The Roots of Grounded Theory

[From a keynote presentation given to the 3 rd International Qualitative
Research Convention, Johor Bahru, Malaysia 23 rd August 2005]

Barney G Glaser, Ph.D.; Hon Ph.D.

I studied sociology at Stanford 1948 to 1952, which
was partially fine but limited in those days. But then I
knew I wanted to be a sociologist. Returning to the USA
from the army in 1955 to study sociology at Columbia
confirmed my goals. I bought the program 100% on doing
sociology as my life work. All I do is sociology in every
facet of life; work, recreation, family etc. My life is
sociology driven and directed.

Now let me give you a quote from Barton’s (1955,
p.246) article of Paul F. Lazarsfeld [PFL], “Analyzing the
logic of research operations to clarify concepts remained a
key to PFL’s life”. It has been the key to my life also.

“All is data” – that now sloganized tenet of Grounded
Theory [GT] – clearly came from PFL per Barton’s words.
Robert K. Merton’s [RKM] brief flicker of light – to admit
to emergence (see Barton, p. 255) – became the key to GT’s
theoretical stance.

The Four Dimensions of being a Sociologist

In buying the program 100%, I bought the four
dimensions of doing sociology – autonomy,originality,
contribution and the power of sociology. All dimensions are
interrelated; they became a part of my sociological identity
and led eventually to my originating GT. Now let’s consider
each of these dimensions of my training, how they affected
me, subsequently found their way into GT and how they
may serve as food for thought in your training.


PhD training is a training for autonomy. One becomes
the doctor, so to speak. One claims one’s own pacing.
One claims one’s own ideas and the connections between
them. One becomes the theorist and/or research author.
Therefore, one must stand on what one has said and

This puts a call on one’s seniors, on faculty and the
social structure of departments to allow the PhD candidate
to do his own thing, irrespective of faculty and supervisor
desires to have the candidate work on their ideas. It puts
a call on author idol worship of “grand theorists”; it puts a
call on theoretical capitalism; it puts a call on supervisor
control and ownership of the candidate’s work in favour of
giving him/her full freedom and license. It is a claim that
the candidate must stand for irrespective of senior or
supervisor obstruction and efforts to the contrary. Try it;
you will like it.

Please remember, I did my dissertation totally on my
own on secondary data from the survey research center at
the University of Michigan. It passed easily. My supervisor
Hans Zetterberg was delighted. PFL was overjoyed by the
core variable and the development of new method analytic
techniques. RKM was confounded since it cast grave doubt
on his famous paper; “Recognition in Science”. My
dissertation was published immediately, given the
recalcitrant forces of action. It was requested, not sold by
me – since I did not have a clue.

Throughout my whole training I resisted the efforts of
both PFL and RKM to co-opt me to work for them and those
who did were not very smart. I had no time for them
personally, just their ideas. It was clear in RKM’s writings
on the sociology of science that the key to creativity was to
study ideas with autonomous freedom in order to put them
together by seeing the connections at will, hopefully, for
maximum yield and creativity.

This PhD stance, of course, fed into my origination of
GT. GT gives total autonomy, by the nature of emergent
discovery and more PhD candidates can claim this
autonomy through GT than do at this point. More can than
do through other methods and consequent subservience to
supervisor demands based on social structural power. As
GT spreads through out the world, researchers are
discovering this autonomy in their own departments. They
are in demand for lectures and consulting. Supervisors
become humbled, often with delight.


PhD candidates are supposed to produce original
research with originality of ideas and methods. RKM clearly
delineated the composition of originality. He said quoting
the literature, “There is seldom an idea that hasn’t been
said before somewhere else. Originality comes with the
putting together of ideas into new connections.” Yes, I
studied the sociology of science and wrote the famous
paper on comparative failure in science. See my reader
(Glaser, 1993).

RKM implicitly put a call on relevance with the new
connections idea. I saw clearly that my research for a
dissertation was generating a whole new and very relevant
set of connections leading to a theory of recognition in
science that was relevant, worked and fit. Probably that is
why it was published so quickly. Virtually 75% of the
chapters were published as papers. Earned relevance
became a strong requirement of GT analysis.

This of course, fed into the origination of GT: It had to
be field-wide with fit, relevance and workability (explaining
what’s going on). GT provides new (valid) categories and a
theoretical interrelation between then based on theoretical
codes (Glaser, 2005). I was the originator of GT as a
discovery method; Anselm Strauss did not have a clue
about these ideas on emergence (Glaser, 1992).

PFL of course seeded me with four important
methodological beginnings. Firstly, the index formation
model based on accumulation and summing of indicators
from survey data to generate indexes or concepts is
fundamental in GT. GT is just a simple index formation,
inductive method based on using any type of data. That’s
all. Sorry qualitative researchers!

Second, PFL’s discovery of the interchangeability of
indicators used to generate concepts was major. No
matter what indicators were used in multiples of three, the
generated concept had the same relationship to other
concepts. Based on this, he confessed to me one day in
privacy, that crude indexes gave the same findings as
elegant, perfected indexes based on latent structure
analysis. So the latter was a waste and expensive.

With this notion I was off and running and further
developed the analytic techniques of consistency analysis
that I used with his elaboration analysis model and mine of
theoretical saturation. The interchangeability of indicators
and theoretical saturation subsequently became prime
ingredients of GT procedures for generating substantive
theory. These two procedures led to essential delimiting of
research content, data collection and time for generating
theory with completeness, depth and scope. It allowed
dissertations to go very quickly, rather than take the
laborious long time exhaustion always heard expressed by
candidates (Glaser, 1978).

Third, PFL missed this one. So near and yet so far. He
missed the constant comparative analysis approach
(Glaser, 1978, chapter 9). It is so simple. At the time, in
order to do a survey, a researcher from the Bureau for
Applied Social Research would go into the field to do
qualitative research on what to ask as questions in a
survey; that is, as indicators. They summed the indicators
with Likert scales into an index to get the concept. It never
occurred to them to systematically and carefully compare
the indicators’ meanings to generate conceptual properties
of the soon to become index or concept. The power of this
procedure to generate theory is phenomenal. What a
theoretical yield of discovery. What a miss! The constant
comparison technique became the influential analytic
procedure of GT to generate and discover theory.

Lastly, PFL showed clearly in the academic mind that
core variable analysis explained so much of what is going
on and resolved the main concern of the participants.
Lazarsfeld and Thielens (1958) proved the core variable
analysis model has great yield. I used it in my dissertation
with the recognition index and it literally opened up the
data to a plethora of findings about the quest and
consequences of recognition. I transferred the analytic
notion of core variable to qualitative data and did the book
on the core variable “awareness of dying” (Glaser &
Strauss, 1965). This book was a big hit. It became a
classic and was subsequently published in several
languages. Thus I made core variable analysis the key to
generating GT. The core variable, as you know, is the
category that all other categories and their properties are
related to, and by these relationships explain what is going
on to continually resolve a main concern (Glaser,1978,
chapters 4 and 6). Connecting these methodological ideas,
of course produces the originality emergent in the GT
method. All this was beyond Anselm Strauss because he
was an expert in qualitative analysis – which means mostly

Truth is stranger than fiction, yet fictions rule the world
as they are built into and are a part of vested social
structures. Thus socially structured vested fictions are a
functional requirement of formal organization and the social
organization of life. The effect of these fictions often leads
to a miss of what is really going on in a social arena. Thus
these fictions usually lead to preconceived professional
problems upon which to do research despite their non-existence.

I cannot count the number of PhD candidates, using
the GT method, that have called me to ask what to do
about researching for a professional problem that is just
not there. I always recount the story about RKM when he
had a large grant to study his theory of professions. He
hired six PhD candidates to do the research. None got
their degree since what they were to study, his theory of
professions, yielded only independent correlations [no
findings and no data]. What a tragedy!

To discover what is going on using GT is first to
discover the problem or main concern in a social area to
discover it conceptually, which is not necessarily in the
participant’s view. Preconceived problems seldom, if ever
work, unless fictions are needed; but not to worry, the
discovered problem will, in the end, relate back to the
professional one in some way. Again GT ensures originality
on this dimension: the problem. What one is supposed to
study does not often produce a study! Thus GT is exciting,
motivating and fast as discovery emerges. Product proof is
in the making. Discovering the problem is just the
beginning of the originality in GT generation; subsequent to
which is the discovery of new concepts and their
connections which are then modeled by a theoretical code
(amplifying causal looping, basic social process, typology,
continuum, etc.).

Could I have done originality or origination of GT
without my claiming my freedom and autonomy? No, I
would have just been a pleaser of seniors. PhD candidates
bring problems of autonomy to me constantly and I always
answer the same thing. “Get the degree and you will be
autonomous after if not before”. See chapter six in the
Grounded Theory Perspective (Glaser, 2003).


PhD candidates are in training to contribute to science.
They are supposed to innovate and to contribute to their
field. I bought this aspect of the program hook, line and
sinker immediately when entering the sociology
department at Columbia University. I have, of course, by
now succeeded in producing a series of about 20 books –
both monographs and methodology. I have published
many papers in peer review journals – too many to count.
Two of my books have been translated into four different
languages. As I have said, my dissertation was published
immediately. Discovery (1967) has sold thousands of
copies and still sells 39 years later. These publications
have fostered the use of GT; correctly or not. I started
Sociology Press to keep my books perpetually in print and
to satisfy demand since typically publishers drop books
after a few years. It also keeps me in touch with those
doing GT throughout the world.

Thus the mandate to publish and therefore to contribute
to the field at large worked in my case. Keep in
mind, however, that I discovered in my dissertation that
one achieves the most recognition not by publishing and
peer reviews, BUT by being subsequently referred to,
used and footnoted. The noted take the cake.

So the problem I confronted as a PhD candidate and
you do too is how to get into print as fast as possible.
Careers hang on it! Let me give you a few ideas.

First: you write papers for your Professor. Give me a
break! What difference does it make what he or she says
or how they grade it. For the autonomous PhD candidate,
this is too particularistic and holds little or no career
prospect – as yet. Better to send the paper to a journal for
peer review and possible publication. Peer review notes
will give the author knowledge of how the field will receive
his work and what needs to be done to improve it. These
are the true gatekeepers for a career in the “publish or
perish” academic world. Why wait? Readiness is in the
hands of peer review not a particularistic professor. Make
every paper count and send it out there for the “test”.
Submit, submit, submit! I required all my students to write
papers for submission, NOT for me. It takes a clear view of
one’s autonomy and originality to do this with hope of
success, but many are pleasantly surprised. Remember, in
the final analysis, you are being tested by the field, not
your professors with their immediate social structural

Also, if your lectures are good enough to take your
time, then take notes carefully, as you probably do, and
write them up into a paper to submit. See if the class is
worth it. I did it and was amazed at what others thought
anonymously of RKM’s role theory, which I wrote up and
submitted the American Sociological Review. It made me
realize how important it was to ground concepts
systematically from systematic research. I was told it was
reified gibberish, by whom, I do not know. And don’t worry
about intellectual capital; it is over-rated in the academic
professions where one gives to the field as much as it can
and will take. This is not heresy.

Also, two more grounded items to remember. First,
when circulating a working paper for comments, never put
on “citations only permitted with the permission of author”
or “no parts of this paper can be used without permission
of author”. Rather, say “when using parts of this paper
please give proper citation and help yourself”. Be delighted
if someone wants to quote you.

Second, there is no such thing as full coverage in GT,
there are no misses. It is what you do that you offer as
contribution – not what you did not do. Full coverage is
impossible. Thus any senior colleague who points out
misses is just wrong or off track, since he missed the fact
that a GT fosters flooding out in all directions with general
implications and research possibilities and new ideas. GT is
very stimulating to what is next, not what is missed.

Built into the GT rigorous procedures package is the
goal of ending the research – the generated theory – as a
publishable product. It is a carefully delineated set of
procedures for doing so. Otherwise why do it? And GT
produces contributions to the field. So many PhD students
using GT are being published; it amazes me. It is
practically a sure thing to see one’s originality as fostered
by GT reaching a wide public in whatever the field.


The power of GT is phenomenal. Sociology itself is
very powerful and GT, by discovering and conceptualizing
latent patterns, potentiates that power. It is the mandate
of the PhD candidate to use this power humanely, morally,
as often as possible. It was my mandate. I use it
everyday in every facet of my life, using GT studies I know
of and doing GT all the time by keeping notes.

GT potentiates the power of sociology through its
conceptual categories and their properties integrated into
theory to explain the continual resolving of what is going
on in an action area. The discovered categories have
earned relevance with tremendous grab and endurance.
They are remembered decades after their discovery.

This conceptual relevance provides high impact
dependent variables to explain and vary a theory. GT’s
discovered in vivo substantive categories have great
meaning to people reading and using GT. They fit the
action scene so that people can virtually see the GT in
action and application. GT has much general implication;
that is, one can see the application of a GT in many other
substantive quarters. And with conceptual modification
through constant comparison, one can use the GT in areas
different than that in which it was initially grounded. See
for instance, Wendy Guthrie’s mystiquing and pseudo
friending in her study of client control (Guthrie, 2000). She
found this in a veterinarian practice but it is seen

GT grab can be favourably compared with immaculate
conjecture of concepts generated by ‘grand old men’.
Deductive, immaculate conjecture, usually from logical
deduction, often has little or no power since it is not
grounded in data with earned relevance. Its use is
preconceived and doomed to little or no relevance unless
forced on the data. Witness again the professions study as
featured in Glaser (1998, chapter 6). In order to avoid the
miss of preconceived problems and concepts, PFL would
always suggest running all items against all items to
discover the patterns that emerged from multirelationships
and then write up the patterns.

GT conceptualization has tremendous grab. Its
endurance and power overwhelms the power of description.
Description is stale-dated soon after the research whereas
conception goes on forever. The grab of GT is also found in
its jargon. Many now give it lip service to justify and “OK”
otherwise ungrounded qualitative research. It is powerful
even at this rather “empty” level. As I have said it draws
people from all over the world, by its excitement of
discovery and its truth and its quest of appropriateness to
the task. This worldwide use indicates both its power and
adds to it. GT is powerful also in its ability to use all data
and in its procedural pacing which allows the flex time we
all need in PhD work. And, GT is powerful in its sure
approach to achieving by doing its lock-step procedures for
getting a research project finished. Finishing is necessary
and very fateful for the PhD candidate.


Barton, A. (1955), “The concept of property space in social
research” in, The Language of Social Research: A
Reader in the Methodology of Social Research,
Lazarsfeld & Morris Rosenberg(eds.), New York, pp.

Glaser, B.G. & Strauss A. L. (1965), Awareness of Dying,
Aldine, Chicago.

Glaser, B.G. & Strauss A. L. (1967), The Discovery of
Grounded Theory, Aldine, Chicago.

Glaser, B.G. (1978), Theoretical Sensitivity, Sociology
Press, Mill Valley CA

Glaser, B.G. (1992), Basics of Grounded Theory, Sociology
Press, Mill Valley CA

Glaser, B.G. (1993), Examples of Grounded Theory,
Sociology Press, Mill Valley CA

Glaser, B.G. (1998), Doing Grounded Theory, Sociology
Press, Mill Valley CA

Glaser, B.G. (2001), The Grounded Theory Perspective:
Conceptualization contrasted with description,
Sociology Press, Mill Valley CA

Glaser, B.G. (2003), The Grounded Theory Perspective II:
Description’s remodeling of grounded theory
methodology, Sociology Press, Mill Valley CA

Glaser, B.G. (2005), The Grounded Theory Perspective III:
Theoretical coding, Sociology Press, Mill Valley CA

Guthrie, W. (2000), Client Control, Unpublished PhD
Thesis, Department of Marketing, University of
Strathclyde Business School, Glasgow Scotland

Lazarsfeld, P.F. & Thielens, W. (1958), The Academic Mind,
The Free Press, Glencoe ILL