“Whither the Pragmatism?”
by Mark Wilcox, PhD
University of Durham Business School, Mill Hill Lane, Durham, DH1 3LB United Kingdom
email
: mark.wilcox@durham.ac.uk
telephone : +44 (0)191 –374 – 1429
(An earlier version of this paper was presented at the
8th Annual Deming Scholars Seminar at Fordham University, NY in
February 2002)
Abstract
The fact that Shewhart and Deming were
influenced by the work of C.I Lewis (1929) is undisputed Deming (1991) . How he
influenced their work is less clear.
Lewis’ work was arguably a radical contribution to the pragmatist
methods being developed by a few eminent philosophers in the 1920s. Despite their apparent obsession with
Lewis’ theory of knowledge, neither Shewhart nor Deming referred to
“pragmatism” in their writing. This
paper is the first step of an exploration to try to uncover “pragmatism” in Shewhart's work. It is analogous to an archaeological
investigation - so whither the pragmatism?
Introduction
This
paper is an exploration seeking to uncover elements of pragmatic theory in
Shewhart’s theory of statistical quality control. Indeed this is the first
stage of a bigger project. I am treating this as an archaeological site. Three important artefacts lay under the
surface of a field of study, called quality control and management. These are the works of Shewhart, Lewis and
Deming. On the surface, we have the
remnants of the late 20th century quality movement. Below the surface, we find the “conditioning
text” (Xu:1997) of Dr Shewhart and “1st order texts” of writers such
as Deming, Juran and Feiganbaum. It was
they who went on to develop and expand the scope of quality control and
management from Shewhart's epochal theories.
“A conditioning text” is a concept used to describe how a significant
piece of work marks the beginning of a new period in history as an epoch and
thus shapes the future development in a particular field of study. In this case, we are referring to the work
of Dr Shewhart as a conditioning text.
While Shewhart’s work is often recognised by writers on quality
management, as the root of the modern quality management ideas, he is not well
known outside this field of study. Neither is his work widely read outside the
specialist groups, who continue to follow the work of Deming for example.
It
is the conditioning text of Shewhart that will be uncovered here. I will examine his work in relation to
Lewis’ theory of knowledge (conceptualistic pragmatism). In particular, I will be examining the
notion of “management–by-prediction” as a source of pragmatic theory of
knowledge. It is quite well known how Shewhart became acquainted with Lewis’
work in the early 1930s. This event
happened shortly after he published his major work in 1931, but before the 1939
book edited by Deming. Lewis’ book was apparently difficult to understand. Both
read it several times in an attempt to digest the content (Deming:1991). We can only assume they thought it was
worthwhile. Many have since followed in
their tracks, and floundered through the esoteric text, like wading through
treacle.
Despite
the well-hackneyed folklore, few have ventured into the relationship between
Shewhart, Deming and Lewis’ work in any detail. This paper marks the beginning of a project, to try to understand
the relationships in the work of these three internationally recognised
scholars who appeared, unwittingly, to be around at the same time and whose
work, despite being over 70 years old has still to reach its full potential. It is only right and proper that their work
receives more attention than has been previously afforded.
First,
some background on how this paper has been constructed. A few days before I sat down to write this
paper, a small a poem appeared on one of the many cluttered notice boards in my
home. It read…
“The past is
history.
The future a
mystery.
And today is
a gift.
That is why
it is called the Present” Anon.
If
ever a poem were to encapsulate the theme of a paper, this one did, in so few
words. In due course, this paper will
focus on how knowledge of the past informs our interpretation of the
present. An interpretation of the
present – in turn - helps to predict the future. I will use this poem to summarise how statistical methods can be
used to predict future behaviour of a process, but also the important role
pragmatic methods may play in the process.
The
paper unfolds in four parts. Part 1 is
a brief summary of the pragmatists' work in the late 19th and early
20th century. Lewis’ distinctive
contribution to this body of knowledge forms part 2. In part 3, I attempt to show how pragmatic methods may have
influenced the later development of Shewhart’s original work. A conclusion forms part 4.
The Pragmatist Movement
In
this section I will sketch out a brief summary of the pragmatist movement,
prior to Lewis' contribution in 1929.
The so-called mainstream writers in the pragmatist movement were Charles
Peirce (1839- 1914), William James (1842-1910) and John Dewey (1859-1952) (c.f.
Chung-Ying:1969; Bird:1986; Harrison-Barbet:2001). Pierce was probably the first of more recent philosophers to take
issue with metaphysicians such as Bradley and Royce and their debates on the
nature of reality and truth for example. Peirce wrote two notable papers in Popular Science Monthly in 1877 and 1888
entitled “The Fixation of Belief” and “How to Make Our Ideas Clear”
(Bird,1986:4). Crucially, Pierce
defined pragmatism as a theory of meaning
not truth. He described the philosophical debates on truth as absurd, being
conducted in an abstract form by his contemporary metaphysicians. For Pierce, truth was a function of the practical based on doubts and beliefs found
in human enquiry. Equally, he
conjectured that truth had a public
characteristic as opposed to the subjectivity
of Cartesian dualism. Therefore, truth
was a manifestation of the practice of human inquiry with (publicly) shared
meanings. Indeed, his pragmatism, as a
theory of meaning, intended to be scientific, in that certain concepts and
phenomena could be observed under specified conditions and the results translated into practical effects.
At
least two dominant schools of philosophical thought had formed, with quite
opposing views on how to proceed with the nature of human inquiry. “Rationalists” maintained that ‘human mind’
and sensory perceptions were unreliable as the basis for rigorous debate. Their thesis was that knowledge was beyond
or outside the limits of the mind, hence it was a priori – before experience.
On the other hand, empiricists held a view that only way we could form
judgements on the nature of reality was through the senses. At the risk of grossly over-simplifying some
complex philosophical debates, we find the pragmatists pitched in-between these
two schools, trying to form radical new agenda. Again, an oversimplified account will have to suffice for the
brevity required for this paper. The
pragmatists were focussing on what was practical and instrumental in human
inquiry, rejecting dogma in the process.
Peirce’s
work was essentially utilitarian and focussing on desired consequences. He was also adopting an inductive – hypothesis logic in his method. As a relatively lone
voice at that time, Peirce was receiving quite severe criticism, and his work
failed to get much popular attention for at least another twenty years. The pragmatist movement in general was
subsequently accused of being anti-intellectual by other philosophers (see
Bird,1986).
William
James worked mainly as a psychologist.
He was referred to by Shewhart (1931) in this respect, but
interestingly, not for his work on pragmatic methods. James was a close friend of Peirce and indebted too him for his
advise in developing his own work.
However it should be noted that they also disagreed on how the notion of
pragmatism should be developed as a theory.
James' work can be found in a variety of published lectures and papers
(see James: 1913; & 1917/1935).
While there is not the time or the space to unpack the fullness of the
debates in this paper, it is worth noting several points that have a bearing on
the following arguments.
James
carefully laid out his arguments for pragmatic method by referring to Peirce at
the beginning of his paper, entitled What
Pragmatism Means (1907 – published in James, 1917/1935: pp.,198-217)
“Mr
Peirce, after pointing out that our beliefs are really rules for action, said
that, to develop a thought’s meaning, we need only determine what conduct it is
fitted to produce: that conduct is for its sole significance. And the tangible fact at the root of all our
thought-distinctions, however subtle, is that there is no one of them so fine
as to consist in anything but a possible difference in practice. To attain perfect clearness in our thoughts
of an object, then, we need only consider what conceivable effects of a
practical kind the object may involve- what sensations we are to expect from
it, and what reactions we must prepare.
Our conception of these effects, whether immediate or remote, is then
for us the whole of our conception of the object, so far as that conception has
positive significance at all.
This
is the principle of Pierce, the principle of pragmatism. It lay entirely unnoticed by anyone for
twenty years…” (p.200).
James,
a devoutly religious man, wrote several papers on religion, e.g. The Varieties of Religious Experience (1902)
and The Will to Believe (1897) whilst
simultaneously developing his pragmatic method. Here, we can see the evolution of the practical features of
pragmatism. Specifically, the notions
of truth and utility have a particular part to play. James interpreted “practical” to mean the particular import that
a belief has in the life of the
individual. Put succinctly…what practical difference does it make if an
idea is true or not? In science, a concept has to be verifiable to be true
– but verified ideas serve our need
to predict and cope with our
environment. We will see later how
Shewhart used the notion of belief in his work.
In
his interpretation of “practical”, James sought to demonstrate
how philosophical debates, pale into insignificance when concrete examples were
sought to elucidate or illuminate issues.
“There
can be no difference anywhere that
doesn’t make a difference elsewhere-
no difference in abstract truth that doesn’t express itself in a difference in
a concrete fact and in conduct consequent upon that fact, imposed on somebody,
somewhere, and somewhen” (op-cit:201).
James dismisses the ‘absolute’ nature of metaphysical debates,
suggesting, theories should become instruments, not answers to problems. And in a beautifully eloquent phrase … “We
don’t lie back on them (theories) we move forward, and on occasion make nature
over and over again by their aid.
Pragmatism unstiffens all our theories, limbers them up and sets each
one at work. Being nothing essentially
new, it harmonises with many ancient philosophical tendencies. It agrees with nominalism, for instance, in
all ways appealing to particulars; with utilitarianism in emphasising practical
aspects; with positivism in its disdain for verbal solutions, useless questions
and metaphysical abstractions.
All
these, you see, are anti-intellectualist tendencies. Against rationalism as a pretension and a
method pragmatism is fully armed and militant.
But at the outset, at least, it stands for no particular results. It has no dogmas, no doctrines save its
method” (ibid.:203).
Pursuing
the practical theme, James developed the notion of “cash-value”, which relates to the utility of the concept or theory
being employed. Concepts and theories
all have a relative cash-value for
helping the inquirer interpret the present..
To
briefly explain James ideas (and at the risk of doing a grave injustice to his work)
in this respect, he developed a notion of a “stream of consciousness”,
developed as part of his theory of radical empiricism and psychology. In essence, the stream of consciousness aims
to describe how sentient beings experience activities on a daily basis. He appeared to suggest a "Heracleitean
state of flux" whereby experience of reality, creates an awareness of
personal identity and a level of consciousness. The "here and now" are understood using concepts
developed from past experience. Past experiences
have a relative cash-value to the given situation – just as the rooms down the
corridor of the hotel. Past experience
provides a cash-value for interpreting a complex stream of consciousness.
James
suggested in his paper entitled “What is Pragmatism” that there was nothing new
in what the pragmatists were arguing, quoting eminent philosophers back to
Plato, Aristotle and Socrates. Not
surprisingly, Plato had developed similar arguments in his imaginary dialogue
with Theaetetus, for his Theory of Knowledge.
The Platonic dialogue deconstructed various theories of knowledge, and
at one stage Socrates was considering the Protagorean thesis of “man is the
measure of all things” and the "Heracleitean state of flux". Like James’ stream of consciousness, we see
a similar exposition from Socrates, but with some careful manoeuvring to avoid
the argument, that if everything is in a state of flux – then knowledge of
anything is impossible, which leads to sceptisism and solipsism.
“Socrates: Their doctrine that ‘being’ (so-called) and ‘becoming’ are
produced by motion, ‘not-being’ and perished by rest, is well supported by such
proofs as these: the hot or fire, which generates and controls all other
things, is by itself generated by movement and friction – both forms of
change…. And so with the condition of the soul. The soul acquires knowledge and is kept going and improved by
learning and practice, which are of the nature of movements (Plato;1934:36-7)
It
would appear that Plato was an early advocate for continual improvement!
In
sum, in this short introduction to the pragmatist movement, I have been
highlighting some of the important issues being addressed by these scholars
almost a century ago. The focus of
their work was the practical element of interpreting human experience, in an
attempt to interpret experience of reality and predict the future. By taking this a stage further, we were able
to see how this notion led to a stream of consciousness, and the Heracleitean
state of flux.
Examining
the world from this view point is to take a systems perspective of the
universe. From this systems
perspective, we have to try to understand what constitutes knowledge and
reality. Indeed the epistemological questions
are how do we know what we know from this view point? So, to provide more focus for the paper, we concentrate on Lewis'
contribution to pragmatism and pragmatic methods and his theory of knowledge.
Mind and The World Order: Outline of a Theory of Knowledge
In
this section of the paper I will focus on Lewis’ notion of a priori and how this influenced his form of pragmatism and
pragmatic method. I will then show how
his theory of a priori influenced the
interpretation of the present to predict the future. It has to be said that
Lewis’ work does not lend itself to being condensed. Every page appears to be an important element of his detailed
argument, treading carefully between the solipsist traps of empiricism, and the
'impractical' consequences of a priori
theories. Venturing forth with cautious
optimism, I will offer an overview of specific features of Lewis' treatise.
For
clarity and historical accuracy, Lewis was considered to be 'outside' the
mainstream pragmatists referred to above.
He did however recognise their earlier contribution, which spurred him
to develop his own particular brand of pragmatism, labelled “conceptualistic
pragmatism” (Lewis,1929;xi). Lewis’ work differed from the others in
several ways, but especially because of his doctrine of a priori truth. He adopted
a polemic argument against the so-called metaphysicians with this doctrine. In
chapter 1 he outlines his arguments slowly and meticulously, while
simultaneously attacking the metaphysicians along the way. Three theses are developed, and are worth
repeating in full, particularly as there is insufficient space to consider
these in detail in this paper.
“The construction here attempted turns principally
upon three theses: (1) A priori truth
is definitive in nature and rises exclusively from the analysis of
concepts. That reality may be delimited a
priori, is due, not to forms of intuition or categories which confine the
content of experience, but simply to the fact that whatever is dominated “real”
must be something discriminated in experience by criteria which are
antecedently determined. (2) While the delineation of concepts is a priori, the application of any
particular concept to particular given experience is hypothetical; the choice
of conceptual systems for such application is instrumental or pragmatic, and
empirical truth is never more than probable. (3) That experience in general is
such as to be capable of conceptual interpretation, requires no peculiar and
metaphysical assumption about the conformity of experience to the mind or its
categories; it could not conceivably be otherwise. If this last statement is a tautology, then at least it must be
true, and the assertion of a tautology is significant if it is supposed that it
can be significantly denied” (Lewis,1929: x-xi).
Let us try to use these three theses in relation to
the poem at the beginning of this paper. Put simply, Lewis argued that
philosophers had no special insight, enabling them to pose as prophets. Metaphysics had become a dumping ground for
abstract debates with a tenuous link to science. In contradistinction – suggested Lewis - progression in science,
is the sole activity of mind. Minds do
not invent new ideas in an abstract world, devoid of reality. Therefore, a pivotal feature of Lewis’ work
was the primacy given to mind and its relationship to, and integral part of
reality. To support, and win this
argument, he had to show flaws in the work of metaphysicians, while avoiding
the solipsist traps or dilemmas of extreme scepticism.
Lewis makes some interesting points while
deconstructing the work of metaphysicians: science – he argues – is not just
reporting facts – but significantly an activity that distinguishes speculative from reflective inquiry. His
argument is aimed at the speculative
naval gazing (metaphysics) which has little practical value. However, he supports reflective inquiry, which helps one move forward in practice. Not wanting to dismiss theory per-se, he insisted on its practical
value for interpreting the present.
What mattered, was, how theory was defined, generated and used.
The nature of his polemic against metaphysics
is manifest here…
“But that
philosophical legerdemain which, with only experience for its datum, would
condemn this experience to the status of appearance and disclose a reality more
edifying, is still with us … But at least we must observe that such metaphysics
turns away from one type of problem which is real and soluble to another which
may not… If metaphysicians try to work on the problem of defining reality –
paradoxically they will create something esoteric and transcendent of ordinary
experience. Any metaphysics which
portrays reality as something strangely unfamiliar or beyond the ordinary grasp
stamps itself as the thaumaturgy, and is false upon the face of it” (Lewis,
1929:pp8, 9).
In discussing the nature of reality, he
argues that the ordering of experience leads to categories that stand at the
side, above or before experience, and are therefore a priori. In contrast,
experience of reality can only be understood in terms of experience. The correct understanding of reality must be
in empirical clues. “If they are not
contained within that segment of experience which constitutes the phenomenon
itself, they must be discoverable by the conjunction with what precedes or
follows” (ibid:. 12-13).
Uniquely, mind is capable of interpreting and
judging if something is real or unreal.
It is the reflective nature of mind, drawing on past experience, to
interpret the present, that allows us to predict, with probability, the future
behaviour of some observed phenomena.
Mind also acts in a similar way when interpreting the canons of logic,
the distinction between good and evil in ethics, and the distinction between
real and unreal in experience. Classifying
experience into categories is an activity of mind – not metaphysics. Therefore, the way we interpret the present
is a judgement of what appears real and unreal to a mind. More critically, mind is capable of bringing
experience to itself to make interpretations and discriminations and to make an
evaluation. And here we see the sum of
his argument for the primacy of mind in his methodology.
“Thus philosophy is, so to speak, the mind’s
own study of itself in action; and the method of it is simply reflective. It seeks to formulate explicitly what from
the beginning is our own creation and possession” (ibid:18). Therefore the existence of a priori set of categories independent of
the mind is untenable in his methodology.
Having established the premise of his
treatise, the activities of a (philosophical) mind are examined in more
detail. We have to be critical as
opposed to descriptive. This is part of
the reflective function of the mind, which should be dialectical and Socratic
because the “truth” is a function of mind – it has just to be brought to the
fore more clearly.
Lewis tries to define his alternative
position on a priori, because his
methodology implies mind’s activities are reflective, discriminating and
evaluating. They therefore rely on an
interpretation of past experience to “prove” x, which he then argues, is an act of “persuasion”. Interpreting a priori for his methodology he compares it with the notion of a
purpose. The mind is capable of
generating a purpose, which is not necessarily drawn from experience, but must
take shape in experience. If a purpose has no use-value, then it will
soon disappear from mind’s focus.
Similarly “what is a priori
and of the mind is prior to the content of the given, yet in another sense not
altogether independent of experience in general” (ibid:.24). Therefore, mind is in control of the a priori – and by definition a function
of mind. An activity of mind must be to
choosen from its vast repertoire of past experience the criteria for
interpretation of the real.
“The content of experience cannot evaluate or
interpret itself. Nevertheless the
validity of such interpretation must reflect the character of experience in
general, and meet the pragmatic test of
value as a guide to action (ibid:.27) (emphasis added).
Thus, as Lewis concludes his first chapter we
have a firm basis of the arguments to follow in relation to his interpretation
of mind and a priori.
“The reflective method is pragmatic in the
same sense that it is empirical and analytic.
It supposes that the categories and principles which it seeks, must
already be implicit in human experience and human attitude. The significance of such fundamental
conceptions must always be practical because though and action are continuous, and because no other origin
of them can be plausible than an origin which reflects their bearing on
experience. Further, it claims for
philosophy itself the pragmatic sanction that reflection is but a further stretch
of that critical examination of our own constructions and interpretations by
which we free them from inconsistency and render them more useful” (ibid:.34)
(emphasis added).
In sum, we can have seen how Lewis has
attempted to dismiss the grand metaphysical theses, hinging on an independent, a priori notion of reality. He has made mind the pivotal part of his
argument and methodology for a pragmatic method of interpreting reality – hence
the title of his book. Mind is not a
purely descriptive organ that passively processes data – that would lead to a
deterministic view of mankind. The mind
is reflective, building up its own a
priori database from experience, to analyse the present and future
experiences. The mind is also
discriminating and evaluating, making judgements between what is real and
unreal in “world order”.
While Lewis does not elaborate too much on
notions such as the James' stream of consciousness, he does acknowledge that we
are in motion – life does not stand still – this is world order. The final chapter of the book, entitled Experience
and Order, covers this topic and will not be further considered here.
For the purpose of this paper, my object is
to tease out the notion of management-by-prediction as part of a pragmatic
theory of knowledge. We have already
seen how he suggests we should interpret the past, but how does one predict the
future from the present? Figure 1
attempts to show the complex interaction of the three notions of history,
present and future.
Figure 1 The HPF process of prediction.
Time
|
H1 |
P1 |
F1 |
|
|
|
|
|
|
H2 |
P2 |
F2 |
|
|
|
|
|
|
H3 |
P3 |
F3 |
|
|
|
|
|
|
H4 |
P4 |
F4 |
|
|
|
|
|
|
H5 |
P5 |
F5 |
Key: H=History; P=Present; F=Future
Figure 1 may help to understand Lewis’ riddle,
quoted by Shewhart (1939) “Knowing begins and ends in experience; but it does
not end in the experience in which it begins” (Lewis,1934, quoted in
Shewhart,1939:80). Starting with P1 an
interpretation takes place based on knowledge of concepts from the past in H1 (knowing begins and ends in experience). Interpretation of P1 leads to predictions of
the future using existing knowledge of concepts. The concomitance of P1 and H1 leads to a prediction for F1. Moving across Figure 1 in time sequence, P1
become H2 and F1 becomes P2 (but it does
not end in the experience in which it begins). P2 is then interpreted with concepts and knowledge, gained from
H2, to predict the future in F2. F2 now
becomes P3. P3 is interpreted using H3,
which was previously P2, and before that F1.
P3 is the predictor for F3 and so on – ad infinitum.
If we briefly reconsider Lewis’ riddle we can
see more clearly what he meant by following H3-P2 –F1, which all have the same
root, but their substantive form changes as the process moves on. A metamorphosis occurs, and knowledge of the
concepts grows with experience.
However, the HPF process of prediction, while attempting to understand
the process being described by Lewis, may appear deterministic at first
sight. However, this is clearly not the
case, because ‘we’ are ‘swimming’ in a stream of random data, which is our
experience of reality. We assimilate,
interpret and discriminate as we move on through this world order, with mind as
acting as our arbiter in the process.
Lewis acknowledges the random nature of
experience. How can we possibly know
what we do not know for instance?
Similarly, one thing does not follow naturally from another, which is
what would be required for us to be 'certain' about knowledge and predicting
the future. If the world was so
orderly, nothing would be worth learning, because we would not be able to do
anything about it. (Ibid:355)
To create a sense of order in the here and
now, Lewis uses a notion called an “instant mental reaction to experience”
(ibid:358) which constitutes an ‘island of knowledge’. This is an important feature of Lewis’
thesis in his attempt to avoid the solipsist traps and sceptisism. In short, we
have to recognise the limitations of instant
mental reactions to experience as they flow through minds. Minds are limited in what they can
assimilate at any one time. We may
never know what we do not know in the world order. But significantly, there are no absolutes, so we become pragmatic
in how we interpret the present, in an attempt to predict the future. This is not a compromise to the
metaphysicians position, this is the nature of reality per-se in a pragmatist theory of knowledge.
Finally, we must recognise an important distinction
in Lewis’ work, in relation to the a
priori. He distinguishes between
the analytic and interpretative nature of the mind, which is itself, an a priori truth about our ability to
perform these functions from the a priori
truths of logic and mathematics.
“The a
priori is knowable simply through the reflective and critical formulation
of our own principles of classification and interpretation … It is the a priori element of knowledge which is
the pragmatic, not the empirical” (Lewis, 1929:232, 266).
In contrast, logical and mathematical
solutions and interpretations are in one sense absolute in the metaphysical
sense, because they are consistent and adequate for the purpose. They are equally pragmatic in being useful
in the right conditions.
“On the one side we have the abstract
concepts themselves with their purely logical implications. The truth about these is absolute in the
fashion in which pure mathematics offers the typical illustrations. Such purely abstract a priori truth answers only to the criteria of consistency and
adequacy. It is absolute and
eternal. On the otherside, there is the
actual brute facts of given experience…It is between these two, in the choice
of conceptual systems for application and in assigning of sensuous denotation
to the abstract concept that there is a pragmatic element in truth and
knowledge. In this middle ground of
trial and error of expanding experience and the continual shift and
modification of conception in our effort to cope with it, the drama of human interpretation
and the control of nature is forever being played” (ibid:.272).
To briefly summarise this short section of
Lewis’ treatise we have seen how he wove his own distinctive path to maintain a
grip on certain features of a priori
(maths and logic). Mind was the sole
creator and arbiter of theory generated from experience. Thus he had two distinct a priori positions. This stance also set him against the pure
empiricists, who relied on the senses as the source of knowledge. Hence, in true pragmatist fashion, he trod a
'middle ground' relying on a thesis of practical value and expediency to
progress through random streams of information called the world order.
A
Search for Pragmatism in Shewhart’s Theory
We have to look quite closely and carefully to find
just how Lewis may have influenced Shewhart and Deming's work. Given that Shewhart had not read Lewis’ work
when he published his first book in 1931, there is a remarkable similarity in
the way they addressed their subject areas, with different purposes. Lewis was developing a radical theory of
knowledge in the field of philosophy.
Shewhart was addressing the problems of quality control as a research worker in a
firm, mass-producing telephone equipment.
Both were concerned with the notion of prediction as a pivotal part of
their theories. Lewis focussed on the
role of mind in interpreting reality, and what the conditions were to be able
to 'predict' into the future.
For Shewhart, the notion of
prediction was central to his thesis, because it lay at the root of his
definition of control…
“ a phenomenon will be said to be in control when, through the use of
past experience, we can predict, at least within limits, how the phenomenon may
be expected to vary in the future. Here
it is understood that the prediction within limits means that we can state, at
least approximately, the probability that the observed phenomenon will fall
within given limits… The specific problem that concerns us at the present
moment is the information of a scientific basis for prediction, taking into
account the element of chance, where for the purpose of our discussion, any unknown cause of a phenomenon will be
termed a chance cause” (Shewhart, 1931:6,7) (emphasis in original).
In Shewhart's thesis, being
in a state of statistical control was an operation involving an
“engineer”. Being able to predict was a
key feature of this activity. Five
steps were defined to maintain this “state” forming part of Shewhart’s theory
of knowledge. It provides the criterion
for a process of a state of statistical control and as such is an important
artefact on this site of investigation.
To see the importance of these five steps, we have to first understand
an issue that Shewhart had taken with existing modus-operandi in engineering and manufacturing.
Shewhart took issue with the
“exact” nature of engineering discourse and epistemologies. In a polemic aimed at engineers we find a
similar type of argument used by Lewis against the metaphysicians.
“It is conceivable that some
time man will have a knowledge of all the laws of nature so that he can predict the future quality of a
product. This might be considered a
goal for applied science, but indications today are that it is not a practical
one…In fact if we are to believe, as do many of the leaders of scientific
thought today, that possibly the only kind of objective constancy in this world
is of a statistical nature, then it follows that the complete realisation of
the sixth stage is not merely a long way off but impossible” (Shewhart,
1931:352-3) (emphasis added)
Thus, just as Lewis took
issue with the grand theories of the metaphysicians, Shewhart took issue with
engineers' deterministic and objective view of the world. In a similar vein he attempts to show the fallacies
with “industrial programs” (Shewhart
,1931:354).
Shewhart argued how these
procedures were flawed because the sources of variation appear in the last four
stages of the process, thus leading to 100% inspection.
“One could specify what he
wanted, someone else could take this specification as a guide and make the
thing, and an inspector or quality judge could measure the thing to see if it
met specifications. A beautifully
simple picture!
The whole picture, however,
is radically different just as soon as we admit that we have only a probable
science” (Shewhart,1939:44).
Shewhart’s theory is the
antithesis of this approach. The
application of statistical methods to mass production requires a revolution to
the “exact” thinking of engineering. The
notion of being able to predict the future performance of a process based on an
analysis of past performance indicates the fundamental difference to
engineering discourse and a remarkable similarity to pragmatic methods.
To demonstrate how he could
apply statistical methods to mass production, Shewhart used what he described
as the three steps of the “old methods” of production.
“The three steps constitute
a dynamic scientific process of acquiring knowledge. From this view point, it might be better to show them as a form
of spiral gradually approaching a circular path which would represent the
idealised case where no evidence is found in step III to indicate a need for
changing the specification (or scientific hypothesis) no matter how many times
we repeat the three steps. Mass
production in this way constitutes a continuing and self-corrective method for
making the most efficient use of raw and fabricated material” (Shewhart,
1939:45).
The cycle of improvement,
referred to here, was the forerunner of the PDSA Shewhart cycle. To make this theory and process work,
Shewhart had to address a wide range of issues covering topics related to
pragmatism with philosophy of science, logic and psychology.
Shewhart was adept at using
small figures and models to get his ideas across. A major part of his thesis in relation to prediction, and the
cycle of improvement, were the postulate of three components to knowledge,
evidence, prediction and degree of belief.
Chapter 3 of Shewhart (1939)
is worthy of much closer attention than can be afforded here. This is where he appears to use Lewis’ work
more than in other parts of this book.
If we make a comparison with the way his theory was developed in his
1931 book, we can see how the notion of control and prediction were mutually
dependent concepts, developed on the basis of reliable methods of measurement,
sampling and probability theory. These
concepts remain central to his thesis, but in the 1939 book, we may detect the
'pragmatic influence' of Lewis on his thinking.
Original data as evidence E Prediction
P
Degree of belief in prediction P based on evidence E
(Shewhart, 1939:86) Figure 11
Here, Shewhart made a
distinction to show how statisticians may prefer to use more emotive language
than scientific, while engineers mainly use the latter.
“Strange as it may seem in
the face of this situation the statistician sometimes rushes in to help the
scientist and engineer do a scientific job and forgets that a lot of his
professional lingo is more emotive than scientific witness for example, the
statisticians use of such phrases as statistical facts, confidence limit,
probable error, most probable value and best estimate, to mention only a few”
(p84).
This discussion predicates
Figure 11 above and is analysed in some detail. The three components of knowledge form the basis of Shewhart’s
epistemology. How we know what we know
in Shewhart’s thesis, draws on the combination of scientific and emotive
language. He acknowledges that there is
no quantitative method of measuring the degree of belief in evidence in his
model. This is where he has to be
pragmatic and draw on notions such as intuition. Using another model to try to explain the dynamic nature of
systems thinking, and how statistical methods fit, we can see the links with
Lewis’ riddle described above.
|
Previously observed |
Practically verifiable |
Only theoretically verifiable |
|
X1, X2…X1,…Xn |
Xn+1, Xn+2
…Xn+f |
Xn+f+1,
Xn+f+2 … |
|
|
Future |
|
Present
(Shewhart, 193:133)
In explaining this model,
Shewhart makes another link to Lewis, using a short quotation. This has since gained some currency since it
(re)appeared in Deming (1993).
First Lewis (1929) ..”if
there is to be any knowledge at all, some knowledge must be a priori” (quoted in Shewhart
1939:88). Shewhart’s own phrase to make
a similar point was “What has just been said illustrates the generally accepted
conclusion that we can not have facts
without some theory” (op-cit.) (emphasis in original).
This relates to a crucial
and very sensitive issue in this type of epistemology. How mind interacts with our experience of
reality is the subject of numerous epistemological debates and theories. Here, we can see how Shewhart took Lewis’
premise as part of his own theory. However,
we have to dig deep into the text to establish what theory Shewhart may have
had in relation to mind. In the 1931
book, the extended bibliography refers readers to texts on philosophy, logic
and psychology for example. Indeed,
Shewhart was quite eclectic in his reading, demonstrating a scholarly approach
to constructing a new theory.
“Philosophy for hypotheses
on the nature of reality, functions of laws, theories and causal explanations,
into the field of logic because it represents what we know about formal methods
available in the theory of deduction and induction. How do data depend on the mind?
What are the effect of factual experience and the effect of reasoning
upon the observer? Perhaps more
important, however, is that the mental experience involving reasoning
influences to a marked extent what we serve” (p.482).
Apparently, Shewhart did not
pursue the issue of mind in any detail at this point. However, if we study the text carefully we can find some
illuminations on the role of the engineer.
In Shewhart’s theory the ubiquitous engineer and statistician play
pivotal roles. Engineers have to be in
charge of designing products and processes, as well as being trained in
statistical methods. Take this
following passage as an example.
“The operation of control is
in this sense a dynamic process involving
a chain of actions, whereas the criterion of control is simply a tool used in
this process. The successful quality
engineer, like the successful research worker, is not a pure reason machine but instead is a biological unit reacting to and acting upon an ever changing
environment” (Shewhart, 1939:38) (emphasis added).
Shewhart does not appear to
develop or define the notion of biological unit, but it does indicate a close
relationship with the product and processes, suggesting, mind engages in an
activity of continual improvements.
Part of this process involves the notion of management by
prediction. Shewhart suggests there are
at least three types of prediction to be considered, referring back to one of
his previous examples.
“…the first, the type of
prediction involved in a student range …. the second, the type involved in the
tolerance range…. The third type of prediction, P3 is that involved in estimation and will be considered
first since it is the one with which the statistician is most familiar….
Obviously all scientific predictions must have definite meanings and we shall
accordingly choose the following.
Criterion of meaning: Every sentence in order to have definite scientific
meaning must be practically or at least theoretically verifiable as either true
or false upon the basis of experimental measurements either practically or
theoretically obtainable by carrying out a definite and previously specified
operation in the future. The meaning of
such a sentence is the method of justification” (Shewhart, 1939:93-94).
The three types of
prediction are built into a figure to show how data is collected and used for
prediction, and by implication control (Shewhart, 1939:89) (see Fig 22, ibid:
102 for the complete figure involving prediction).
X1,
H1
C1
Where X is the data being gathered, H
is the person collecting the data, and C
is the conditions. The concomitance
of the three signs in the figure represents the original data being collected,
to be used in various forms of prediction.
From this we consider the final point from Figure 11, which is the
notion of belief. The notion of belief
is duly qualified in Postulate II :
“In what follows we need to
keep clearly in mind that the statement that the quality of product is in a
state of statistical control involves a prediction P which may or may not be true, and it involves the evidence E for believing in the prediction. The
statement itself is a probable inference.
I shall assume the basic.
Postulate II. The objective degree of
rational belief p$’ in an inference involving a
prediction P based upon evidence E is not an intrinsic property like a truth but inheres in the inference
through some relation of the prediction P
to the evidence E” (Shewhart,
1939:42) (emphasis in original)
Shewhart now combines
several concepts to explain the process being discussed. A combination of mental, physical and
mathematical concepts to assist in making predictions based on evidence
provided at any one time. The
prediction is not a truth, but a
probable inference. The notion of
prediction is complicated, and worthy of more space than can be provided
here. But we should note that the state
of statistical control was defined as a state,
an operation, and a judgement. The
judge of quality had to be familiar with all the consequences associated with
Figure 11 above, whilst recognising the rules of probable inference and rules
of evidence (ibid:43).
The final point left to
raise here, is the role of Shewhart’s control charts. The control charts have been described as the voice of the
process, or the process talking to us (Burr quoted in Deming:1986). So how do they reflect the pragmatic methods
being searched for? The control charts
display the information by which the engineer can interpret whether the process
is in or out of control. Random and non-random
data indicates whether the processes are stable or unstable. Unstable processes, display non-random
patterns of variation on the control chart, making it impossible to predict the
future performance of the process.
Conversely, stable processes display random patterns of variation, and
therefore may be used to predict with a degree of certainty the future
behaviour of the process.
It follows therefore that
control charts are inherently pragmatic in the way they convey information from
the process. The information allows the management and control of the future
performance of the process. An
indicator of non-random data on a control chart requires the engineer to
interpret the signal on the basis of past experience and knowledge. Control charts ‘tell’ part of the story,
which the engineer has to interpret as a “biological unit”. The engineer uses theory and knowledge
gained from past experience to interpret the signal. The small dot on a control chart is the ultimate symbol for
management by prediction, it is a sign sending a message to the engineer’s
mind. Is it a signal or noise? What is the process saying? The interpretation of that small dot, or
sign and any decisions to act on it, will be taken on an “economic” basis. The economic basis is sufficient to
interpret it as a pragmatic decision.
Economic theory provides a basis and justification for the rationale
behind the decision. Economic theory is
the means to interpret the data and predict whether any actions need to be
taken.
Conclusion
The aim of this paper was to
prepare the ground for uncovering elements of pragmatism in Shewhart’s
work. Shewhart and Deming were both
apparently influenced by Lewis, but for some reason they did not use a
'pragmatic' discourse in their own writing.
This investigation on the surface of this archaeological site has
revealed similarities between Shewhart and Lewis’ theses. Indeed, we found strong pragmatist
tendencies in Shewhart’s 1931 work, written before he read Lewis.
The application of
statistical theory to mass production meant observing the production process as
a system. Systems are dynamic and
theories of knowledge pertaining to systems thinking may be problematic,
because every thing appears to be in a state of flux. This raised the problem of how sentient beings react with, or act
upon a system that is in a state of flux?
We have seen a degree of semantic manipulation by philosophers wrestling
with this age-old problem. Shewhart
appears to have addressed this in a unique manner with the application of
statistical methods and the development of control charts. The control charts
display variation as data gathered from the process. Data has to be interpreted by the engineer, and thus a pragmatic
activity takes place, because of the requirement to be economic. The state of
statistical control is an ideal, but probably unattainable goal, and thus the
decision to act is not final, but a step along a never ending path. A ‘justifiable’ decision made to day, on an
interpretation of evidence, need not be ‘justifiable’ tomorrow because the
conditions will have changed.
If we return to the poem
used at the beginning of this paper, we can see how Lewis and Shewhart
attempted to address the riddle. Lewis,
from the basis of a philosophy driven by practical results. Shewhart, from the basis of trying to solve
a quality problem in a high-volume, mass-production industrial system. Both addressed the problems with defining
and using a priori logic - Lewis more
than Shewhart. Both addressed the
problems of measuring the present - Shewhart probably more than Lewis in sheer
practical terms. And they both
addressed the problems of prediction.
This paper has only scratched the surface and prepared the site for
further, more rigorous investigation.
Future work will involve
examining the closer relationship between sampling, probability and statistical
theories and laws with the precepts of pragmatic methods. It is clear, there is a great deal of
overlap, but one wonders where this begins and ends.
However, Shewhart had
developed the major tenets of his work before he read Lewis’ work, and we know
from his bibliography, that several other philosophers influenced him. For the time being, it appears plausible to
suggest that Shewhart and Deming used Lewis’ work as a form of post-hoc rationalisation. So the work goes on…the site remains open.
References:
Bird, G. (1986) William James, Routledge Kegan Paul, London.
Chung-Ying, Cheng., (1969) Pierce’s and Lewis’s Theories of Induction, Martinus Nijhoff: The Hague.
Deming, W.E., (1986) Out of The Crisis, MIT Press, Cambridge, Massachusetts.
Deming, W.E. (1991) “A Tribute to Walter A. Shewhart On the Occasion of the 100th Anniversary of His Birth” SPC INK Newsletter Winter 1991. SPC Press, Inc.
Deming, W.E., (1993) (2nd edition) The New Economics, MIT Press, Cambridge, Massachusetts.
Harrison-Barbet, A., (2001) (2nd edition) Mastering Philosophy, Palgrave, Basingstoke.
James, W., (1913) Talks to Teachers on Psychology: and to students on some of life’s ideals. Longmans Green and Co, London.
James, W., (1917/1935) Selected Papers on Philosophy, Dent and sons, London
Lewis, C.I., (1929/1956) Mind and The World Order: Outline of a Theory of Knowledge. Dover publications: New York.
Plato, (1934)( Edited by Cornford, F.M) Plato’s Theory of Knowledge: The Theaetetus and the Sophist translated with a running commentary. Routledge &Keegan Paul, London.
Shewhart, W.A., (1931/1980) Economic Control of Quality of Manufactured Product ASQC: Milwaukee.
Shewhart, W.A., Deming, W.E., (ed) (1939/1986) Statistical Method from The Viewpoint of Quality Control. Dover Publications: New York.
Xu, Qi., (1997) The Making of Total Quality Management (TQM): A supplementary examination. PhD Thesis University of Durham.