therein, which occurs in other areas of social life as well.
This illustrates the consequences BSP sociology would have for the
manner in which sociology theoretically divides the empirical world.
BSPs as basic uniformities of social life, cut across the boundaries by
which sociology has traditionally been sub-divided. Thus, one of the
major ways in which we render the world sociologically should reflect
this basic uniformity.
Authors
Barney G. Glaser Ph.D., Hon. Ph.D.
The Grounded Theory Institute
P.O.Box 400
Mill Valley, CA 94942
USA
Judith A. Holton
10 Edinburgh Drive
Charlottetown, PE C1A 3E8
Canada
Correspondence:
Tel: 415 388 8431
Fax: 415 381 2254
Email: bglaser@speakeasy.net
Judith@islandtelecom.com
References
Beeson, Diane. (1973). Women in Karate: Neutralization of Sex Roles
(Master’s Thesis). San Francisco: University of California.
Bigus, Odis E. (1972). The Milkman and his Customer. A Cultivated
Relationship, Urban Life and Culture, July, pp. 131–165.
Glaser, Barney G. and Strauss, Anselm L. (1967). Awareness of
Dying. Chicago: Aldine Publishing Co.
Glaser, Barney G. (1969) Second Deeds of Trust: How to Make Money
Safely. Mill Valley, CA, Balboa Publishing Co.
Glaser, Barney G. and Strauss, Anselm L. (1971). Status Passage: A
Formal Theory. Chicago: Aldine Publishing Co.
Glaser, Barney G. and Strauss, Anselm L. (1968). Time For Dying.
Chicago: Aldine Publishing Co.
Hanson, Richard R. (1976). In Quest of Optimal Health: The Natural
Health Movement in the United States (Ph.D. Thesis). Davis,
CA: University of California.
Lindemann, Constance (1974). Birth Control and Unmarried Young
Women. New York: Springer Publishing Company.
Olesen, Virginia L. and Whittaker, Elvi W. (1968). The Silent Dialogue:
A Study in the Social Psychology of Professional Socialization.
San Francisco: Jossey-Bass Inc.
Strauss, Anselm L. (1969). Mirrors and Masks. San Francisco:
Sociology Press, 1969.
Weber, Max. (1947). Theory of Social and Economic Organization.
Glencoe, IL: Free Press.
Yabroff, Lawrence J. (1972). Faulting: Why They Don’t Hear(Master’s
Thesis). San Francisco: University of California.
