Policy Sciences 19:33-59 (1986) 33

9 Martinus Nijhoff Publishers, Dordrecht - Printed in the Netherlands

Between knowledge and politics: Three faces of policy analysis

DOUGLAS TORGERSON
Administrative and Policy Studies Program, Trent University, Peterborough, Ontario KgJ
7B8, Canada

Abstract. Various conceptual schemes have been employed to make sense of the diverse policy
literature. Attempting to understand policy analysis in terms of its political and historical sig-
nificance, this essay points to three distinct "faces," distinguished with regard to differing relation-
ships between knowledge and politcs: one where knowledge purpo~s to replace politics, one where
politics masquerades as knowledge, and one where knowledge and politics attain a measure of
reconciliation. Historically, these three faces may be viewed, to an extent, as periods in the develop-
ment of policy analysis: from positivism, to its critique, to present post-positivist efforts.

"...it is not easy to imagine a world free o f politics..."

Harold D. Lasswell

Public policy analysis often appears as a jungle o f diverse and conflicting

modes o f inquiry, full o f inconsistent terminologies, divergent intellectual
styles - perhaps, indeed, incommensurable paradigms. There have been vari-
ous attempts to survey the field and to determine what, if anything, might lend
some coherence to the apparent chaos. Often, however, these attempts b e c o m e
so engrossed with details and nuances in the varieties o f policy analysis that
we end up with a m a p which is almost as tangled and confusing as the jungle
itself. 1
In attempting to make sense o f policy analysis, this essay takes the risk o f
erring in the opposite direction - o f so simplifying the m a p that it bears little
relation to the actual terrain. Different maps, however, have different purposes;
and they simplify or exaggerate accordingly. The p r i m a r y purpose here is to
understand policy analysis in terms o f its political and historical significance. 2
F r o m this point o f view, we m a y begin by disregarding all the undeniable
differences within policy analysis and attempting to identify a single p h e n o m e -
non. Broadly conceived, then, policy analysis m a y be said to comprise those
activities aimed at developing knowledge relevant to the formulation and im-
plementation o f public policy.3
As we attempt to understand the political and historical significance o f this
p h e n o m e n o n , we o f course quickly c o m e to see that it has no uniform appear-
ance, but projects distinctly different'aspects, or "faces." It is in this sense that
it is possible to identify three faces o f policy analysis. As we consider these
34

three faces in turn, we shall see in each a different relationship between knowl-
edge and politics - and, moreover, a different assessment of reason and its
place in public affairs. 4
This essay might be viewed as a contribution to the contextual orientation
of policy analysis (cf. Torgerson, 1985). The importance of such contextual
orientation was repeatedly stressed in the work of Harold D. Lasswell (e.g.,
1971: Ch. 2), who called in particular for the elaboration of "developmental
constructs": tentative - indeed, frankly speculative - models which would
serve to locate the analyst and his or her work in an historical framework.'In-
deed, there is an historical dimension in the appearance of the three faces of
policy analysis: from the heyday of social science positivism, to a later critique
and disillusionment, to current efforts to develop a "post-positivist" direction.
Here the three faces are somewhat artificially arrayed - one, two, three -
in an historical sequence, each new face emerging from past developments. It
might even be tempting to speak of distinct, progressive stages of policy analy-
sis. But history is not so neat. We do glimpse in the third face the potential
for a more adequate relationship between knowledge and politics, but there
have been earlier anticipations both in theory and practice. Despite everything,
moreover, the first face still haunts the field, and this is where we find that rea-
son, so to speak, begins to betray itself.

The first face

Images from an old dream tend to dominate professional thinking about pub-
lic policy. It is a dream of the abolition of politics - of putting an end to the
strife and confusion of human society in favor of an orderly administration
of things based upon objective knowledge. This dream of reason in public af-
fairs was prevalent in the Enlightenment of the 18th century and was reasserted
with the advent of positivism in the 19th century. The early positivists an-
nounced the dawn of a new age of humanity: a smooth, efficient industrial
civilization, established and managed not by the dictates of political interests,
but by the dictates of geniune knowledge: the findings of the modern natural
and social sciences. 5
The Positive spirit would rigorously distance itself from the speculations of
theology and metaphysics, confronting the world objectively in order to ob-
serve the facts and determine the lawful order of nature and society. The do-
main of mystery and ambiguity would be abandoned in order to know what
could be known clearly and certainly. For it was through providing precise and
reliable knowledge of lawful regularities that science could promote the pro-
gress of human civilization through the early turmoil of industrialization to-
ward the ideal of a mature and harmonious industrial order. Knowledge would
replace politics.
 35

The ideal of replacing politics with knowledge remains, implicitly or ex-

plicitly, as a predominant theme in much of contemporary policy analysis (cf.
Schick, 1969, 1971). Indeed, one gets the sense that professional policy analysis
is not really of this world - this all-too-human world of conflict, confusion,
and doubt. Consider the conventional posture of the policy analyst. The guid-
ing outlook is a technocratic one - that is, one of detachment in developing
knowledge of society in order that the knowledge can subsequently be applied
to society. The place of the analyst in society as a social being dealing with
others of his kind tends not to be raised as a significant matter for discussion.
Removed from society, social science produces the knowledge from which to
fashion an effective social technology; and the analyst - as both scientist and
technician - becomes one who performs remote operations on an essentially
alien object. 6
Policy analysis today bears the unmistakable imprint of the positivist heri-
tage. First articulated in the 19th century, the spirit of positivism was reasserted
in the early 20th century by the crusade for scientific management in industry
and government (see Haber, 1964; Waldo, 1948). The most immediate in-
fluence on policy analysis, however, has perhaps been the impact on its
methodological orientation of logical positivism. As a movement in philoso-
phy, logical positivism was especially significant in the period before the Sec-
ond World War. Still, the implications of this movement for the conduct of the
social sciences became particularly important later, as there were determined
efforts throughout the 1950's and 1960's to cast the social sciences in the mold
of the positivistically conceived natural sciences.7
The relationship between methodological theory and actual research prac-
tice is a complex, rather than a simple and direct, one (cf. Gunnell, 1975). In
the case of policy analysis, the influence of positivism has been pervasive not
only in letter, but also in spirit. That is to say, policy analysis has been in-
fluenced not only by the rather cautious and relatively recent doctrines of logi-
cal positivism, but also by the exhuberant industrialism promoted by the posi-
tivism of the 19th century.
Something of the influence of positivism, both old and new, upon the de-
velopment of policy analysis has been captured by Richard French, who argues
that the positivist and technocratic orientation of the social sciences reached
its peak in the 1960's:

There was a belief on the part of social scientists that their disciplines were
approaching the kind of certainty that characterizes the natural sciences,
and hence that their results could be implemented within public programs
with the same kind of confidence that attends the use of a "hard" technolo-
gy (1984: 19).
36

This confidence in the knowledge to be gained from the policy oriented social
sciences was accompanied by a distinctly positivist conception of the nature
of knowledge itself. All real knowledge, in this view, was scientific knowledge
- that is, restricted to the observation of facts, to logical inference, and to the
determination of regular relationships among facts, s With this kind of knowl-
edge, events could be explained in terms of antecedent conditions; and, con-
versely, future events could be reliably predicted. Clearly, such knowledge
would allow for the control of future events and could thus be employed in the
development of a reliable social technology (cf. Fay, 1975: Ch. 2).
What this knowledge could not provide was any guidance in determining
which goals this social technology should be employed to achieve. Indeed, even
to raise this question would be to expect too much of knowledge. Goals were
a matter of values, not facts. The scope of knowledge was restricted to logical
and factual statements and could not be expanded to statements of value. In
other words, reason had no place in the determination of norms - in helping
human beings to decide what ought to be and what they ought to do.9
Naturally, this restriction on the scope of knowledge created certain tensions
for those who embarked upon the development of a social science designed to
be applied to the guidance of public policy. But these tensions could be felt
only by those who took seriously the sharp dichotomy which logical positivism
introduced between reason and commitment. Perhaps somewhat reluctantly,
they acknowledged that policy analysis Could be used in principle by anyone
for any purpose whatsoever. There was no normative principle intrinsic to in-
quiry which could guide its use. Whether to commit policy analysis to the
cause of tyranny or democracy was, for example, a question for which reason
had no answer: it was a matter ultimately of what values one happened to be
committed to. 10
Early positivism, however, had not been bothered by such considerations.
For in the view of the early positivists, the development of modern science,
technology, and industry was part of the objective order of things: they arose
as part of the predetermined path of human progress. Although thus predicat-
ed on an incredibly comprehensive knowledge of past and future, this belief
apparently resolved for early positivism the question of its commitment. 11 In
practice, policy analysis has tended implicitly to accept this solution also. Yet
there is a striking difference between the commitment of early positivism and
that of contemporary policy analysis. Positivism saw itself as part of the de-
velopment of world history, devoted to promoting the realization of a mature
and orderly industrial civilization. Making its appearance when industrialism
was in ascendance policy analysis could - in contrast - take the existing or-
der of things for granted, and could largely ignore the problem of commitment
by limiting itself to the narrower focus of making "minor adjustments in exist-
ing mechanisms" (Easton, 1971: 81). 12
 37

The second face

The first face of policy analysis, then, is the face o f Enlightenment. Policy
analysis appears as the necessary adjunct to the Enlightenment vision o f a ra-
tional civilization, reasserted by positivism as a vision of industrial order and
progress. However, there is another face o f policy analysis, just as there is a
somber side to the Enlightenment vision. The utopian prospect of a bright and
efficient future has especially been countered by the dark forebodings of the
dysutopian novel. 13 In particular, it is Zamyatin's We which perhaps has most
effectively evoked the horror of a "clear, rational world" with a "regular, pre-
cise, mechanical rhythmn" and a "mathematically infallible happiness"
(Zamyatin, 1972: 77, 83, 1). 14 This book and ones like it - such as Huxley's
B r a v e N e w W o r l d and Orwell's 1984 - draw sharply into question the enlight-
ened hopes of positive thinking. Indirectly, these works also challenge the ra-
tionale of conventional policy analysis.
The first face of policy analysis is the one we see when we are uncritically
enamored o f the old Enlightenment dream and its subsequent revisions. The
second face o f policy analysis makes its appearance as the old dream turns into
something o f a nightmare. No longer can policy analysis be accepted without
question as an adjunct to the necessary unfolding o f a benign and objective
pattern of development - of the orderly progress of a rational civilization. If
policy analysis first appeared as the victory of knowledge over politics, it now
appears as the domination of politics over knowledge.
Let us consider the testimony o f Douglas G. Hartle. In the early 1970's, Har-
tie was a major proponent of what, in retrospect, he now calls "technocratic"
policy analysis.15 Hartle underwent a dramatic and remarkable transforma-
tion, and since 1976 he has published a series o f works highly critical o f his
previous orientation. He provides a striking example o f a person who was ini-
tially fixated on the first face of policy analysis - and then was shocked and
disillusioned when, in the midst of political realities, he directly encountered
the second face of policy analysis. Listen to his conclusion: "changes i n . . .
techniques and processes.., should be considered as a continuous battle of the
'war o f all against all' ":

Those who advocate changes in techniques and processes are for the most
part, consciously or unconsciously, guns for hire; they seek changes in the
existing power structure. They are responding, sometimes without aware-
ness, to the incentive systems that apply to themselves and the groups they
serve (Hartle, 1976a: 24).

What Hartle came to realize, in other words, was that conventional policy anal-
ysis, with its technocratic orientation, was blinded to political reality - that
38

it failed fundamentally to understand the nature of its context: ~o "the advo-

cates of these new technocratic solutions talk and write as if they are trying
to persuade an all-powerful, all-wise dictator. Politics, human nature and in-
stitutional structures are essentially ignored" (1978: 95). Hartle's critique is
particularly intriguing because he directs it explicitly not only at others, but
also at himself.
Political neutrality now can be grasped as an illusion which tends to suppress
critical questions about the political context in which policy analysis is applied
(cf. Gunnell, 1976: 93). In a critique of prevailing professional attitudes, Henry
S. Kariel has clearly revealed the hidden political dimension of conventional
policy analysis:

Attuned to a settled, predefined political reality, trusting it, policy science

remains effortlessly on the side of the accredited interests. Its practitioners,
themselves not disposed to interfere with established definitions of reality,
perceiving such transactions to be unscientific, readily make the prevailing
problems of society those of their science. It becomes their proper task to
search for the type of knowledge which can be used to govern effectively
within established political frameworks, to integrate men in established sys-
t e m s . . . (1972: 106).

Of course, the "abuse" of policy research is a well-known phenomenon (cf.

Beneviste, 1977; Horowitz, 1970) - such an open secret, indeed, that we find
this extraordinary admission in a letter to a scholarly journal:

The company I was with developed a concept - "Sell it to the guy who's
buying." We built this study to fit his brain capacity and also to agree with
his political bias, and frankly we would not have stayed in business other-
wise (Kennedy, 1978: 383).

However, as Kariel suggests, the hidden politics of policy analysis goes deeper
than the problem of such flagrant abuse. While claiming an allegiance to rea-
son, conventional policy analysis not only serves particular interests, but also
reinforces the order and ideology of the established political world.
In the context of the modern, administrative state, policy analysis tends to
project a technocratic aura, reaffirming the old positivist notion that science
somehow will deliver the precise and certain solutions we need to maintain the
progress of the existing order. Even though significant, intrinsic limitations to
conventional policy analysis are now widely recognized, 17 such technocratic
expectations persist. How is this to be explained? Let us consider the answer
of Laurence Tribe, who has offered what is still perhaps the most important
critique of conventional policy analysis:
 39

[A]nalysis is often intended not only to aid the decision-maker in choosing

a course of action, but also to help in persuading others of the justifiability
and wisdom of his choice. The usefulness of analysis in such advocacy is
drastically reduced whenever it does not at least appear to point objectively
and unambiguously toward a particular alternative. Thus, the users of
policy-analytic techniques are under constant pressure to reduce the many
dimensions of each problem to some common measure in terms of which
"objective" comparison seems possible - even when this means squeezing
out "soft" but crucial information merely because it seems difficult to quan-
tify or otherwise render commensurable with the "hard" data in the prob-
lem (1972-73: 627).

Here Tribe points not so much to the abuse of policy analysis by political in-
terests as he does to the pervasive yet subtle influence on analysis of the nature
of politics itself. The allegiance to reason suggested by the apparent objectivity
turns into an unconscious betrayal: technocratic style and imagery become
part of political rhetoric; irrational claims and expectations are advanced un-
der the banner of reason. ~8 Politics wears the mask of knowledge. The old
dream that reason could escape politics now appears as the political naivete
of conventional policy analysis (cf. Baum, 1982; Torgerson, 1984: Ch. 6).

The third face

The third face of policy analysis can today still be seen only dimly, if at all.
Here something of the old dream of the Enlightenment remains, but has been
decisively altered. The commitment to reason in public affairs remains, but the
meaning of reason has changed along with the meaning of politics. In the first
face of policy analysis, knowledge replaced politics; in the second, politics
overwhelmed knowledge. In looking at the third face of policy analysis, we
glimpse the potential for a relationship in which politics and knowledge are no
longer deadly antagonists.
It is the critique of conventional policy analysis which exposes the second
face, showing the dark side of the Enlightenment vision: an irrational tech-
nocracy, in which the apparent escape from politics is but a mask for the sur-
reptitious exercise of power. This very critique is largely predicated on the
emergence of a new direction in the philosophy of the social sciences, which
has been articulated in Richard J. Bernstein's well-known book, The Restructur-
ing of Social and Political Theory (1976). In the policy literature, French
(among others) has referred to the book and has drawn a parallel between this
philosophical development and changing views of policy analysis:
40

Even as philosophers are discovering the inadequacy of the positivist view

of the social sciences - that is, admitting that the social sciences are far in-
deed from the certainties of the natural sciences - policy analysts and
government decision-makers are ruefully recognizing the technocratic
hubris... (1984: 20).

It is a short step to draw the conclusion that the widely noted post-positivist
turn in the philosophy of the social sciences has profound implications for the
conduct of policy analysis. Indeed, a number of recent works have begun to
explore this idea. 19
Once the second face of policy analysis is recognized, the spell of positivism
is broken. It becomes apparent that the narrow, positivist conception of reason
has fostered an intellectual style which is insensitive to its own nature and con-
text - which is, in a word, irrational. 2o Yet we do not necessarily reach a state
of utter disillusionment; the critique of conventional policy analysis does not
end with the debunking of a myth. For the critique is itself inspired by the ideal
of rational inquiry - of an open, uninhibited search for knowledge. The cri-
tique thus brings us to a key question: can there be a place for reason in the
real, political world in which public policy is made and implemented?
To approach this question, we need to consider what has been called the
"politics of methodology" (Amy, 1984: 210). Positivism encourages an image
of the neutrality of research because it divorces the analyst from the object of
analysis; the researcher becomes, in principle, a detached, neutral observer of
facts. The neutral analyst, who gains knowledge o f society, readily fills the role,
moreover, of the neutral technician, who applies this knowledge to society.
Whether or not to include such policy analysis within political life is a political
question, however, involving the manner in which collective decisions are to be
made~ 21 Such policy analysis presupposes a choice, implicit or explicit, of a
particular form of political relationships, one which divides experts and
citizens.
It should further be recognized that the adoption of any methodological
posture - whether right or wrong - is inescapably a form of political action.
The gulf between the expert and the citizen thus appears not as politically neu-
tral, but as an artifact of the administrative state (cf. Torgerson, 1984: Ch. 2).
It should not be surprising, moreover, that the changing orientation in the phi-
losophy of the social sciences carries with it implications not only for the con-
duct of policy analysis, but also for the conduct of political life. If positivism
removed the analyst from the human world and turned him into a neutral ob-
server, then post-positivism returns the analyst to this world as an active par-
ticipant.
The methodological point is this: the analyst must develop not only a knowl-
edge of society, but also a knowledge in society, z2 Social life is conducted,
 41

quite simply, on the basis of common sense - the generally shared meaning
of words, gestures, and institutions which constitutes a particular mode of cul-
tural understanding. Anyone who seeks to study social life must possess such
understanding; one must be "in on" the generally shared meaning of things
(Taylor, 1971: 13). Research thus can no longer be divorced from society, but
becomes in principle a particular mode of social action. The appropriate mod-
el of the analyst, in other words, is not the detached observer, but the
participant-observer. No longer is participant-observation a particular research
technique; broadly conceived, it becomes the very foundation of all research
effort (cf. Cicourel, 1964: 40).
In any policy issue, moreover, there will no doubt be questions which can
be answered with relative assurance and precision. Yet policy issues also point
beyond these matters to questions that pertain to the broader political and
historical context. These questions, which typically resist precise and unam-
bivalent answers, are ones which the conventional analyst usually seeks to
avoid - favoring, instead, comfortable questions amenable to an established
repertoire of research techniques. However, no mode of analysis which claims
to be rational can ultimately avoid facing the broader questions: for these
questions involve the nature and identity of the analytic project itself. To fail
to ask the broader questions means being blind not only to one's context, but
also to oneself.
In rejecting both the self-assurance and detachment of conventional analy-
sis, the post-positivist orientation offers insights which are, of course, by no
means novel. Indeed, the current emergence of post-positivism in the philoso-
phy of the social sciences has clearly been supported by a mode of discourse
which pre-dates and has resisted positivism - i.e., the tradition of political
thought as interpreted and preserved by certain key contemporary figures. 23
From this perspective, the chief defect in conventional policy analysis would
be its lack of "political understanding" (cf. Wolin, 1969; 1977; Jung, 1979).
The authoritative posture of self-assurance and detachment relies upon a
methodological orientation which removes one from the everyday language
and experience of politics, obscuring political reality and the part one plays in
it. Inattention to context, history, and the complexity of political life thus
results in a "crisis of political education" (Wolin, 1969: 1070-1077):

Political life does not yield its significance to terse hypotheses but is elusive,
and hence meaningful statements about it often have to be allusive and in-
timative. Context becomes supremely important, for actions and events oc-
cur in no other setting. Knowledge of this type tends, therefore, to be sugges-
tive and illuminative rather than explicit and determinate (p. 1070).

Early proponents of a policy orientation in the social sciences were not una-
42

ware of the need for such political understanding. The influence of traditional
political thought thus loomed large in Karl Mannheim's proposal for a "scien-
tific politics" to guide social development (Mannheim, 1936: Ch. 3; Mann-
heim, I940; Kettler et al., 1984). Here Mannheim identified the "task of politi-
cal education" as keeping alive a flexible orientation to changing situations:
"This ability to reorient oneself anew to an ever newly forming constellation
of factors constitutes the essential practical capacity for the type of mind
which is constantly seeking orientation for action" (Mannheim, 1936:
176-177). Harold D. Lasswell was also fully conversant with the traditional
literature of political thought, and its influence is easily discernable in his argu-
ment that a project of contextual orientation is needed in the development of
the policy sciences (cf. Torgerson, 1985).
While these early proposals tended to envision policy analysis with a dis-
tinctly human face, they were largely eclipsed and misinterpreted - rendered
marginal - as positivist notions and technocratic dreams came to dominate
the social sciences. Lasswelrs work is especially significant in this regard be-
cause it reflects a deep tension in policy analysis between technocratic and par-
ticipatory orientations. Critics of Lasswell have often held him up as the arche-
type of the positivist and technocrat, and they have found support for this view
in some of his early formulations. However, Lasswell's work is predicated on
a methodological framework which, by drawing attention to the significance
of contextual orientation, emphasizes the inevitable participation of the in-
quirer in a human world. Although striking a rather technocratic note in some
early writings, moreover, Lasswell came to be concerned mainly to advance a
policy-analytic profession which would counter bureaucratic and oligarchic
tendencies; he envisioned a profession which would promote both the general
enlightenment of the population and the widespread sharing of power in a
democratic policy process. What we witness in the development of Lasswell's
work, then, is a methodological-political convergence which centers on partici-
pation.
Without directly refering to Lasswell, Kariel has clearly articulated the na-
ture and significance of such a convergence:

It may yet become possible to see what is entailed by defining research as

a form of action - the full presence of the researcher in his field of concern.
Present in a field that includes his multiple unrealized parts, ideally his en-
tire self, he is impelled to realize, moreover, that he simply cannot know
what is least costly to actual and potential human possibilities without in-
volving whoever is affected by the policy, himself included, in the designing
of their society... Participation... becomes an epistemological imperative;
action and knowledge are inextricably one (Kariel, 1972: 107).24
 43

Just as positivism unaerues the dominant technocratic tendency in policy anal-

ysis, so the post-positivist orientation now points to a participatory prospect.
Here the methodological posture complements a particular commitment
which has been voiced on the fringes of the policy literature: a commitment
to promote a policy process which both permits and encourages greater citizen
participation. 25
In the context of a prevailing technocratic orientation, the participatory
potential in policy analysis was easily ignored and obscured - even when
stressed by such a key figure as Lasswell. If this potential is now becoming
more visible, it is at least partly because recent developments in the philosophy
of the social sciences have drawn into question the narrow notion of reason
fostgred by positivism. No longer is reason restricted to calculation and
mechanization, certainty and precision; now reason both requires and pro-
motes the inquirer's knowledge of self and context - a knowledge which
orients one for action and analysis amid the contingencies of social and politi-
cal life. Indeed, it is perhaps only with the advent of post-positivism that the
significance of Lasswell's proposal for contextual orientation can be fully
recognized and appreciated.
Yet from a post-positivist perspective, even Lasswelrs proposal for an en-
lightened, democratically committed policy science profession remains some-
what questionable (cf. Torgerson, 1985). For it is apparent that a focus on
professional development could obscure what is really at stake - the quality
of public life in advanced industrial society. The full emergence of the third
face of policy analysis depends not simply upon the professional development
of the field; it depends upon a broader context of social and political forces
which typically tend to inhibit moves toward a participatory orientation. If
technocratic analysis is an artifact of the administrative state, then a participa-
tory focus challenges an established set of power relationships and anticipates
major changes in the political world (cf. Pateman, 1975; Macpherson, 1977:
Ch. 5).
The tension in policy analysis between its technocratic tendency and its par-
ticipatory potential becomes especially evident when we consider innovations
which have been developed to deal with problems perceived in the course of
industrialization (cf. Torgerson, 1980, 1981). Efforts to give explicit considera-
tion to social and environmental implications of technical innovation and in-
dustrial expansion - technology assessment, environmental impact assess-
ment, social impact assessment - do often exhibit a narrow technocratic
orientation. Nonetheless, these efforts have been predicated on a certain disil-
lusionment with positivist notions concerning the progress of history toward
a mature industrial civilization. The institutionalization of these forms of poli-
cy research represents a recognition of previously unforeseen problems in the
development of industrial society. At least implicitly, then, the advent of im-
44

pact assessment is founded on a view of the context of analysis which is at

odds with a previously conventional one.
Both in theory and in the actual conduct of impact assessment, moreover,
there is at times a tendency to give explicit consideration to the wider context
of analysis. Here the research project itself may come to be viewed as a social
process which, no less than the findings, can affect the character and quality
of social and political life. In such a mode of impact assessment there may,
at decisive points, also be active involvement in the process by the social groups
being studied. Here the focus shifts from a fixation on analytic technique to
a concern with the process of inquiry. Presently, we shall examine an extraordi-
nary example of such "participatory" impact assessment. 26 For the moment,
however, we need to consider further the distinction between analytic tech-
nique and the process of inquiry.
The intrinsic limitations of conventional analytic techniques have been
clearly recognized by modest proponents (see, e.g., Nash et al., 1975). These
limitations, evident to sophisticated observers even within a positivist frame-
work, become especially clear in the issues which policy analysis encounters
with the advent of impact assessment. This point has been emphasized by
Giandomenico Majone, who argues in effect that the presence of "trans-
scientific" issues 27 in the assessment of technological innovation propels poli-
cy analysis beyond the technocratic orientation:

To cope with the problem posed by technology assessment, [policy] analysis

must become a sort of generalized jurisprudence. Before worrying about de-
cision rules, utilities, optimality and all the other categories of decision
analysis, one must be able to assess the adequacy of argument, the strength
and fit of evidence, the relevance and reliability of data, the intrinsic limita-
tions of scientific tools, the pitfalls lurking in every technical conclusion. To
get to the "truth," the analyst will have to rely not on models and al-
gorithms, but on advocacy and the adversary process. The supreme analytic
achievement is no longer the computation of optimal strategies, but the de-
sign of procedural rules and social mechanisms for the assessment of incom-
plete and often contradictory evidence (Majone, 1977: 174).

The key thrust of Majone's argument is that - now more than ever - analysis
deals with a range of issues which shatters the technocratic expectation of pre-
cise and certain solutions. The assumptions necessary to proceed with the usu-
al calculations now appear unreasonable because they effectively detach analy-
sis from relevant uncertainty and complexity.
Majone is aware that the rationality presupposed by inquiry is seldom close-
ly approximated, especially when analysis finds itself influenced by political
allegiances. Still, he proposes to change the existing communicative context of
 45

inquiry in a manner which will aid in the collective achievement of "truth."

His argument makes it clear that progress in this direction requires less the abo-
lition than the recognition of the political context. The advocates who meet
one another to argue out the judgments involved in public policy may well rep-
resent competing interests: the point of a "generalized jurisprudence" is to
draw these interests out into the open - demanding reasons for policy alterna-
tives, rather than allowing initiatives to be pushed through covertly. 28 Here he
emphasizes the need for policy analysis to overcome a lack of methodological
principles corresponding to "legal notions like reasoned decision, proper form,
and rules of evidence" (Majone, 1982: 289). Beyond this, he identifies and en-
dorses a key principle for a rhetoric oriented to truth: to "make the weaker case
stronger" (Majone, 1977: 174-175). It is in the interest of rational inquiry, he
argues, not only for unconventional ideas to be tolerated, but also to be en-
couraged and seriously considered (Majone, 1982: 289).
Concerned with the maintenance of an existing mechanism, the first face of
policy analysis displayed a fascination with technique - and a neglect of the
significance of process. The disillusionment in the second face came essentially
from a realization that policy-analytic techniques are not the detached, neutral
devices they might appear to be - that they are, indeed, imbedded in a politi-
cal process. The emergence of a third face begins with this realization: it is un-
derstood that the theory and practice of policy analysis are rooted in inherently
political choices. To decide upon a particular mode of inquiry is necessarily
to affect the life and values of a society while influencing the shape of the
wider political process. The key point is that the choice may either inhibit or
enhance the rationality of political life.

The third face: a practical perspective

Today the third face of policy analysis appears mainly as a methodological

prescription predicated on a new direction in the philosophy of the social
sciences. To picture the third face clearly, we need to shift from a theoretical
to a practical perspective. Yet this is no simple matter, for the third face of poli-
cy analysis remains, at best, an emerging tendency, something which must be
anticipated. To be sure, it has been anticipated before. Lasswell alone advanced
and experimented with a series of proposals - the decision seminar, the social
planetarium, the power sharing prototype - which had distinctly participato-
ry overtones (Lasswell, 1971: Chs. 4, 8; cf. Bolland and Muth, 1984; Muth and
Bolland, 1981; Kariel, 1969: 137-139). There have been other innovations in
a similar vein, such as public participation in planning (Arnstein, 1969; Stewart
et al., 1984), experiments with policy mediation (Amy, 1983), the proposal for
permeable and internally diverse "technical communities" (MacRae, 1976), the
46

use of science courts (Mazur, 1977), and the promotion of informed citizen in-
tervention at public hearings (Salter and Slaco, 1981; Weller and Jackson,
1982). However, no such proliferation of examples can provide a full and clear
picture. Indeed, we have only fragments before us, not a total configuration.
The form of the third face can perhaps be vaguely imagined, but its full reali-
zation remains a practical task.
Nonetheless~ a better glimpse of the third face can be gained if we look to
the relatively new field of impact assessment. We have noted that the tension
between the technocratic and participatory dimensions of policy analysis is es-
pecially acute in this area. This is not surprising because impact assessment
sprang from broad social concerns questioning the pace and direction of in-
.dustrialization (Torgerson, 1980). In this context, specific, narrowly defined is-
sues have had a tendency to blossom into broader considerations about the de-
velopment of society. For example, impact assessment controversies over the
safety of nuclear power plants have not only raised issues about the adequacy
of conventional analytic techniques, but have evolved into a general debate
about the direction of energy policy: the "hard path" versus the "soft path."
Voicing criticism of both technocratic policy analysis and the centralized ad-
ministrative state, soft path proponents have, indeed, argued for a more par-
ticipatory policy process within the framework of a decentralized, democratic
pattern of social, economic and political development (cf., e.g., Lovins, 1977;
Orr, 1979).
It is in the field of impact assessment, moreover, where we find a policy-
analytic project which dramatically displays the essential features of the third
face. Indeed, the Mackenzie Valley Pipeline Inquiry, conducted by Mr. Justice
Thomas Berger for the Canadian federal government, likely remains the most
extensive impact study yet undertaken. 29 What could have been a routine ana-
lytic exercise became a key part of a Political and historical transformation in
the Canadian North. An interpretation of this experience is likely to provide
as clear a practical perspective on the third face as is now possible given the
present stage of development in policy analysis - and in the public life of
modern society.
Established in 1974 by the federal Cabinet of Canada, the Berger Inquiry,
as it came to be known, was directed to examine "the social, environmental and
economic impact regionally" of a proposal to transport natural gas from Prud-
hoe Bay in Alaska and from the Mackenzie Delta in the Northwest Territories
south by pipeline through the Mackenzie Valley (Berger, 1977: Vol. 1, appen-
dix). Although vigorously promoted by major international petroleum corpo-
rations, the proposal was not approved by Cabinet; another pipeline project
to transport gas from Alaska was approved instead. The negative recommen-
dation of the Berger Inquiry in 1977 was one factor - among others probably
more significant - which influenced the pipeline decision (cf. Bregha, 1979).
 47

Here, however, the real significance of the Inquiry lies in the approach it adopt-
ed to policy analysis as a process (cf. Gamble, 1978; Berger, 1977: Vol. 2,
223-230).
Viewed as a process, the Berger Inquiry can be described as a relatively open
forum allowing the articulation of divergent interests and perspectives. The In-
quiry solicited and received scholarly reports on various dimensions of the
pipeline issue. The research, however, was not simply accepted at face value.
Impact statements countered one another; study challenged study. With re-
plies, rebuttals, cross examinations, debate moved at a high level of sophistica-
tion. Scholarly research, moreover, remained only one aspect of the process.
For members of the public, both regionally and nationally, were not only per-
mitted, but encouraged to speak their minds. Here three significant aspects of
the Inquiry can be identified.
1. Information. All parties to the Inquiry were required to submit a list of
all documents in their possession pertinent to the work of the Inquiry. This
allowed all parties to request the documents of others - and, if their requests
were not met, to ask the Inquiry to issue a subpoena. In the end, all parties
cooperated and no subpoenas were issued.
2. Funding. Despite the evident reluctance of the federal Cabinet, Berger
provided funding to native groups, environmental groups, northern municipal-
ities, and northern businesses - in order, as he putlt, to "enable them to par-
ticipate on an equal footing (so far as that might be possible) with the pipeline
companies" (Vol. 2, p. 225).
3. Form o f the lnquiry. The conduct of the Inquiry revolved about an exten-
sive series of public hearings. Berger divided the hearings into two types: for-
mal and community. The formal hearings involved sworn testimony and the
cross-examination of witnesses. This was the domain of the expert witnesses,
mainly academics and private consultants who had carried out research in the
Canadian North. Even in the formal hearings, Berger sought to avoid "a too
rigid observance of legal rules of admissibility, for that might have squeezed
the life out of the evidence" (p. 227). Nonetheless, the community hearings
provided a sharp contrast. They were held in all the major cities of southern
Canada and in all the settlements in the Mackenzie Valley region. There was
no sworn testimony, only Berger cross-examined witnesses, and all participants
were allowed to speak in their own languages. A forum encouraging active
public participation was carefully designed. This encouragement was met by
an extraordinary response, especially from native people, who on the whole
presented their views thoughtfully and eloquently. Indeed, for Berger, the com-
munity hearings appeared decisive:

No academic treatise or discussion, formal presentation of the claims of na-

tive people by the native organizations and their leaders, could offer as corn-
48

pelling and vivid a picture of the goals and aspirations of native people as
their own testimony. In no other way could we have discovered the depth of
feeling regarding past wrongs and future hopes, and the determination of
native people to assert their collective identity today and in the years to
come (Vol. 2, p. 228).

While it promoted a high level of public involvement, the Inquiry also consid-
ered the question of development from a broad perspective. Beginning with
one specific project, the Inquiry examined related plans and proposals and
concluded that at stake was a decision crucial to the whole pattern of future
development in the area.
In his report, Berger emphasizes that a basic divergence arose during the
hearings between two opposing "philosophies" (Vol. 2, pp. 3-6; Vol. I, pp.
148 ff.). One is generally committed to the rapid, full modernization of the
North - industrialization, the expansion of private business, and the creation
of employment opportunities for both whites and native people. From this
viewpoint, the North, Berger suggests, is seen as a "frontier." In contrast, he
maintains, the other perspective regards the region as a "homeland." Indeed,
Berger claims that the latter is the overwhelming view of the aboriginal peoples
whose ancestors had inhabited the land since prehistorical times. This position
stresses the importance of self-determination. Development is not altogether
rejected; yet the conventional notion of development as a virtually natural and
irresistable process is replaced by a conception which considers development
as a matter of collective choice. The goal becomes a mode of development
which can support the cultural identity of native people in a way that allows
any cultural change to contain significant self-direction. There remains a role
for industry, business, and wage employment. The accent, however, is put on
renewable resources, co-operative enterprises, and community development.
In his overall assessment, Berger endorsed the second viewpoint, suggesting
that the other was ethnocentric and superficial:

I think the basic reason for this gulf in our belief in the benefits of industrial
employment and the native people's fear of it is that the native people of
the North are not simply poor people who happen to be of Indian, Inuit or
Metis descent. They are people whose values and patterns of social organi-
zation are in ways quite different from those that underlie the modern indus-
trial world (Vol. 1, p. 148).

An adequate assessment of the pipeline proposal, Berger insisted, required that

one overcome any "lingering reluctance to take the views of native people seri-
ously when they conflict with our own notions of what is in their best in-
terests" (Vol. 2, p. 228).
 49

The Berger Inquiry was a significant event for peoples that - as the report
of the Inquiry itself revealed - had long been in a position of acquiesent
subordination to the dominant elements of Euro-Canadian society. The In-
quiry, indeed, became something more than a setting where interests could be
freely articulated; it became a forum also for the clarification and recognition
of interests. In fact, the Inquiry itself had a significant impact, culturally and
politically, on the native peoples. As a recent study has argued, the Berger In-
quiry was the "midwife" to a political transformation in the region which saw
important institutional innovations combined with the emergence of the native
peoples as a significant force (Abele, 1983: 79; cf. Ch. 6).
Considerable controversy arose over the Berger Inquiry, in part focusing on
the analytic adequacy of Berger's report. Here we might consider some exten-
sive quotations pertaining to this controversy. In these passages, it should not
be difficult to glimpse all the three faces of policy analysis. First we look to
one of Berger's critics:

Until recently project evaluation in the public sector, as in the private sector,
consisted of comparing direct benefits which could be measured in terms of
dollars against direct costs which were also measured in dollar values. Spil-
lover effects, the unintentional by-products of the intended course of action,
were recognized to exist but since they often involved impacts which had no
measurable counterpart in a private market, project evaluation has, in recent
years, included a determined attempt to evaluate the spillover, or indirect,
effects of the proposed course of action. This is done through the impact
study which.., provides the experts' best and most specific assessment of
the effect of the proposed course of action on the subject which is the focus
of attention. Such an assessment may be made in dollar terms or, if this is
not feasible, in units most appropriate to the subject under investigation...
The information is used by the responsible person or group, in conjunction
with that provided through direct evaluation, in order to arrive at a decision.
The decision is reached by weighing direct plus indirect benefits against di-
rect plus indirect costs... The decision is made, of course, only after a full
consideration of all the relevant information (Stabler, 1977: 57-58).

The critic deems the Berger report "inadequate" to this standard of analysis.
In his report, however, Berger himself advanced a conception of inquiry quite
at odds with the critic's:

There is a tendency, in estimating the impact of a large-scale industrial pro-

ject, t o . . . minimize the importance of conclusions which are unsupported
by "hard data." Usually those in favour of the project are able to say ap-
proximately how much it will cost, although experience with s o m e . . , large-
50

scale.., projects.., has indicated that the early estimates of costs have been
completely unreliable. But at least there is a set of figures to work with, and
they offer the comforting illusion that you are dealing with hard data (Vol.
1, p. 143).
In considering the social impact of large-scale developments, very few
figures are available... There is a strong tendency to understate social im-
pact and social costs, and there is a tendency to believe that whatever the
problems may be, they can be overcome (p. 143).
Some impacts are easier to predict than o t h e r s . . . Moreover, some conse-
q u e n c e s . . , will be controllable, hut others will n o t . . . It is all too easy to
be overconfident of our ability to act as social engineers and to suppose -
quite wrongly - that all problems can be foreseen and resolved. The nature
of human affairs often defies the planners (pp. 160-161).
If you are going to assess impact properly, you have to weigh a whole se-
ries o f matters, some tangible, some intangible. But in the end, no matter
how many experts there may be, no matter how many pages of computer
printouts may have been assembled, there is the ineluctable necessity o f
bringing human judgment to bear on the main issues. Indeed, when the
main issue cuts across a range of questions, spanning the physical and social
sciences, the only way to come to grips with it and to resolve it is by the exer-
cise o f human judgment (Vol. 2, p. 229).

While the critic advances what purports to be to be an authoritative account

of the correct approach and legitmate place of analysis in public policy con-
siderations, Berger locates analysis within an ideological context. He points to
a prevailing bias and indicates that the bias tends to obscure "the nature of hu-
man affairs." Berger also understands project assessment as a political proc-
ess. No doubt, he was aware that to decide upon the form of an inquiry is -
whatever the decision - a political act. By encouraging the participation of
non-experts - particularly native people - Berger provided a forum for open
discussion in which individuals and communities could not only express their
views, but could develop their capacity both to recognize and articulate their
interests. In form and style, the inquiry not only provided an opportunity for
participation, but also developed into an educative experience promoting the
ability, both individual and collective, for reasoned and effective participation.
Berger thus took a political stance, but we should note that he had no option
for neutrality. Any form o f inquiry whatsoever would have constituted a social
process o f political significance. If Berger had opted for a more conventional
form of inquiry - one with a focus on expert testimony and a minimum of
public involvement - his stand would still have been political, even if less visi-
bly so.
In light o f the Berger experience, we can recognize participation by citizens
 51

in the process of inquiry as being a form of public participation in political

life. Here policy analysis ceased to be an agent o f the administrative state and
- by involving the public in an assessment o f economic, social and political
development - promoted an open forum for deliberation on public issues. In-
deed, when it promotes the development o f such a forum for the discussion
of public affairs, policy analysis actually sets itself at odds with a policy proc-
ess cloistered in the administrative state. The logic of inquiry points to a trans-
formation of public life.

Conclusion

In the third face of policy analysis, there remains a distinction between expert
and citizen. However, the experts have developed an acute awareness of their
own frailty and fallibility. Further, they do not erect a facade to hide the limita-
tions which they can perceive in themselves, but make their humanity appar-
ent. Indeed, experts take care not to mystify non-experts, and the citizenry
guards against mystification. In the communication between expert and citi-
zen, each seeks to identify points where the boundary between expertise and
c o m m o n sense becomes fluid and indistinct; each seeks, that is, to expand the
domain in which active participation can educate a broad range of the popula-
tion for a rational consideration o f both specific policies and general social
directions.
We have asked whether the demands of knowledge and politics can be recon-
ciled - whether rational inquiry has any place in the real world of public poli-
cy. In the third face of policy analysis, we find the potential for some such
reconciliation. Hope for this must surely be tempered by a serious assessment
of the actual politica ! context of policy analysis. 30 Still, at least some theoreti-
cal and practical ground has been prepared for such a development.
The concept of knowledge has been expanded through a recognition that
contextual knowledge is both an end and a means of rational inquiry. Indeed,
it may further be recognized that, since it carries a commitment to unfettered
investigation and dialogue, reason is by no means value free - that, in effect,
it carries a political commitment. An un!nhibited quest for knowledge in pub-
lic affairs would require a change in the prevailing character of public life. 31
The third face of policy analysis, indeed, points toward a new conception o f
politics - or, more correctly, to the revival o f an ancient one. Today we readily
equate politics with strategic maneuvering and conflict, but knowledge and
politics have not always been conceived as being completely antagonistic.
There is an earlier sense of politics in which conflict is balanced with the
cooperation necessary for an open, reasoned discussion of public affairs. 32
To assert that earlier idea of politics, one does not have to begin by attempt-
52

ing to change the entire fabric of existing political institutions. One could in-
stead begin with a single thread: the institution of policy analysis itself. 33 To
embark upon a project of contextual reorientation - dispelling the tech-
nocratic mystique and pointing policy analysis in a post-positivist direction -
would, indeed, constitute a political act in which reason and politics could fi-
nally converge.

Acknowledgements

A preliminary version of this article was presented to members of T.A.S.O. (Re-

search Program for Technology Assessment in Subarctic Ontario), McMaster
University, Hamilton, Ontario, October 11, 1984. For a lively and helpful dis-
cussion, I wish to thank T.A.S.O. Director Richard Preston and the other par-
ticipants. For valuable criticisms and suggestions, I am grateful also to the
anonymous reviewers and to the editor of Policy Sciences.
I would welcome correspondence on theoretical and practical efforts to de-
velop what is here termed a "third face" of policy analysis.

Notes

I. The problem is compounded by the differing analytic schemes which have been used to sort
out the field. See, e.g., Schoettle (1968), Heclo 0972), Dye (1975), Aucoin (1979), Bardes and Dub-
nick 0980), Nagel 0980). One might draw a parallel between the policy literature and the related
administrative literature (cf. Koontz, 1961).
2. This article attempts briefly to reformulate in a more accessible way a view of policy analysis
which has been developed at some length in two earlier studies (Torgerson, 1980, 1984). According-
ly, the article treats in a rather cursory manner issues dealt with more thoroughly in those previous
works.
3. This formulation is meant to be suggestive rather than definitive. The nature of the relevance
is deliberately left unspecified, and the focus on formulation and implementation is not meant to
rule out attention to other aspects of the policy process. The nature of knowledge, moreover, is
a concern to be raised throughout the article. The conception is broad in the sense that it incor-
porates both those policy studies with a directly applied focus and those without. Moreover, no
particular methodological orientation is presupposed.
4. No doubt, many will note the dialectical character of the argument. It is thus necessary to
warn that the rather mechanical and well-known scheme of thesis, antithesis, and synthesis (der-
ived from Fichtr not Hegel or Marx) is likely to be quite misleading. Here we do not have one
distinct mode of policy analysis opposed to another with the opposition resolved in some sort of
synthesis of the two. Instead, policy analysis is conceived as a single, yet multifaceted and dynam-
ic, phenomenon. The dynamic nature of the phenomenon is rooted in an internal tension, a dialec-
tical opposition between knowledge and politics. Through the interplay of knowledge and poli-
tics, different aspects of the phenomenon become salient at different moments - it takes different
forms, projects different "faces." The presence of dialectical tension means that the phenomenon
 53

has the potential to develop, to change its form. However, no particular pattern of development
is inevitable. Cf. Kaufmann (1966: 153-162), Avineri (1968: Ch. 6).
5. For a relevant discussion of the Enlightenment and early positivism, see Torgerson (1980:
62-70). On the Enlightenment generally, see Gay (1966, 1969), Cassirer (1951). Perhaps the most
representative statement of early positivism is Comte (1974). Cf. Aron (1968). The famous phrase
"the administration of things" is, of course, derived from Frederick Engels, but he was clearly in-
spired by the positivist Saint-Simon (Engels, 1978: 42). Cf. Saint-Simon (1952).
6. Just as modern technology has been derived from the physical sciences, it is believed, so so-
cial technology will arise from advances in the social sciences (Helmer, 1966: 3-6). Such a concep-
tion of social technology has received its most sophisticated critique in the work of J/irgen Haber-
mas (see, e.g., Habermas, 1971h, 1974). Cf. Torgerson (1980: Ch. 2; 1984: Ch. 3). Churchman (1979)
can be read as an internal critique of social technology in the guise of a popular introduction.
7. On the distinction between early positivism and the more recent logical positivism, see Gid-
dens (1977). On positivism generally, see Kolakowski (1968).
8. This view received its most important formulation in Ayer (1952).
9. In metaethics and political philosophy, this position is known as "noncognitivism" (see Op-
penheim, 1968:24 ff.).
10. The tensions are especially apparent in the work of Harold D. Lasswell. For an example of
his concerns regarding the commitment of policy analysis to democracy or tyranny, see Lasswell
(1971: 42-43). For significant critiques of his position, see Taylor (1973) and Brown (1976). Cf.
Torgerson (1985).
11. On the paradox of positivism as a philosophy of history, see Habermas (1971a: 71 ff.).
12. Easton argued against this orientation in favor of a broader commitment to human pro-
gress. So did Lasswell (e.g., 1951; 1958; 1962; 1971: Ch. 7; 1974).
13. At a philosophical level, the critique of Enlightenment has been a major theme of critical
theory. See, in particular, Horkheimer and Adorno (1972).
14. While written in Russian, this book was originally published in an English translation
(1924). The book apparently influenced Orwell.
15. As Deputy Secretary, Planning Branch, Treasury Board Secretariat, Hartle was at the center
of efforts to introduce a Planning Programming Budgeting System (P.P.B.S.) into the federal
government of Canada. The Canadian experience is especially interesting with regard to the tech-
nocratic orientation of conventional policy analysis because of the sympathetic "policy philoso-
phy" of the Prime Minister of the time, Pierre Trudeau (see Doern, 1977). For a penetrating analy-
sis of this period in Canada, see French (1984: Chs. 1-2). Also of. Torgerson (1984: 52-66).
16. Hattie's work contains interesting similarities, in both substance and tone, to that of Aaron
Wildavsky. See, e.g., Wildavsky (1966, 1979).
17. For a modest statement in favor of conventional analytic techniques, see Rivlin (1971). For
rather profound criticisms, however, see Tribe (1972), Self (1975), Lovins (1977). Cf. Nash et al.
(1975).
18. The work of Tribe (1972, 1972- 73) is central to this critical assessment of conventional poli-
cy analysis. Also see Wildavsky (1966), Self (1975), Van Gunsteren (1975), Fay (1975), Kramer
(1977). By now the critique of conventional policy analysis has become vocal enough for one au-
thor to identify a school of "anti-policy analysts" and to offer some comments in defense of the
orthodox view (McAdams, 1984). For a sympathetic treatment of some of the critical writers, see
Torgerson (1984: 16-19, 38-45).
19. While one might question how adequately me assessment quoted from French serves as a
description of major trends in the policy literature, a post-positivist trend is clearly emerging. See
Amy (1984), Torgerson (1980, 1984), Paris and Reynolds (1983), White (1982), Dallmayr (1980-81),
Fischer (1980). This literature does not conclude with a critique of conventional tendencies, but
takes this as a point of departure in attempting to reorient the foundations of policy analysis,
54

20. It is thus that Kolakowski (1968) refers to an "alienation of reason" and Horkheimer (1974),
to an "eclipse of reason." One of the central themes in the post-positivist literature of the philoso-
phy of the social sciences is that the positivist conception of reason needs to be expanded. Such
an expanded concept departs from conventional tendencies to equate rationality with final
products of cognition (e.g., accurate facts, correct strategies, logical consistencies) and, instead,
grounds rationality in processes of critical self-reflection and collective discourse. A rational judg-
ment does not have to achieve absolute certainty; indeed, it remains open to later revision or rejec-
tion. What is essential to a rational judgment is the process of individual and collective delibera-
tion through which it is formed - a process, in principle, of uninhibited exploration and criticism.
Reason, in this sense, is the testing ground of a knowledge which is always open to question. (Cf.
Apel, 1978, 1979; McCarthy, 1979: Ch. 4.)
21. However obvious this may appear, it is something which is often obscured through what
Tribe (1972: 77) identifies as a conventional tendency to "collapse process into result." Fay (1975:
52) argues, similarly, that since the employment of policy analytic techniques requires "that people
interact with one another in certain definite ways," the use of these techniques is founded upon
an implicit judgment of "moral value." Cf. Elkin (1975).
22. This formulation is influenced by Lasswell (1970: 3) where "knowledge o f the policy proc-
ess' and "knowledge in the process" are described as "two separable though entwined frames of
reference." Cf. Lasswell (1971: 13).
23. Bernstein, indeed, makes this connection explicit (1978:57 ff., 185 ff.). Cf. Arendt (1958),
Berlin (1962), Habermas (1974: Ch. 1), Oakeshott (1962), Strauss (1975), Wolin (1960, 1962).
24. In an earlier development of this idea, significantly, Kariel (1969: 138) discussed a research
project of Lasswell as an "innovation... establishing a democratic forum for sharing power." Cf.
Kariel (1966).
25. For a review of participatory themes in the policy literature, see Bryden 0982). Also see the
proposal by Wildavsky (1979: Ch. 11) to develop "citizens as analysts." The concern to promote
participation is apparently not confined to writers of any one political ideology (cf. Berger and
Neuhaus, 1977; Macpherson, 1977; Pateman, 1970).
26. On the distinction between "participatory" and "technocratic" impact assessment, see Tor-
gerson (1980: Ch. 6).
27. Weinherg (1972) defines such issues as ones which can be formulated in scientific terms, but
which cannot he answered with scientific assurance - at least, not in a time frame relevant to hu-
man interests. As industrial society enters new territory - both geographically and technologically
- it generates a wide range of pressing "transoscientific" issues.
28. For a much earlier, complementary discussion from the field of jurisprudence, see Hart and
McNaughton (1958).
29. For an excellent discussion of the Inquiry by a member of its staff, see Gambel (1978). The
character of the Inquiry is brilliantly captured in a 1977 film available from the National Film
Board of Canada, The Inquiry Film: A Report on the Mackenzie Valley Pipeline Inquiry. The
most important scholarly treatment is Abele (1983). The Berger report itself contains an appendix
which examines "The Inquiry Process" (1977: Vol. 2, 223-230). Interestingly, Dryzek (1982:
324-325) briefly discusses the Inquiry and suggests that it "ilustrates the potential for effective
policy analysis within the hermeneutic mold." Unfortunately, Dryzek's very fine article is marred
by his attempt to view hermeneutics - i.e., interpretation - as a particular "model" of inquiry.
He thus obscures the point that interpretation - whether or not made explicit - is necessarily
a feature of all social inquiry (cf. Taylor, 1971). The question is how consciously and how well the
task of interpretation is undertaken. The Berger Inquiry can be seen to embody implicitly key fea-
tures of Lasswelrs proposed approach to policy inquiry. In particular, Berger's focus on the histor-
ical dimension in terms of competing perspectives on development would appear to be in accord
with Lasswelrs call for the use of "developmental constructs." For a discussion which links the
 55

Berger Inquiry to Lasswell's approach, see Torgerson (1980: 16-26, 178-189) - a discussion
which is closely followed in the present treatment of Berger. Further on Lasswell and the place
of interpretation in a post-positivist restructuring of inquiry, see Torgerson (1985).
30. This admonition should not be seen as a perfunctory one (cf. Arnstein, 1969; Pateman,
1970: Ch. 4).
31. While reason perhaps can never be completely at ease in a world of politics, it is equally
the case that reason can never be politically neutral. The interest in genuine knowledge is, indeed,
typically at odds with other forces in the policy process; and this point becomes especially appar-
ent when we are concerned to comprehend the political world itself. With its well-known conflict,
deception and secrecy, the world of politics typically resists efforts to investigate and disclose it
fully. This means that a serious commitment to rational inquiry here carries a political commit-
ment seldom perceived in the policy literature. To try to know the political world fully would neces-
sarily be to counter any resistance to investigation; the effort, in effect, would be to change an
existing policy process into one more amenable to inquiry - that is to say, into a more open policy
process. This point obviously has implications for any effort to develop a professional ethics. (Cf.
Torgerson, 1984: Chs. 4, 8.)
32. Bernstein (1976: xxi-xxii, 185-87) draws a connection between this Aristotelian concep-
tion of politics and the emergence of post-positivism in social and political inquiry. Also see
Habermas (1974: Ch. 1), Arendt (1958" Ch. 2). Cf. Kariel (1966). In explicit connection with policy
analysis, see Elkin (1985).
33. Here, of course, the political context of policy analysis would remain decisive. Effective
change in the institution of policy analysis could encourage or reinforce changes in other institu-
tions, but a movement for change within policy analysis would ultimately need to be encouraged
and reinforced by these wider changes.

References

Abele, Frances (1983). "The Berger Inquiry and the politics of transformation in the Mackenzie
Valley," Ph.D. Thesis, York University. Toronto.
Amy, Douglas J. (1983). "Environmental mediation: An alternative approach to policy stale-
mates," Policy Sciences 15: 345-365.
Amy, Douglas J. (1984). "Toward a post-positivist policy analysis," Policy Studies Journal 13:
207-211.
Apel, Karl-Otto (1978). "The conflicts of our time and the problem of political ethics," in Fred
R. Dallmayr (ed.), From Contract to Community. New York: Marcel Dekker, pp. 81-101.
Apel, Karl-Otto (1979). "Types of rationality today: The continuum of reason between science
and ethics," in Theodore F. Geraets (ed.), Rationality Today. Ottawa: University of Ottawa Press,
pp. 307- 340.
Arendt, Hannah (1958). The Human Condition. Chicago: University of Chicago Press.
Arnstein, Sherry R. (1969). "A ladder of citizen participation," Journal o f the American Insti-
tute o f Planners 35: 216-224.
Aron, Raymond (1968). "Auguste Comte." In his Main Currents in Sociological Thought, Vol.
1. Garden City, NY: Anchor Books, pp. 73-143.
Aucoin, Peter (1979). "Public-Policy Theory and Analysis," in G. Bruce Doern and Peter Au-
coin (eds.), Public Policy in Canada: Organization, Process and Management. Toronto: Macmil-
lan of Canada, pp. 1-26.
Avineri, Shlomo (1971). The Social and Political Thought o f Karl Marc Cambridge: Cam-
bridge University Press.
56

Ayer, A. J. (1952). Language, Truth and Logic. New York: Dover Publications, 2nd ed.
Bardes, Barbara A., and Dubnick, Melvin J. (1980). "Motives and methods in policy analysis,"
in Stuart S. Nagel (ed.), Improving Policy Analysis. Beverly Hills, CA: Sage Publications, pp.
101-127.
Baum, Howell S. (1982). "Policy analysis: Special cognitive style needed," Administration and
Society 14: 213-236.
Beneviste, Guy (1977). The Politics of Expertise. San Francisco: Boyd and Fraser, 2nd ed.
Berger, Peter L., and Neohaus, Richard John (1977). To Empower People: The Role of Mediat-
ing Structures in Public Policy. Washington, D.C.: American Enterprise Institute for Public Policy
Research.
Berger, Thomas R. (1977). Northern Frontier, Northern Homeland: The Report of the Macken-
zie Valley Pipeline Inquiry, Vols. I - 2 . Ottawa: Supply and Services Canada.
Bernstein, Richard J. (1976). The Restructuring of Social and Political Theory. Philadelphia:
University of Pennsylvania Press.
Bolland, John M., and Muth, Rodney .(1984). "The decision seminar: A new approach to urban
problem solving," Knowledge: Creation, Diffusion, Utilization 6: 75-88.
Bregha, Franqois (1979). Bob Blair's Pipeline: The Business and Politics of Northern Energy
Development Projects. Toronto: James Lorimer.
Brown, Peter G. (1976). "Ethics and policy research," Policy Analysis 2: 225-240.
Bryden, Kenneth (1982). "Public input into policy-making and administration: The present situ-
ation and some requirements for the future," Canadian Public Administration 25: 81-107.
Cassirer, Ernst (1951). The Philosophy of the Enlightenment. Fritz C. C. Koelin and James P.
Pettegrove (tr.). Princeton: Princeton University Press.
Churchman, C. West (1979). The Systems Approach. New York: Dell Publishing, rev. ed.
Cicourel, Aaron V. (1964). Method and Measurement in Sociology. New York: The Free Press.
Comte, Auguste (1974). "Plan for scientific operations necessary for reorganizing society," in
Richard Fletcher (ed.), The Crisis of Industrial Civilization: The Early Essays of Auguste Comte.
London: Heinemann Educational Books, pp. 11-181.
Dallmayr, Fred R. (1980-81). "Critical Theory and Public Policy," Policy Studies Journal 9:
522- 534.
Doern, G. Bruce (1977). "The policy-making philosophy of Prime Minister Trudeau and his ad-
visers," in Thomas A. Hockin (ed.), Apex of Power: The Prime Minister and PoliticalLeadership
in Canada. Scarborough, Ont.: Prentice-Hall of Canada, 2nd ed., pp. 189-196.
Dryzek, John (1982). "Policy analysis as a hermeneutic activity," Policy Sciences 14: 309-329.
Dye, Thomas R. (1975). Understanding Public Policy. Englewood Cliffs, N J: Prentice-Hall, 2nd
ed.
Easton, David (1971). The Political System: An Inquiry into the State of Political Science. New
York: Alfred A. Knopf, 2nd ed.
Elkin, Stephen L. (1985). "Regulation as a Political Question," Policy Sciences 18: 95-108.
Engels, Frederick (1978). Socialism: Utopian and Scientific. Moscow: Progress Publishers.
Fay, Brian (1975). Social Theory and Political Practice. London: George Allen and Unwin.
Fischer, Frank (1980). Politics, Values and Public Policy: The Problem of Methodology. Boul-
der, CO: Westview Press.
French, Richard D. (1984). How Ottawa Decides: Planning and Industrial Policy Making
1969-1984. Toronto: James Lorimer, 2nd ed.
Gambel, D.J. (1979). "The Berger Inquiry: An impact assessment process," Science 199:
946-952.
Gay, Peter (1966, 1969). The Enlightenment: An Interpretation, Vols. 1-2. New York: Alfred
A. Knopf.
 57

Giddens, Anthony (1977). "Positivism and its critics." In his Studies in Social and Political The-
ory. New York: Basic Books, pp. 29-89.
Gunnell, John G. (1975). Philosophy, Science and Political Inquiry. Morristown, N J: General
Learning Press.
Gunnell, John G. (1976). "Social scientific knowledge and policy decisions: A critique of the
intellectualist model," in Phillip M. Gregg (ed.), Problems of Theory in Policy Analysis. Lexing-
ton, MA: Lexington Books, pp. 29-49.
Haher, Samuel (1964). Efficiency and Uplift: Scientific Management in the Progressive Era
1890-1920. Chicago: University of Chicago Press.
Habermas, Jiirgen (1971a). Knowledge and Human Interests. Jeremy J. Shapiro (tr.). Boston:
Beacon Press.
Habermas, Jiirgen (1971b). Toward a Rational Society. Jeremy J. Shapiro (tr.). Boston: Beacon
Press.
Habermas, Jiirgen (1974) Theory and Practice. John Viertel (tr.). Boston: Beacon Press.
Hart, Henry M., Jr., and McNaughton, John T. (1958). "Evidence and Inference in the Law,"
Daedalus 87: 40-64.
Hartle, Douglas G. (1976a). "Techniques and processes of administration," Canadian Public
Administration 19: 21- 33.
Hat'tle, Douglas G. (I976b). A Theory of the Expenditure Budgetary Process. Toronto: Univer-
sity of Toronto Press.
Hartle, Douglas G. (1978). The Expenditure Budget Process in the Government of Canada.
Toronto: Canadian Tax Foundation.
Heclo, Hugh (1972). "Review article: Policy analysis," British Journal of Political Science 2:
83-108.
Helmer, Olaf (1966). Social Technology. New York: Basic Books.
Hol:kheimer, Max (1974). Eclipse of Reason. New York: Seabury Press.
Horkheimer, Max, and Adorno, Theodor W. (1972). Dialectic of Enlightenment. John Cum-
ming (tr.). New York: Seabury Press.
Horowitz, Irving Louis (1970). "Social science mandarins: Policymaking as a political formu-
la," Policy Sciences 1: 339-360.
Kariel, Henry S. (1966). The Promise of Politics. Englewood Cliffs, N J: Prentice-Hall.
Kariel, Henry S. (1969). Open Systems: Arenas for Political Action. Itasca, IL: E E. Peacock
Publishers.
Kariel, Henry S. (1972). "Commentary on Lane's paper," in James C. Charlesworth (ed.), In-
tegration of the Social Sciences through Policy Analysis. Philadelphia: American Academy of Po-
litical and Social Science.
Kaufmann, Walter (1966). Hegel: A Reinterpretation. Garden City, NY: Anchor Books.
Kennedy, James (1978). Letter-to-the-editor, Technological Forecasting and Social Change 2:
383.
Kettler, David, et al. (1984). Karl Mannheim. London: Tavistock Publications/Chichester: Ellis
Horwood.
Kolakowski, Leszek (1968). The Alienation of Reason: A History of Positivist Thought. Nor-
bert Guterman (tr.). Garden City, NY: Doubleday.
Koontz, Harold (1961). "The management theory jungle," Academy of Management Journal
4: 174-188.
Kramer, Fred A. (1977). "Policy analysis as ideology," in Fred A. Kramer (ed.), Perspectives on
Public Bureaucracy. Cambridge, MA: Winthrop Publishers, 2nd ed., pp. 203-220.
Lasswell, Harold D. (1951). "The policy orientation," in Daniel Lerner and Harold D. Lasswell
(eds.), The Policy Sciences: Recent Developments in Scope and Method. Stanford: Stanford
University Press, pp. 3-15.
58

Lasswell, Harold D. (1958). "Clarifying value judgment: Principles of content and procedure,"
Inquiry 1: 87-98.
Lasswell, Harold D. (1962). "The public interest: Proposing principles of content and proce-
dure," in Carl J. Friedrich (ed.), The Public Interest. New York: Atherton Press, pp. 54-79.
Lasswell, Harold D. (1965). World Politics and Personal Insecurity. New York: The Free Press.
Lasswell, Harold D. (1970). "The emerging conception of policy sciences," Policy Sciences l:
3-14.
Lasswell, Harold D. (1971). A Pre-View of Policy Sciences. New York: American Elsevier Pub-
lishing.
Lasswell, Harold D. (1974). "Some perplexities of policy theory," Social Research 14: 176-189.
Lovins, Amory B. (1977). "Cost-risk-benefit assessments in energy policy," George Washington
Law Review 45: 911-943.
Macpherson, C. B. (1977). The Life and Times of LiberalDemocracy. Oxford: Oxford Universi-
ty Press.
MacRae, Duncan, Jr. (1976). "Technical communities and political choice," Minerva 14:
169-190.
Majone, Giandomenico (1977). "Technology assessment and policy analysis," Policy Sciences
8: 173-175.
Majone, Giandomenico (1982). "Reforming standard-setting," Journal of Policy Analysis and
Management 2: 285- 290.
Mannheim, Karl (1936). Ideology and Utopia: An Introduction to the Sociology of Knowledge.
Louis Wirth and Edward Shils (trs.). New York: Harcourt, Brace and World.
Mannheim, Karl (1940). Man and Society in an Age of Reconstruction. Edward Shils (tr.). New
York: Harcourt, Brace and World.
Mazur, Allan (1977). "Science courts," Minerva 15: 1-14.
McAdams, John (1984). "The anti-policy analysts," Policy Studies Journal 13: 91-101.
McCarthy, Thomas (1978). The Critical Theory of Jiirgen Habermas. Cambridge, MA: The
MIT Press.
Muth, Rodney, and Bolland, John M. (1981). "The social planetarium: Toward a revitalized civic
order," The Urban Interest 3: 13-25.
Nagel, Stuart S. (1980). "Introduction: policy analysis research, what it is and where it is going,"
in Stuart S. Nagel (ed.), Improving Policy Analysis. Beverly Hills, CA: Sage Publications, pp.
7-13.
Nash, Christopher, et al. (1975). "Criteria for evaluating project evaluation techniques," Jour-
nal of the American Institute of Planners 41: 83-90.
Oakeshott, Michael 0962). "Political education," in his Rationalism in Politics and Other Es-
says. London: Methuen, pp. 111-136.
Oppenheim, Felix E. (1968). Moral Principles in Political Philosophy. New York: Random
House.
Orr, David W. (1979). "U.S. energy policy and the political economy of participation," Journal
of Politics 41: 1027-1056.
Pateman, Carole (1970). Participation and Democratic Theory. Cambridge: Cambridge Univer-
sity Press.
Paris, David C., and Reynolds, James E (1983). The Logic of Policy Inquiry. New York: Long-
mart.
Rivlin, Alice M. (1971). Systematic Thinking for SocialAction. Washington, D.C.: The Brook-
ings Institution.
Saint-Simon, Henri (1952). Selected Writings. Felix Markham (tr.). Oxford: Basil Blackwell.
Salter, Liora, and Slaco, Debra (1981). Public Inquiries in Canada. Ottawa: Science Council of
Canada, Background Study 47.
 59

Schoettle, Enid Curtis Bok (1968). "The state of the art in policy studies," in Raymond A. Bauer
and Kenneth L. Gergen (eds.), The Study o f Policy Formation. New York: The Free Press, pp.
149-179.
Schick, Allen (1969). "Systems politics and systems budgeting," Public Administration Review
29: 137-151.
Schick, Allen (1971). "Toward the cybernetic state," in Dwight Waldo (ed.), Public Administra-
tion in a Time o f Turbulence. New York: Chandler Publishing, pp. 214-233.
Self, Peter (1975). Econocrats and the Policy Process: The Politics and Philosophy o f Cost-
Benefit Analysis. London: Macmillan Press.
Stabler, J.C. (1977). "The report of the Mackenzie Valley Pipeline Inquiry, volume 1: A socio-
economic critique," The Musk-Oz" A Journal o f the North 20: 57-58.
Stewart, Thomas R., et al. (1984). "Citizen participation and judgment in policy analysis: A case
study of urban air quality policy," Policy Sciences 17: 67-87.
Strauss, Leo (1975). "What is political philosophy?" in his Political Philosophy: Six Essays by
Leo Strauss. Hilail Gildin (ed.). Indianapolis: Pegasus.
Taylor, Charles (1971). "Interpretation and the sciences of man," Review o f Metaphysics 25:
3-51.
Taylor, Charles (1973). "Neutrality in political science," in Alan Ryan (ed.), The Philosophy o f
Social Explanation. Oxford: Oxford University Press, pp. 139-170.
Torgerson, Douglas (1980). Industrialization and Assessment: Social lmpact Assessment as a
Social Phenomenon. Toronto: York University Publications in Northern Studies.
Torgerson, Douglas (1981). "SIA as a social phenomenon: The problem of contextuality," in
Frank J. Tester and William Mykes (eds.), Social Impact Assessment: Theory, Method and Prac-
tice. Calgary: Detselig Enterprises, pp. 66-92.
Torgerson, Douglas (1984). "The communicative context of policy analysis: The problem of
strategic interaction in the policy process," Ph.D. Thesis, University of Toronto. Toronto.
Torgerson, Douglas (1985). "Contextual orientation in policy analysis: The contribution of
Harold D. Lasswell," Policy Sciences 18: 241-261.
Tribe, Laurence H. (1972). "Policy science: Analysis or ideology?" Philosophy and Public Af-
fairs 2: 66-110.
Tribe, Laurence H. (1972-73). "Technology assessment and the fourth discontinuity: The limits
of instrumental rationality," Southern California Law Review 46: 617-660.
van Gunsteren, Herman R. (1976). The Quest f o r Control: A Critique o f the Rational-Central-
Rule Approach in Public Affairs. London: John Wiley and Sons.
Waldo, Dwight (1948). The Administrative State: A Study o f the Political Theory o f A merican
Public Administration. New York: The Ronald Press.
Weinberg, Alvin M. (1972). "Science and trans-science," Minerva 10: 209-222.
Weller, Phil, and Jackson, John (1982). "Lessons in the need for public participation," Alterna-
tives 10 (2/3): 5 - 8 .
White, Jay D. (1982). "Public policy analysis: Reason, method, and praxis," D.P.A. Thesis, Ge-
orge Washington University. Washington, D.C.
Wildavsky, Aaron (1966). "The political economy of efficiency: Cost-benefit analysis, systems
analysis, and program budgeting," Public Administration Review 26: 292-310.
Wildavsky, Aaron (1979). Speaking Truth to Power: The Art and Craft o f Policy Analysis. Bos-
ton: Little, Brown.
Wolin, Sheldon S. (1960). Politics and Vision: Continuity and Innovation in Western Political
Thought. Boston: Little, Brown.
Wolin, Sheldon S. (1969). "Political theory as a vocation," American Political Science Review
63: 1062-1082.
Zamyatin, Yevgeny (1972). We. Mirra Ginsburg (tr.). New York: Bantam Books.