Conceptual uncertainty in U.S. civil-military relations scholarship

Material Information

Conceptual uncertainty in U.S. civil-military relations scholarship three criticisms of civil-military relations literature
Alternate title:
Conceptual uncertainty in US civil-military relations scholarship
Plotts, Thomas C
Publication Date:
Physical Description:
vi, 122 leaves : ; 28 cm

Thesis/Dissertation Information

Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Political Science, CU Denver
Degree Disciplines:
Political science


Subjects / Keywords:
Civil-military relations -- Study and teaching -- United States ( lcsh )
Civil supremacy over the military -- Study and teaching -- United States ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 117-122).
General Note:
Department of Political Science
Statement of Responsibility:
by Thomas C. Potts.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
45264526 ( OCLC )
LD1190.L64 2000m .P56 ( lcc )

Full Text
Thomas C. Plotts
B.A., University of Nebraska, 1992
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Political Science

This thesis for the Master of Arts
degree by
Thomas C. Plotts
has been approved
Kenneth M. Dolbeare


Plotts, Thomas C. (M A, Political Science)
Conceptual Uncertainty in U.S. Civil-Military Relations Scholarship: Three Criticisms
of Civil-Military Relations Literature
Thesis directed by Visiting Professor Kenneth M. Dolbeare
Contemporary U.S. civil-military relations scholars seem to be stuck with
forty-year-old constructions of basic philosophical conceptions of democracy that
may inadvertently lead to flawed analyses of the relative health of relations between
American military leaders, soldiers, and civilians. Considering that the primary
question of civil-military relations research is to discern whether or not there exists
sufficient civilian control of the military, well-understood meanings of basic
conceptions of democracy are an absolute requisite for conducting meaningful
research and analysis.
There is undoubtedly a long tradition in American democratic thought that
suggests that democracy contains often contradictory meanings, all of which are
apparently valid visions of what democracy actually means in principle. For the
majority of civil-military relations scholars, democracy has come to be represented
almost exclusively by the American conceptions of representative government and
the separation of powers. As a result, most scholars focus on relationships between
the military and government institutions in order to determine the health of
American civil-military relations. Additionally, the social scientific obsession with
quantifiable, scalar measurement contributes to the possibility that the central
question of civil-military relations scholarshipis there more or less civilian control of
the militaryis the wrong question to ask.
This thesis explores alternative analytical possibilities that emerge when the
normative idea of democracy is expanded to include populist visions. It also explores
the way our conceptions of democracy in turn affect research decisions concerning
the agency relationships to analyze in order to pronounce that relations between
military and civilian actors are unhealthy or healthy. Finally, it explores the
relationship between scalar modeling prevalent in the social sciences and those
same choices of agency variables made by civil-military relations researchers.
The purpose of this thesis is to ultimately appeal for a disengagement from
the standard, timeworn arguments that have permeated civil-military relations
scholarship for the last forty years. Reinvigorating the study Of political relationships
between soldiers and their civilian masters requires that broader understandings of

the basic philosophical premises of civilian control of the military be examined from
non-institutional perspectives.
This abstract accurately represents the content of the candidate's thesis. I
recommend its publication.
Kenneth M. Dolbeare

1. PROBLEM STATEMENT AND GOALS.................................1
Iconic Democracy in Civil-military Relations Literature.6
Agency Theories and the Meaning of Gvilian and Civil....10
Scalar Measurement in Civil-Military Relations...........20
ANALYTICAL OUTCOMES........................................25
The Myth of Single-Tradition Democracy...................31
The Relationship Between Military Policy Influence and the
Separation of Powers................................... 36
The Problem of Delegation................................42
How Democratic Assumptions Affect Analytical Outcomes:
An Illustration..........................................45
The Mysterious Fourth and Fifth Branches of Government:
Civilian and Military................................ 58
Checks and Balances, Agency Theory, and the Problem of
Delegation: Part II......................................64
Citizens and Civilian Control of the Military............69
What is Meant by Civilian Control?.......................74

Six Prevailing Analytical Models in Civil-Military Relations
Model One: The Equilibrium Principle...................87
Model Two: Input Versus Output Models..................89
Model Three: Attitudinal and Social Cleavage Models....91
Model Four: Single Principal-Single Agent Models.......93
Model Five: Multiple Principal-Single Agent Models.....94
Model Six: Multiple Principal-Multiple Agent Models....96
Is "More or Less" an Appropriate Question?.............. 97
Linear Modeling and Analytical Incoherence...............100
SUBJECTIVE CONTROL..........................................107
Terms and Multiple Meanings..............................109
Recognizing Multiple Legitimacies........................112
Emancipating Civil-Military Relations Research from
National Security Scholarship............................114
WORKS CITED .......................................................... 117

"Civilian control of the military is a basic principle of the American
Constitution"; so runs the commonplace. It is the thesis of this article that
the dichg could hardly be more inaccurate..."
Samuel P. Huntington, 1956
"...there are no standards by which to evaluate the existence of civilian
control ((of the military)) how well it functions, and the prognosis for its
continued success."
Richard Kohn, 1997
In the intervening forty-one years between Huntington's publication of
"Civilian Control and the Constitution" and Kohn's "The Forgotten Fundamentals of
Civilian Control of the Military in Democratic Government", the study of civil-military
relations in the United States has suffocated under the weight of conceptual
uncertainty in three critical areas. The first area of uncertainty concerns the meaning
of democracy as it relates to a tradition of civilian control of the military in a
democracy. The second area of confusion concerns the imprecision with which
political actors are analyzed for the purposes of determining the health of American
civil-military relations. The third area of uncertainty can be seen in the proliferation
and stubborn adherence to analytical models that attempt to answer the question,
"Is there more or less civilian control of the military in the United States?" This thesis

should adequately demonstrate and explain why such confusion exists in a majority
of contemporary literature about U.S. civil-military relations.
Since the publication of Huntington's seminal work in civil-military relations,
The Soldier and the State in 1957, academic discussion of U.S. civil-military relations
has revolved around a set of three principles. The first principle begins with the
universally accepted maxim that the military must be inarguably under the control of
civilians in any democratic polity. The second principle is that the military is capable
of posing a threat to democratic process both violentlyin the case of a coupor
peacefully, as dominant policymakers in the arena of national security decision-
making. The glue that binds these first two principles together is the acceptance of a
third principle: that civilians must formulate the political goals of any democratic
society, whether such civilians are elected or lawfully appointed governmental
It is my contention that all three of these discussion points for contemporary
civil-military relations scholars were inadvertently dictated by Huntington's earlier
work. By examining the relationship between the Constitution and how civilian
officials manage the military, Huntington began a process by which the
Constitutionand its provisions separating institutional powerbecame semantically
interchangeable with democracy, civilian control of the military became a principle
relevant only to government officials responsible for daily management tasks, and

the focus for scholars with empirical ambitions centered on answering the question
of whether or not there was more or less civilian control of the military in the United
States. It is my contention that this direction in civil-military relations scholarship
was based on mistaken assumptions about democracy, is unjustifiable, and has
resulted in virtual gridlock for those engaged in civil-military relations scholarship.
While there is undoubtedly good reason to subscribe to the principle that no
democracy can tolerate military domination in politics, there seems less reason to
suggest that military policy activity is necessarily undemocratic. It is my belief that
the American system of checks and balances often leads scholars to confuse military
policy participation with inappropriate political dominance by military leaders. The
traditional inability, or reluctance, of civil-military relations scholars to distinguish
between the provisions of the U.S. Constitution and other legitimate forms of
democratic government creates many problems for analyzing the health of domestic
civil-military relations.
The three principles of contemporary civil-military relations scholarship
manifest themselves as two dominant research questions for academics. Gibson and
Snyder (1998) articulate the central focus of most civil-military scholars when they
argue that "One of the leading questions ((of civil-military relations)) whether the
U.S. military is too influential at higher levels of...decision-making" (p. 194). The
other dominant analytical question, according to Thomas Ricks (1997), is whether or

not an increasingly isolated military culture can continue to unquestionably obey the
orders of democratically elected officials. Both of these questions are certainly
important as they relate to the integrity of democratic government Both of these
questions are highly contested by a variety of scholars. Both questions, however,
tend to be examined through smudged lenses forged from vaguely understood, or at
least woefully articulated, conceptions of what we mean when we refer to
"democracy", "civilians", and "civilian control of the military".
I will systematically analyze a variety of civil-military relations literature in
order to demonstrate where such conceptual confusion exists and the impact of such
confusion on analytical efforts by civil-military relations scholars. The quantity of
articles and books employed as cross-sectional evidence is highly representative of
the entire body of expert work in this area. My inclusion of several noteworthy
exceptions should assist the reader in understanding just how pervasive this
confusion is and how often it can be incorporated, sometimes mistakenly, into
analytical tracts that are frequently influential in policy-making circles in the United
The purpose of this thesis is to appeal for a temporary disengagement from
the standard, timeworn arguments that have permeated civil-military relations
scholarship for over forty years. The largest obstacle in this discipline is that
increasingly fossilized language, carrying with it much less substantive meaning than

it did when it was originally used, consistently serves as a restraint for scholars who
may wish to examine the core questions of civil-military relations from a fresh,
reinvigorating perspective. Incoherent conceptions of democracy in their most
general meaning, confusion over which civilians should be in control of the military
and what such control actually looks like, and scalar models developed as a result of
an institutional fixationin order to more "easily" measure complex variables, all
conspire to relegate civil-military relations scholarship to the trash bin of

Iconic Democracy in Civil-Military Relations Literature .
In 1956, Samuel Huntington wrote an article discussing the relationship
between civilian control of the military as a principle and the articles of the United
States Constitution. In "Civilian Control and the Constitution" (1956), Huntington
never uses the word democracy. The word democratic is only used twice. The first
reference is to the idea that "in a democratic society, the armed forces were
democratically organized" (Huntington 1956,677). The second reference concerns
an obligation of the citizenry to defend the nation. Huntington calls this a
"democratic principle". In this article, what characterizes relevant discussion of U.S.
civil-military relations, at least for Huntington, is the United States Constitution as a
particular governing document. There is no attempt to generalize a principle of
civilian control of the military that would apply to any arguable, or non-American,
vision of democratic government.
Since Huntington published this article, numerous scholars and experienced
observers have written on the topic of civil-military relations in the United States.
Some of these scholars appear to have a populist vision of the democratic process.
Others cling to Madisonian republicanism as the ideal democratic framework, and
still several others see democracy as a substantive political system that requires the

existence of specified egalitarian conditions. I will focus primarily on the first two
groups since they comprise the bulk of the authorship of contemporary civil-military
relations literature. My purpose in this section is to demonstrate the distinct versions
of democracy that exist in a variety of academic works.
Christine Scriabine (1983) is one of the authors who incorporate populist
democratic assumptions in their analysis. In an article explaining why some military
leaders run successfully for public office and some are rejected, Scriabine argues
that the issue of civilian control of the military is tied to the popular will. The military
are servantsfirst and foremostof the people. Robert Kaplan (1996) echoes this
political relationship between the military and the people in an article written from
the point of view of several military officers. Bill Kauffman (1999) writes about the
values of West Point cadets, and how those cadets (and he) see their particular
value system as being representative of a communitarian strain of democracy. The
collective values inculcated at the United States Military Academy (USMA) reflects a
democracy premised on cooperation, not individualism. Daniel Sutter (1999) refers
to democracy as a majoritarian process that is legitimated by popular will. More
importantly, Sutter sees the political relationship between the general population and
political leadership as continuous. The people, in other words, exercise constant
influence on the formal government beyond merely an election. Patrice McSherry
(1999) is one of the few writers I have found who specifically refers to the existence

of a "U.S.-style democracy", though she does not mention whether her vision of this
American form is a populist, majoritarian or republican assumption (McSherry, 2).
Andrew Bacevich (1999) approaches civil-military relations from a Jeffersonian
perspective when he discusses the democratic centrality of the citizen-soldier. Harry
Marmion (1971) echoes the Huntingtonian assertion that citizens have an obligation
to defend their nation. In the context of his book, The Case Against A Volunteer
Army, Marmion is decidedly populist in his assumption of what a democracy should
be. For these scholars, the primary civil-military relationship is between an
ephemeral "people" and the warriors serving on their behalf. One way of
consolidating the overarching undertones of Marmion, McSherry and Bacevich would
be to suggest that they view the military as a "people's army".
An alternative vision of democracy can be found in the works of authors who
assume that the American system of institutional checks and balances is, or should
be, the point of focus in evaluating issues of civilian control of the military. This
assumption is most frequently discovered in literature generated by scholars of
international relations writing on military matters, or domestic scholars closely
affiliated with national security institutions. Timothy Boylan (1999) explicitly ties his
discussion of civilian control to systems of checks and balances. Daniel N. Nelson
(1998) assumes that international discussion of military subordination to civilian
authority in democracies must somehow resemble the American manifestation of a

democratic regime. Nelson also discusses Morris Janowitz's argument that the
military must be excluded from involvement in domestic politics because the military
cannot serve as a legitimate check on civilian authority in the American system.
Ambassador Robert E. Hunter (1996), in a State Department Dispatch, assumes that
the U.S. model of democracy establishes universal principles that are
interchangeable with any vision of democracy. In a speech about civilian control of
the military "in a democracy", Hunter makes it dear that what he sees as necessary
conditions for a healthy relationship between the Polish Armed Forces and their
government can only be evaluated by U.S. principles of civilian control (Hunter, 2).
Allison Graham and Robert Beschel (1992) also discuss democracy as a singular
system, generally manifesting characteristics of the United States model. This article
is somewhat anomalous since it contains a discussion of the nineteenth century
American dispute between direct and representative democratic systems. The
authors, however, relegate that discussion to an historical reference that is no longer
relevant. James Burk (1999) contends that American democracy is elite driven, and
that evaluations of civilian control of the military invariably involve power balance
issues between the elite and the military alone. His is an essentially republican form
of democracy, but a democracy evolved from multiple democratic traditions.
John O' Loughlin (1998) and Joanne Freeman (1999), neither writing
specifically about the military, specifically criticize the imprecise notions of

democracy that can be found littering work in national security issues. William
Chitdck (1988) is one of the rare authors that explicitly approach the problem of
civil-military relations from multiple democratic traditions.
The relevance of including literature not specifically discussing civil-military
relations is to demonstrate a larger ignorance of multiple democratic traditions in the
larger field of international relations, which is the primary discipline from which civil-
military relations scholars emerge. The tendency of international relations scholars to
focus on a monotonic conception of democracy is a major problem with
contemporary civil-military relations research.
Aaencv theories and the meaning of civilian and civil
Robert Kaplan (1996) assumes that the term civilian in "civilian control of the
military" is referring to "managers" (Kaplan, 78). For him, civilian refers to any
civilianelected or appointed-^that the military is obligated to obey under the
principle of civilian control. Kauffman (1999) produces a narrative of civilian
authority that describes the relationship between the Secretary of War and the
military officers of the USMA. For Kauffman, one of the initial conflicts of civilian
authority was between an appointed cabinet officer and the military officers he was
tasked with overseeing. Boylan (1999) assumes a dualistic view of the term civilian
that refers to both elected and appointed officials, but adds the element of

conflicting authority between these officials as well as between military and civilian
leaders. For Boylan, who analyzes civilian control through the methodological
framework of agency, or principal-agent, theory, the term civilian refers to multiple
layers of principals. Ambassador Hunter explicitly defines the term civilian for his
audience. "In fact, in democracies, the idea of civilian control is quite different and
more fundamental: a people's acceptance that decisions concerning the full range of
defense policy must be under the scrutiny and ultimate authority of democratically
elected officials" (Hunter, 2). Chitbck suggests that civilian is a broad term,
encompassing both government officials as well as many civilians outside of the
government apparatus, particularly those affiliated with defense industries. Though
he is not arguing that the military obey an order from the CEO of a defense
contractor, he does suggest that civilian control often involves influence issues
between private sector civilian actors and the military. The relevance of this article is
that it introduces a murkier relationship of military influence on civilian authority by
adding a non-govemmental principal to the analytic mix.
Richard Kohn (1997) defines civilians as officials " the people."
One of the more prolific writers on U.S. civil-military relations, Kohn conveniently
extends his definition by referring to civilian officials "...elected or appointed by those
elected." (Kohn, 3). Christopher Bourne (1997) aptly describes the confusion
generated when the term civilian is not dearly specified. Regarding the debate over

the 1986 Goldwater-Nichols Act, Bourne characterizes Senator Goldwater as having
hostile opinions about the abilities of many civilian leaders to manage military and
national security affairs. Goldwater suggests that not all elected officials command
the same principled obeisance. For Bourne, there is a hierarchy of civilian legitimacy
to be obeyed. 1 Samuel Fitch (1998) makes the distinction between elected and
appointed officials, but adds the element of constitutional apportionment of power to
his idea of properly constituted civilian authority. To some extent, this point mirrors
an assumption of constitutional authority present in many arguments assuming the
primacy of a system of checks and balances.
Charles Dunlap (1992), one of the most frequently cited sources in civil-
military relations for his essay 'The Origins of the American Military Coup of 2012",
overlooks significant distinctions in democracy. He separates civilian authority
between popular will and elected officials and consequently paints a fictional
narrative of competing civilian factions that ultimately push the military into choosing
to abide by popular opinion and take over the reigns of government. Issues of
hyperbole aside, Colonel Dunlap inadvertently poses a control problem in cases
where legitimate civilian authorities not only disagree, but place vastly different
value on democracy. Paul Christopher (1995) defines legitimate civilian authority as
"legally elected officials", but then confuses elected and appointed officials by
referring to the Secretary of Defense as an equally legitimate source of civilian

authority as the Senate Chair of the Armed Services Committee (Christopher, 5).
Christopher Schnaubelt (1997) refers to legitimate civilian authority strictly in terms
of governmental agencies. There is, for him, no distinction between elected and
appointed officials when discussing issues of civilian control of the military. For
Sharon K. Weiner (1997), the preeminent civilian authority is Congress. Weiner does
not justify why Congress should be considered the most legitimate civilian authority
in military affairs and national security management. She simply assumes it in her
agency theory analysis of oversight mechanisms. Sam Sarkesian (1998) refers to
legitimate civilian authority as emanating from "civilian councils" in his argument that
military leaders have the duty to engage in political discourse.
One of the clearest explications of what the term civilian means, particularly
for adherents to agency theory as the best framework of analysis, is Peter Feaveris
description of the legitimate role of delegation in civil-military relations. Feaver
(1997) accepts legitimate civilian authority as both elected and appointed, but
introduces the more sophisticated concept that the delegatory powers of the two
types of civilians are not at the same level of legitimacy. David Isenberg (1988)
approaches the civil-military relations kettle with the Huntingtonian idea of objective
control of the military, which means that military leaders are expected to be
politically neutral. Isenberg generally assumes that civilian means any elected or
appointed civilian officials. David Johnson (1996) and William Gregor (1996) also

assume an expansive meaning to the term civilian. The level of democratic
legitimacy of civilian authority, for these two scholars, is identical for elected and
appointed officials.
An interesting twist on the question of what constitutes civilian authority is
presented by Mady Wechsler and David Segal (1983), when they argue that military
officers in leadership positions (at the national level) are virtual civilians as a result
of what their job responsibilities are. For them, the term civilian approaches a job
description. If a soldier is engaged in a position that requires political interaction with
civilian policy makers, the soldiers shed their role as warriors and become another
legitimate group of appointed civilian officials who just happen to wear uniforms.
The aforementioned articles are fairly contemporary representations of the
definitional assumptions in U.S. civil-military relations literature. If one is willing to
go back to the founding of the United States, it can be seen that the evolution of
what Huntington referred to as the "cliche" of civilian control of the military has its
roots from the ratification debates on the United States Constitution. Further,
references to civilian control of the military are as amazingly imprecise as they are
Alexander Hamilton, writing in Federalist No. 24. implies that the ultimate
authority over the military resides with Congress. By advocating the constitutional
clause setting the maximum budgeting for a standing army at two years, Hamilton

would suggest that the responsibility for civilian oversight of the military rests with
Congress. This observation is generally mirrored in contemporary agency theory
literature focusing on congressional mechanisms of oversight. What is lesser known
is Hamilton's distinction between responsibility and control. Whereas the material in
The Federalist demonstrates Hamilton's desire for a standing army, it does little to
shed light on the subtle nuances of civilian authority in military matters. Much like
Huntington began by analyzing the constitutional military clauses and concluded that
civilian control was spread amongst branches of government, Hamilton assumed that
real military control was more than just the power to raise and disband a standing
army. There are issues, for him, of control on the battlefield and the creation of
strategic and tactical plans. This introduces a complication of the term civilian as it
relates to the act of control.
Richard Brookhiser (1999), in his biography of Hamilton, describes Hamilton's
appeal to George Washington at the outset of the Whiskey Rebellion. Washington
had already turned down one offer to participate in the suppression of an
insurrection. When Henry Knox had asked Washington to march at the head of a
militia force against Daniel Shayes, Washington dismissed the request by mocking
the idea that his influence as a military commander was necessary to prevent Shayes
from stalling further foreclosures of farms. In his response, Washington indicated
that " not government" (Brookhiser, 116). General Washington took a

different approach to the Whiskey Rebellion. As the sitting President, Washington
saw his duties as COmmander-in-Chief as virtually military duties (Brookhiser, p.
118). Hamilton took every effort to convince President Washington that the roles
were identical. As a result, instead of ordering Hamilton or Henry Lee to march out
and put down the tax revolt themselves, Washington abandoned the "civilian" sense
of the presidency and became another figure: the supreme military leader, or "top
general", of the United States. From the beginning, it seems dear that there were
often cases where civilian authority stopped when the power of the commander-in-
chief was exercised. In short, Washington did not necessarily view the position of
commander-in-chief as strictly "civilian".
Another confusing assumption of civilian control originated with Jeffersonian
republicanism. Jefferson initially believed that legitimate civilian authority was vested
in state governments. This divergence remains important today, for those who
subscribe to a legitimacy of state authority frequently assume that state governors
and assemblies have certain control authority over state militias. William Watkins
(1999) describes the difference of the Jeffersonian view of civil-military relations in
much the same way. For Hamilton, the legitimate civilian actor of control was first,
the president as commander-in-chief and then Congress, as the oversight
mechanism for the President. For Jefferson, the legitimate civilian authority was the
governors of the states themselves as well as "the People". In spite of the fact that

Jefferson's presidential record indicates an abandonment of this philosophy, this is a
considerable difference in meaning when applied to the term "civilian control of the
Amar and Hirsch (1998) imply that most of the difference of opinions
regarding a civilian hierarchy in military management and the U.S. Constitution result
from differing interpretations of the meaning of the militia clauses in the
Constitution. Though they do not discuss civilian control of the military perse, these
authors adroitly demonstrate why assumptions of democracy affect the
interpretation of the legitimacy of civilian authority relating to military issues in their
discussion of the draft and a popular right to bear arms.
Between the nation's founding and the present, assumptions of what civilian
control of the military mean abound in different types of literature throughout the
nation's history. Joanne Freeman (1999) addresses the evolution of a murky, popular
meaning of democracy beginning with the election of 1800. Shifts in support for the
Republican or Federalist positions during this time were indicative of how democracy
should look and who should be considered the legitimate civilian authority in it. John
O'Loughlin (1998) discusses the divergence between contested ideas of democracy
within the United States and increasingly iconic democratic assumptions for those
engaged in the study of international relations. For O' Loughlin, the act of comparing
democracy against non-democratic forms of government has instigated the shift to

this unitary meaning of democracy. John May (1980), discusses how the differences
in democratic preferences affect policy outcomes as well as evidentiary
interpretations. He concurs with O'Loughlin that one of the reasons for the drift to a
dominant vision of democratic government was the tendency to consolidate
democracy for the purposes of comparative analysis with non-democratic states.
From the establishment of West Point, to the Civil War and the Korean
Conflict, military and civilian leaders alike have discussed civilian authority in vague
and expansive terms. In three editions of the United States Army's Soldier's
Handbook (1884,1891,1908), the pamphlets begin with a general reference to the
principle of civilian control of the military when it reminds soldiers that they must
obey, and be respectful to, "civilian authorities" (1884 Handbook, 2:1891 Handbook,
5). When General Matthew Ridgway assumed command of the United Nations forces
in the Korean War, replacing Douglas MacArthur in one of the more infamous
examples of civilian control issues, he explicitly states that the duly constituted civil
authority for all forces is the United Nations (Ridgway Memorandum to U.N. Forces,
Joseph Dawson III (1996) writes about the problem of civilian authority for
military governors in conquered territories, specifically in the Mexican War. Though
Dawson specifies the typical legitimate sources of civilian control, he introduces
another historical agent of civilian authority over the military: the Supreme Court

Dawson notes how the military governor of Mexico, Colonel Alexander Doniphan,
was compelled to obey a Supreme Court ruling regarding the confiscation of Mexican
property. This ruling was in direct opposition to the political instructions from the
executive branch that such property could be confiscated.
Much of what I am trying to demonstrate here is summed up adequately by
David E. Albright (1980), who argues that the legacy of Huntington's work was to
compel future scholars of civil-military relations to view civilians as a group "opposed
to the power of the military" (Albright, 554). Though I would trace this phenomenon
back much further in American history, the point is well taken. It seems as if
scholars of civil-military relations in the United States have constructed a branch of
government called "civilians", which opposes and checks another branch of
government called the "military". In some cases, a few scholars have readily applied
the checks and balances argument the other way around, suggesting that perhaps
this "military" branch has a legitimate checking role of its own.
One should get a clear sense that there are three distinct categories of
civilians that can be assumed in any discussion on civil-military relations. The first is
the civilian as a democratically elected official. The second is the civilian as a lawfully
appointed official. The third is the idea of civilian that most closely corresponds with
the term, civil. This third civilian is the American people themselves.

Scalar Measurement in Civil-Militarv Relations
In this section, I will present several sources that implicitly or explicitly rely
on some scalar, or linear, method of evaluation. Most scholars tend to evaluate any
problem of civilian control of the military as "more or less" of such control. This is an
inherently linear proposition. All but one of the following references has been cited in
the previous two sections of this chapter.
Scales of measurements generally have polar extremes that are used to
describe movement along the line(s) on the scale itself. In civil-military relations,
there is a general consensus on what one endtypically referred to as the ultimate
negative state of civilian control of the militaryis. This is the certainty, imminence
or actuality of a military coup. That is the first referent. The second referent is the
problem area, particularly for U.S. civil-military relations. There are a variety of
visions describing, however clumsily, what the ideal state of civilian control of the
military in a democracy is. Most literature in which claims are made as to the state of
civil-military relations uses the coup as one referent, and something else as the polar
opposite. As we will see, this unidentified "other" can be a serious problem.
Daniel Nelson sums up the dilemma nicely. He says, "...principal emphasis is
placed on how a state and its army behave. No one defines such norms with any
precision, but everyone seems to think they exist"(Nelson, 137). Nelson argues that
you can somehow invent a "prototype" civil army and use it as a comparative tool.

Another scholar emphasizing comparative evaluations is David Albright. He resists
the temptation to construct an arbitrary linear model, though he ends up
constructing another arbitrary comparative model by advocating the use of a Soviet
participatory model with which to challenge Huntington's universalist assumptions
about civilian control of the military in democracies (Albright, 557). Feaver proposes
a "deductive model" of civil-military relations, which suggests that you must decide
three issues under conditions of uncertainty. The importance of Reaver's suggestion
is the introduction of the uncertainty principle in defining an referent opposed to the
military coup (Feaver, 3).
At first glance, it seems that some scholars are aware of the modeling
problem in evaluative literature of civil-military relations. This is undoubtedly
accurate. I will argue that these acknowledgements are insufficient in addressing the
modeling problem, as the first two authors' use of comparative analysis eventually
end up scaled models with identical referent problems and that Reaver's model,
while the most useful, presents additional problems associated with heuristic
Fitch's work, like that of many serious scholars of Latin American military
regimes, presents one of the more sophisticated linear models available. His
referents for analysis are based on the idea of "military intervention in a democracy".
Though the term "intervention" can prove problematic, it suits my purposes here to

point out that this is still a linear model with the same referent problem (Fitch, 30).
Bourne's model of analysis is evaluating "more or less" civilian control along a linear
scale that uses the coup as one pole. His predictive efforts are focused on the
movement along the scale in terms of alterations in the structure of checks and
balances, with no referent for the polar opposite (Bourne, 5). Another scholar with
evaluative and predictive pretensions is Richard Kohn, who unabashedly engages in
linear analysis with one very important exception. His model is institutional, and
movement along his scale cannot be used for evaluation unless institutional
responses to events or crises in civil-military relations are taken into account Like
Feaver, this linear model mutates into a heuristic analysis of institutional behavior
and still offers no tangible referent opposite the military coup (Kohn, 2).
William Chittick relies on statistical modeling of social attitudes that he agues
are evidence of "militarism". It is a linear model in that it seeks to evaluate the state
of civilian control of the military in terms of degrees of militarism. The problem of
what, precisely, constitutes a militaristic attitude presents many of the identical
heuristic analysis problems to which some of the aforementioned scholars are
vulnerable. Thomas Ricks (1997) adopts yet another linear model predicated on the
analysis of voting patterns by the military. His poles are more definite, with one end
being complete support of the Republican Party and the other end complete
disengagement with the same party. I will ultimately argue that Ricks' referents are

false ones in the context of his own analysis. Christopher Gibson and Don Snider
(1998) present the last linear model that attempts to adopt clearly defined scale
referents. These two scholars suggest linear analysis based on agency theory, but
using the evaluative measuring of "issue networks" in determining erosion or
enhancement of civilian control of the military.
Gregory Foster's analysis of U.S. civil-military relations closely resembles
Professor Kohn's model. His model is linear, but heuristically broad. Though it has
the same referent problems of other linear models, Foster implicitly criticizes the
predominant method of linear analysis that emphasizes taking "snapshots" of points
on the scale as a basis for evaluative claims. Finally, Sutter comes the closest of
anyone to actually defining his referents, however broadly. His "negative" referent is
the certainty of a coup, and his positive referent is the impossibility of a coup.
Though these referents are highly arguable as to their usefulness, it seems
interesting that of all the authors I have examined, Sutter is the only one to
expressly point out both of his referents of measurement.
My point in this section is to briefly explain how scholars of civil-military
relations tend to examine their subject by means of scaled evaluation. Though there
are tremendous variations that, I believe, express an implicit understanding that the
measurement system used by analysts emphasizes the primacy of civilian
governmental control of the military, not civilian control of the military. This

seemingly minor distinction may prove to be the undoing of judicious civil-military
relations research.

The overarching question for scholars of American civil-military relations is
"What kinds of military attitudes, behaviors and policies can constitute a real threat
to the Republic?" A secondary question, often posited by international relations
literati, is "What military attitudes, behaviors and policies can constitute a threat to
democracies?" Scholars belonging to each group frequently juxtapose the distinction
between the American political system and a more general concept of democracy. In
spite of the liberal interchanging of "democracy" and "republic" in literature, the
focus of both groups is aimed directly at potential military threats to any democratic
The nature of these threats, as we have seen, meanders from possibilities of
a military coup to so-called "inappropriate" public statements by military officials.
Given the fertility of such a vast spectrum in which to conduct analysis of U.S. civil-
military relations, it is unsurprising that little discussion is devoted to the specifics of
democracies that military political participation or influence is suggested to threaten.
Why, for example, did General Colin PowelTs 1992 op-ed piece in the New York
Timescriticizing a potential deployment of peacekeeping forces to Bosniaspark
such a heated debate amongst scholars and journalists alike (Gregor 1996,1-2)?
There is a loose consensus among civil-military relations researchers that the Powell

case demonstrates negative military behavior in the realm of democratic politics
because, for reasons not always clear, the general had some responsibility not to
infuse his opinions into the public discourse over a policy proposal. The target of the
military threat, in this case, could reasonably be suggested to be the integrity of the
president's constitutional authority to command the military. This, according to
Gregor, Bourne and many others, is a violation of the professional military ethic of
non-politicization. In all of this discussion, however, another salient question goes
relatively ignored: What part of the democratic dynamic did this public advocacyif
it can be called thatnegatively affect?
Answers to this question are often left to the imagination of a reader of these
articles. It seems reasonable to infer that what some scholars were decrying was the
relationship between a general and the president. In other words, it could be
suggested that the concern generated by Powell's public statement served as an
influential, and contrary, counterweight to the policy preferences of his
constitutionally and legislatively designated commander. My inference here,
however, is just that. In no tract covering this incident was this connection made
dear by any author. It is simply my best guess.
I believe that it is fair to argue that perhaps these writers assume that their
respective audiences share their understanding of democratic processes and
characteristics. It may even be the case that most of these authors simply don't feel

it relevant to spell out the relationship between a public statement and civilian
control mechanisms. If the former assessment is accuratethat the audience is
attuned to the author's democratic visionthen it seems as though my grievance is
trivial. If the latter observation is closer to the truth, then I would argue that the
authors have collectively failed to prove their cases either way. For those writers
furious with Powell's comments, they fail to demonstrate real harm to democracy by
not making the basic connections between action and consequence. For those
writers who believed that Powell's comments were benign, this same rationale
There is, however, a third possibility that may explain the complete lack of
connection in articles on the Powell incident. This possibility suggests that perhaps
there are some authors unaware of the multiplicity of meanings that democracy
carries. What I mean is that some of these authors may be operating with such
ingrained assumptions about democracy in general that they do not know what their
options are. They do not knowor do not incorporate into their analysisthe
possibilities posed by competing democratic traditions and rules that may affect their
conclusions drawn from the Powell op-ed piece. The heart of this chapter is devoted
to a closer examination of this third possibility. I believe that the confusion
surrounding the relationship between military action and negative (or positive)

consequences results, in part, from vague understandings about democracy's actual
elements and characteristics.
The "Powell Incident" serves as an exceptional contemporary event with
which to deconstruct the implicit relationships between the military threat and the
threatened polity. Edward Westermann, while acknowledging the inappropriate
nature of Powell's public remarks on Bosnia, concludes that such remarks do not
constitute a threat to civilian control of the military. Further, he argues that such
remarks are not manifest evidence of a threat to democracy in general, even if such
behavior goes unpunished by civilian authorities (Westermann 1995,79). Richard
Kohn, on the other hand, believes that Powell's conduct revealed a more sinister
trend in U.S. civil-military relations by arguing that the incidentas part of a larger
and increasingly frequent phenomenon in U.S. politicsis a contributor to the
gradual erosion of civilian authority in national security matters and thus threatens
both civilian control of the military and democracy (Kohn 1997, 9).
Why do these scholars disagree in their interpretation of the same event? I
believe that the primary root of disagreement between the two takes us back to
Huntington's distinction between objective and subjective civilian control of the
military. As noted previously, objective control is essentially defined as a politically
neutral, professional military devoted to the value of obeying orders from civilian
masters. Subjective control is characterized by a symbiotic value system between

military leaders and political elites that minimizes the possibilities of military
usurpation of power due to the political similarities between civilian and military
leaders. Perhaps the most enduring legacy of Huntington's The Soldier and the State
is the confusion over what military professionalism is supposed to look like. For
Huntington, there is but one overriding empirical test for the presence of adequate
objective control: Do the military follow their lawful orders (Huntington 1957, xiv)?
This is a relatively simple question to apply as an empirical test of control. However,
contemporary civil-military relations scholars have made assumptions about what
constitutes a professional set of ethics for the modem military, and therein lies the
Westermann subscribesfor the most partto Huntington's belief that
civilian control is best expressed when it passes the test of obedience. May the
military disagree with a president on a course of action? Yes it may. May it refuse to
obey a lawful order? No, it may not Though Westermann condemns Powell's
remarks as being "unprofessional", he does not link this lapse of professionalism to
any probability that an order will be refused. Kohn also agrees with Huntington's
definitional premise of objective control, inasmuch as the military should possess a
strong ethic of political neutrality. It is Kohn's expansion of the concept of neutrality
that distinguishes his analysis from Westermann's. For Kohn, civilian control of the
military is best expressed when all policy matters at the strategic and political level

are developed and debated between civilian officials. Military input, then, should be
strictly limited to tactical developmentareas of military professional expertise (Kohn
The locus of disagreement between Westermann and Kohn, therefore,
illustrates two things. First, both authors accept, to varying extent, Huntington's
standard of objective control. Second, neither author specifically addresses any
portion of the democratic dynamic that is threatened by such remarks, remaining
content to debate the merits of what constitutes sufficient neutrality. There is, for
example, no argument made by Kohn as to how public comments by General Powell
affect power balances between political institutions. Westermann, for his part, does
not indicate how Powell's comments act benignly on specific power arrangements in
a democratic polity. I will argue that the reason such language and analysis is absent
from the works of these two scholars, in addition to a majority of other researchers
in the discipline, is due to a widespread assumption that democracies should look
very much like the American political system. Democracy, in other words, is assumed
to have the same procedural look (such as elections), institutional arrangements
(parliaments and legislatures vying for power with executives and courts) and values
in any society that claims to possess such a government.

The Myth of Single-Tradition Democracy
To utilize an admittedly fantastic example, imagine that the people of the
United States held some form of plebiscite to decide just who they wanted to
develop military strategic doctrine. By some quirk of fate or conscious desire, the
people designate the source of military doctrinal development to be the Joint Chiefs
of Staff of the Armed Forces. In spite of this peculiar new distribution of political
authority, no further alterations are made to the legal fabric of the American
government, with the President retaining the powers of commander-in-chief and
Congress its powers under Articles II and I respectively.
Setting aside the question of constitutional legality for the moment, would
such an event prove to be a violation of the principle of civilian control of the military
in theory? Would this event simply eradicate the tradition of civilian control of the
military altogether, or would it merely redefine the source of civilian authority away
from institutions to popular mandate? More succinctly, would such an event be an
enhancement or threat to democratic principles? It seems clear that the answers one
would give to these hypothetical questions would be predicated on how one views a
hierarchy of legitimacy in democratic theory and practice.
Rarely do dedicated theorists of democratic politics confine characterizations
of democratic governments to particularly Madisonian terms. A cursory examination
of Robert Wiebe's survey of literature devoted to defining what a democracy is or

isn't reveals an interesting absence of Madisonian constructions of democracy
(Wiebe 1995,2-6). Amar and Hirsch mention contemporary fixations by legal
scholars on the so-called "Madisonian dilemma", though in all of this discussion
phrases such as "separation of powers" and "checks and balances" are noticeably
lacking. For Amar and Hirsch, a major characterization of liberal democracy is the
tension between majority rule and the protection of individual rights. The democratic
dynamic, for these two scholars, is between the citizens and the government, not
between institutions (Amar and Hirsch 1998, x-xii). According to Wiebe's
interpretation of Robert Dahl's egalitarian polity, democracy is a system whereby
individual citizens have " equal opportunity to prepare the agenda of public
issues, equal rights in determining the mechanisms for decision-making (italics
mine), equal access to information..." and so forth (Wiebe, 6). Using Dahl's
description as a guide, the scenario of having a plebiscite transferring power to non-
elected officials seems to be a legitimate democratic exercise. Amar and Hirsch echo
several components of Dahl's egalitarian democracy by suggesting that the primacy
of the Constitution is misplaced, and that the planks of the governing document
areand should beopen to public alteration through majoritarian decision (Amar
and Hirsch, x-xiii).
Distinctions between philosophical and procedural democratic characteristics
are nothing novel. In Transitions from Authoritarian Rule: Comparative Perspectives.

Lawrence Whitehead cogently delineates between the cultural democratic traditions
of Europe and the United States: "European definitions of democracy seem to give
more stress to soda! and economic participation, whereas the Americans give almost
exclusive emphasis to the electoral aspect" (Whitehead 1986,17). This appears to
be the first qualitative distinction between democratic systems that Whitehead
draws. He submits another possible distinction; this one, between both European
and American democratic tradition and "Third World" emergent democracies.
Whitehead argues that "...when the leading liberal capitalist governments attempt to
promote democracy internationally, it is the first variant ((of democracyprivate
ownership and anti-Communism)) that meets with their strongest approval"
(Whitehead, 8). Continuing, he suggests that the differences between Socialists and
Capitalists hinge on the former's rejection of democracy as being purely procedural,
similar to Dahl's contention that there is an implicit egalitarian promise made by
democracy (Whitehead, 14-15). Thus we are left with distinctly differing accounts of
what democracy may or may not be. Democracy may be purely procedural or it may
be substantive, with an overt egalitarian bent. It may express a relationship between
popular will and representative institutions, or it may concern popularly constructed
institutions that interact with each other antagonistically. When the word democracy,
then, is bandied about, it carries multipleoften-contradictory^-meanings.

The political tradition of separating institutional powers raises yet another
interpretive problem for precisely identifying the characteristics of a threatened
democratic polity. Bill Granstaff articulates a traditional view of the role of
representation in the formulation and execution of public policy with an important
...the American constitutional system's "democratic connection" requires that every
duly-elected representative possess equal opportunity to influence what the federal
government will do and why. To compromise ((this)) tenet is to fundamentally
compromise its system's democratic spirit Obviously, to systematically exclude
qualified parties, voters, or representatives is to violate an essential criterion for
democratic processes" (Granstaff 1999,7).
Unlike Dahl, Wiebe or Amar and Hirsch, the Granstaff democratic vision focuses on
the relationship between representatives and legislative policy outcomes. Further,
the definition he provides of representative government seems paradoxical when he
mentions the sanctity of not excluding "qualified parties, voters, or representatives".
It can be assumed that what Granstaff means is that it would be unacceptable to
restrict suffrage to groups of citizens, but notas Wiebe pointed out-H:o restrict
their control over the mechanisms of the representative polity they support with
their votes. Granstaffs lack of clarification throughout his book is a testament to the
typical confusion surrounding the meaning of the word democracy. Without
elaborating as to why policy influence should be restricted to representatives as a
matter of course, Granstaff falls prey to the same mistaken assumptions that most of
the authors previously quoted have.

Aaron Wildavsky presents what I think may be the most obvious explanation
for the prolific confusion surrounding the meaning of democracy. For Wildavsky,
American history is riddled with contradictory democratic traditions, many of which
are mutually exclusive with other traditions. Beginning with the colonial charter
system, followed by the Articles of Confederation and then the U. S. Constitution,
Wildavsky suggests that distinct permutations of democratic principles have traveled
a parallel historical path throughout American history. Each of these various
compactsor compact typesrepresents competing theoretical propositions, though
each is undoubtedly democratic (Wildavsky 1998, 122-123). Additionally, the fact
that each of these types of legal frameworks was in turn superseded does not
diminish the pervasiveness of their respective legacies. The Jeffersonian slant of the
Articles of Confederation, with the corresponding emphasis on state autonomy within
a loose system of cooperativeoften voluntaryrules, seems diametrically opposed
to the collection of national power embodied in the Constitution. For those inclined
to think that the historical significance of the Articles is a political artifact, consider
the subsequent adoption of the Ninth and Tenth Amendments as well as
contemporary arguments for or against states' rights that occur frequently in
mainstream American political discourse. This clash is by no means a stroll down
memory lane, but a vital, living political conflict with origins dating back to colonial
America (Wildavsky, 124-125).

The apportionment of governmental power under the U.S. Constitution is but
one facet of American political and democratic tradition. As noted in Chapter Two,
this seemingly simple observationthat democracy is often a crazy quilt of
traditions, icons and imageryis routinely ignored in civil-military relations literature.
Such literature generally focuses on a military threat to institutional power and
relationships but rarely on the target of such a threat: the democratic polity of the
United States as expressed in alternative democratic views. Quite often the
consequences of assuming too much about what democracy is or isn't are trivial in
the study of civil-military relations. Occasionally, however, the outcome can border
on grave, particularly when such vaguely informed or articulated conceptions of
democracy translate into drastic legislative action, such as the Goldwater-Nichols Act
of 1986. For the moment, suffice it to say that as a question of legitimacy, there is
no reason to believe that the majoritarian democratic heritage is any less worthy of
consideration than the institutional relationships promulgated by the Constitution.
The Relationship Between Military Policy Influence
and the Separation of Powers
In the literature review, I noted that several scholars do not study civil-
military relations solely under the assumption that democracy is first, American,
secondly, institutionally antagonistic. The first order of business in this section is to
distinguish between the types of analyses engaged in by those who utilize

alternative democratic frameworks and those who are decidedly Madisonian in
orientation. As mentioned in the introduction, there are two prevailing types of civil-
military relations analysis: the first is research that attempts to ascertain the health
of U.S. civil-military relations inasmuch as the possibility of a coup exists. This
analytical approach generally encompasses research exploring an ostensible social
gap between the military and the remainder of American society (Ricks 1997,2).
The aim of this type of analysis is to determine the conditions under which a
sufficiently alienated military would attempt to grab the reigns of power by force. I
classify this type of research as internally consistent, meaning that vague definitions
of democracy are usually irrelevant since a coup is inarguably antidemocratic
regardless of any theoretical distinctions between types of democracies. The second
analytical approach in civil-military relations literature is much less cataclysmic in
orientation. This type of analysis focuses on military influence in the policymaking
process. For the purposes of this thesis, I will label this kind of research as internally
inconsistent My motivation for this label is that this second type of analysis-
predominant among domestic civil-military relations scholarsis where most
assumptions of what democracy is or isn't have the greatest potential negative
Scholars who implicitly or explicitly perceive democracy as a political system
with multiple traditions and characteristics are generally focused on the possibilities

of a military coup, typically within a longer-term context. It seems clear that
wresting power from democratically elected officials, or the people themselves,
meets any reasonable criteria for an antidemocratic act. This does not mean that
scholars fixated on the American constitutional tradition do not engage in this type
of analysis from time to time. Thomas Ricks (1997) and Ole Holsti (1997), both of
whom can reasonably be categorized as Madisonian-oriented scholars regarding
democratic norms, have worked extensively with the issue of an increasing social
gap between civilians and the military. Scholars who investigate both influence and
value issues frequently drift into both analytical camps. My concern is with the
influence analysis. Influence arguments tend to be unstable, given their myriad of
imprecise assumptions of democracy. When translated into legislative action,
influence analyses are far more flawed than research that studies coup possibilities,
because coups are inarguably antidemocratic and no legislator is going to argue that
point, even though the specific evidence for any case may be highly contested.
Influence arguments, however, are primarily dependent upon the democratic
framework chosen as a comparative baseline when determining the potential threat
posed by military policy encroachment. In structuralist terms, the primary difference
between coup and influence scholarship is that the latter is not satisfied with merely
avoiding the possibility of military rule, but requires that the military not influence
institutional power distribution to favor one branch over another. This second

pursuit frequently elicits claims of healthy or unhealthy civil-military relations based
solely on perceptions of power alignments between the three branches of American
government (Robert Hahn 1997,4-5).
Within the community of scholars who study military influence on the
policymaking process, there are two dominant questions that are asked: the first is
whether or not military influence in strategic policy is so dominant that it constitutes
a fourth branch of government, and the second is whether or not military influence
is sufficiently inappropriate as to favor one branch over another. The 1986
congressional debate over the passage of the Defense Reorganization Act of 1986,
or the Goldwater-Nichols Act (GNA), provides an excellent example of both of the
above influence concerns. The provisions of the GNA were designed to strengthen
the positions of both the Secretary of Defense (at the expense of the civilian service
secretaries) and the Chair of the Joint Chiefs of Staff (at the expense of individual
service chiefs of staff) (Isenberg 1985, 8). The GNA provisions established the JCS
chair as sole military advisor to the President, Congress and the National Security
Council, thus eliminatingor mitigatingthe direct interaction between other service
chiefs and these bodies. The purpose of this organizational change was, in the minds
of the act's supporters, to consolidate military advice and provide a more unified
military opinion that civilian policymakers could rely upon without suffering through

what Barry Goldwater considered endless service game playing due to intense
service rivalries ([Douglas Lovelace 1989,3-4).
The debate that the GNA deliberations provoked was rancorous, to say the
least. Christopher Bourne, for example, decried the changes wrought by the GNA as
movement towards a military general staff. For Bourne, this possibility provoked
imagery of Nazi Germany and Fascist Italy since he believes that such a
concentration and unanimity of military advice historically has led to an increase in
the propensity to go to war (Bourne 1994, 5). The ostensible increases in military
efficiency also concerned scholars like Bourne, who interpreted the search for unified
military command as an antidemocratic tendency that eliminates the give and take
of policy discussion, both between Congress and the military as well as amongst the
military themselves (Bourne, 7). Other researchers take a more benign view of the
GNA's passage. David Isenberg suggests that the consolidation features of the act
have no effect on civilian control of the military. To begin with, argues Isenberg, a
stronger JCS chair does not limit the policy options of civilian leaders, most of who
have their own specialized staffs to provide alternative perspectives from the military
viewpoint. Second, Isenberg cogently notes that there are no institutional changes
resulting from the act that funnel power to or away from one branch to another. He
concludes his argument by suggesting that academics like Bourne ignore the
confusing effects of autonomous military command in the previous defense

organizational structure, a system that Isenberg thought encouraged service leaders
to misrepresent the truth in order to gain a competitive advantage over the other
services (Isenberg, 12-14: 25-26).
Whether or not one agrees with either of these positions is irrelevant for the
purposes of this thesis. The salient point is that both scholars orient their arguments
concerning enhanced or eroded civilian control around institutional relationships. The
primary gripe for Bourne is that somehow, the GNA affects the separation of powers
to the advantage of the executive branch, an argument echoed by Sharon Weiner
(Weiner 1997,16). Isenberg, for his part, argues that such an imbalance has always
existed and that power relationships between branches historically fluctuate, with no
real impact on the integrity of the federal system. Neither of these positions
regarding the GNA reflects a democratic dynamic that suggests the legitimacy of
military political participation if such participation is urged by the people themselves,
or by an act of legislative or constitutional delegation. If either of these scholars
were confronted with the hypothetical example beginning this section, it is likely that
they would scratch their heads and cry "foul!" since there is no precedent for a
popular referendum on specific constitutional duties.
This brings me to another analogy demonstrating how confusing the
assumptions of democracy can be in asserting the relative health of U.S. civil-military
relations. What if, for example, the Secretary of Defense orders the chair of the JCS

to craft a military strategic doctrine with minimal or no input from civilian officials.
Suppose the order sounded something like "General, I want you to draw up some
guidelines on when you think military force should or should not be applied based
solely upon the standard of probable success". In other words, the Secretary asks
the JCS chair to basically instruct him as to what operational conditions the military
thinks are or are not "winnable". The JCS dutifully performs this task, returning with
a document spelling out not only strictly military details as to what kinds of battles
should be fought or avoided due to capability, but also some advice on how different
political contexts can affect the outcome. The report concludes that "political
conditions dictate that the application of force be so overwhelming as to guarantee
victory with a minimum of casualties". The Secretary approves of this policy and is
supported by members of the NSC and the President. Has the military overstepped
its bounds and thus eroded civilian control of the military by being the source of
what could be construed as overtly political advice?
The Problem of Delegation
The previous hypothetical situation actually occurred during 1988 with the
development and subsequent adoption of what is euphemistically known as the
"Powell Doctrine". A great deal of literature has been devoted to studying this event
Some of the literature argues that Powell overstepped his authority by constructing

what amounts to political doctrine concerning when to apply force as a tool of
foreign policy (Kohn, 13). Additionally, many other scholars have argued that since
civilian leadership delegated the task it could not possibly be a usurpation of civilian
mastery of the military machine (Gregor 1996, 18). What is fascinating is that there
is no literature devoted to Powell's options in this case. To alter the analogy
somewhat, suppose General Powell had refused such a directive? According to
Bourne's standards of civilian control, such an actthough explicitly illegalwould
have served to enhance civilian control of the military by refusing to accept what
must be seen as an illegitimate abdication of civilian responsibility. Strictly speaking,
it would not matter that such a control enhancement would have come at the
expense of defying a lawful order! Since Huntington's standard of civilian control
indicates that ultimately, the military may not disobey any lawful order and remain
"controlled", Isenberg would conclude that this hypothetical refusal would be
tantamount to a policy coup. In the parlance of civil-military relations scholars, this is
what is known as the problem of delegation.
Delegation arguments, exceedingly prevalent in literature concerning U.S.
civil-military relations, are critical portrayals of the checks and balances bias in the
discipline. They are institutionally oriented arguments that tie the health of U.S. civil-
military relations directly to the concept of the separation of powers. Les Aspin, prior
to his appointment as the Secretary of Defense, argued that both the Congress and

the President were pleased with the delegation arrangement provided by the GNA's
organizational changes (House Report on the Gulf War 1991, 31-33), Robert Previdi,
on the other hand, thought that such delegation of new powers to the military
constituted an abandonment of constitutionally apportioned power (Robert Previdi
1988, 9).
The influence arguments embodied by the GNA congressional discourse
clearly demonstrate the assumption fixation on the American structure at an
institutional level. This may be all well and good if the claims made in all of these
analyses were restricted to such a framework. This is generally not the case, as we
have seen. Most of the time, the argument may be that the GNA is bad because it
concentrates too much power in the hands of the JCS chair, but the claim made is
that either democracy is endangered or civilian control of the military sui generis is
being jeopardized (Albright 1980, 573-74). There is no discussion of civilian control
of the military that focuses on the threat posed by the military to any other aspect of
democracy other than institutional power balances. As David Isenberg's analysis
implicitly suggests, the type of the democratic system one assumes to be the
threatened target can drastically affect one's assessment of civil-military health.
Isenberg, in response to those that feared that the GNA would create a de facto
general staff, cleverly points out that if America utilized a parliamentary system of
government, the fear of an inordinately powerful general staff would be nonexistent,

since such a model would then more closely mirror Great Britain and Israel rather
than Nazi Germany and Fascist Italy (Isenberg, 21).
As a result, for a scholar to claim that the formation of an American general
staff necessarily threatens democracy is an obvious overgeneralization. While it is
arguable whether or not such an arrangement poses a threat to the institutional
configuration imposed by the U.S. Constitution, it is virtually inarguable that such a
threat to the separation of powersor even to the constitutional apportionment of
poweris not the same thing as a threat to "democracy". It doesn't even necessarily
threaten American democracy, since by rights the Constitution is subject to
amendment and alteration, both legally and morally. This distinction is, as I have
argued, often lost in the confusion over what military policy influence can negatively
affect in the democratic dynamic and the confusion between one permutation of
democratic government and another.
How Democratic Assumptions Affect Analytical
Outcomes: An Illustration
This section will deconstruct and analyze critical rhetorical passages in
Richard Kohn's article, "The Forgotten Fundamentals of Civilian Control of the
Military in Democratic Government". I have chosen this article for three reasons:
first, the very title suggests that Kohn will attempt to make a claim that something
the military did, or is, or does, affects democracy in general. Second, Kohn is a

prolific writer in the discipline with fairly well known opinions in the community that
carry a reasonable degree of credibility. Third, Kohn's rhetorical approach is similar
to most articles written about U.S. civil-military relations from an "undue military
policy influence" perspective. I will proceed through Kohn's use of language
describing characteristics of the democracy that he sees as being threatened,
analyze the propriety of his linguistic choices and conclude by suggesting that
alternate democratic characteristics could have led to a different conclusion. It is my
hope that this exercise clearly demonstrates how vague assumptions of democracy
compound the potential for analytical error.
In his abstract, Kohn says "The challenge to new, as well as mature,
democracies is not only to prevent military interference in government by the
military, but to insure the supremacy of civilians in military affairs. Civilian control is
not a fact, but a process...and measured by the relative influence of the military and
civilian leadership in military policy and decisions" (Kohn, 2). What can we discern
about Kohn's democratic preferences from this statement?
The first striking observation is that it is not clear what Kohn means by the
term "interference". A democracy appears to be a political system that requires an
absence of this interference by military officials. It seems reasonable to conclude
that this interference is the influence referred to in the second sentence. There is no
other clarification of the meaning of "interference" in the remainder of the

document. Substituting influence for interference leaves us with this new description
of what democracy is: Democracy is a political system that requires an absence of
an unspecified influence from military officials. Given my earlier argument that
scholars who focus on influence issues tend to be Madisonian in democratic
orientation, one can find Kohn reinforcing this contention in a following paragraph
where he discusses the necessity of "countervailing power" that has traditionally
checked military policy influence (Kohn, 2). The second observation one can derive
from the abstract is that civilians must retain supremacy in military matters, and, by
extension, one could reasonably assume that they must retain supremacy in the
whole of democratic government. Thus, Kohn's democracy now insists that
unidentified civilians must determine what he refers to as "military policy". The final
observation I have is that once Kohn establishes the democratic imperative of civilian
supremacy, he asserts that the "fundamental" principle of civilian control
consequently derived from this necessary supremacy means it is a process measured
by military influence in policy decisions (Kohn, 2).
From this abstract, we are initially left with the basic democratic shell by
which Kohn will proceed with his argument. Democracy, then, is a political system
that requires an absence of military influence on the policy process. Civilians alone
must dominate this policy process and the control exercised by these unidentified
civilians must then be manifested in yet another mysterious process ("...civilian

control is not a fact, but a process..."). Kohn adds his last initial democratic
ingredient two sentences later when he asserts that civilian control "...requires
constant effort by those involved and support from the populace and the organs of
influence (his countervailing power?) in society" (Kohn, 2).
This last component is important for two reasons. The first reason is that he
introduces public oversight or pressure as a control mechanism, thereby
complicating his institutional democratic dynamic. Note that this public role is
supportive, not primary. The second reason is that this dynamicin spite of the
inclusion of the public elementexpresses a balancing of "natural tensions" similar
to the federalist assumptions behind the design of the constitutional separation of
powers, which seems to validate my argument that Kohn's unrefined democracy is
indeed not all democracies but the one under which he lives. We now know that
Kohn's democratic reference is what he sees as the American model of democracy
with a few arbitrary characteristics (civilian supremacy) tossed in for good measure.
The role of the people appears to be tertiary, with citizens relegated to third party
status whose relevance is to support the legitimate institutional policymaking
function dictated by the Constitution. Is Kohn's vision of democracy as universal as
he claims when he refers to "all new democracies'? Obviously, the answer is no.
Does his democratic shell display a bias for institutional separations of power?
Probably, though it is certainly a deduction on my part. Does Kohn have any

characteristics to add to his blossoming definition of what a democracy really is? You
On the following page of this article, it appears that Kohn is not quite done
defining his threatened polity. These new add-ons, moreover, are disturbingly
inconsistent with his first set of characteristics. He argues that 'The military is
authoritarian and purposely hierarchical, while democratic society is consensual or
participatory and emphasizes egalitarian principles" (Kohn, 3). This statement begins
a descent into a dysfunctional conception of democracy. To begin with, Kohn does
not seem to indicate how hierarchy diminishes political participation in a democracy.
How is hierarchy, in other words, mutually exclusive with participation or even
consensus? Second, the relationship between the hierarchical military and egalitarian
principles is not well clarified either. This statement could contain multiple meanings,
such as "hierarchy" implies that a higher caste gives orders to a lower caste, which
in turn obeys the orders without any deliberation. It could also mean that
egalitarianism literally excludes the possibility of rank differences, something very
much in dispute. If the meaning is the latter, it could certainly prove problematic for
the civilian hierarchy that he acknowledges is functionally legitimate due to the
primacy of the constitutional ordering of position and designation of the president as
"commander-in-chief. If it is the former meaning, that could certainly be grounds

for comparison, but highly unlikely, since civilian leadership is also hierarchical in
that the Secretary of Defense gets his marching orders from the President
In other words, there is every reason to believe that Kohn's inclusion of the
egalitarian element of democracy is an unanchored reference to abstract theory in
direct contradiction to his ordered hierarchy between institutions and the population.
Kohn then proceeds to note that democracy is "individualistic" in nature, something
the military is decidedly not. Apparently, Kohn now equates J.S. Mill's liberalism with
democracy so intimately that any classical democratic conceptionsuch as the
Athenian city-stateno longer qualifies as legitimate. Further, Kohn continues along
utilitarian lines when he says that "democracy...((attempts)) to achieve the greatest
good for the largest number..." (Kohn, 3).
This increasingly perplexing construction of a democratic vision gets even
more mangled when Kohn introduces his perception of what the military value set is
visa vis normative democratic values. "While many of the military's professional
valuescourage, honesty, sacrifice, integrity, loyalty, serviceare among the most
respected in human experience, the norms and processes intrinsic to these
institutions so diverge from the premises of democratic society that the relationship
is inherently adversarial and sometimes unstable" (Kohn 13). This lastand most
critical assertionis frightening in that it is difficult to determine whether or not
Kohn is demonstrating a less-than-ironclad understanding of democratic theory or an

even firmer grasp of contemporary myths about the degradation of American values.
In short, what Kohn is literally implying is that democratic values are cowardice,
lying, self-service, corruption, disloyalty and selfishness. This may at first seem
harsh, since I would agree that Kohn's disposition is highly favorable to the notion of
self-rule. However, these seemingly negative values are unquestionably the
opposites of the values he ascribes to the military ethic, which he does claim is
antithetical to civilian values. While this may be true about American society, it is not
an adequate description of a theoretical democracy.
Further, one would wonder why Kohn would leap at the chance to defend
such a value set over one that he must consider clearly superior to democratic
values. It surprises me to learn that Kohn is not out there stumping for a military
coup as soon as possible. That Kohn fears inordinate military influence is explicit in
his article (Kohn, 14). This seems odd, considering how repellant his democracy
must look at this point. Why the apparent contradiction and lack of clarity? Why is
Kohn's democracy egalitarian, participatory and consensual on the on hand, and
imbued with the values of Lucifer on the other?
My argument is that Kohn, like so many other scholars in this discipline, has
only a rudimentary grasp of democratic theory. It seems to me that what he has
done is confuse democratic normsdemocracy as a political systemwith what he
perceives as the social norms of his own society. This disparity between social and

democratic norms may even be accurate, but by confusing the two, he immediately
forfeits any credibility to his ultimate claim. One leaves Kohn's article not knowing
whether the author fears for his society or democracy as a system. Even ditching
Kohn's dreadful value comparison, his construction of democracy is so permeated by
eclectic and contradictory components that what we are left with is an entirely
unworkable and inappropriate democratic model to use as a comparative tool for
analysis. I believe that Kohn's confusion corroborates Wildavsk/s contention that the
American political tradition is a mixed bag of nuts, with democratic principles
consistently struggling against each other. As Australian democratic philosopher John
May points out,"... a clear understanding of what democracy is a major
condition for behaving democratically" (John May 1980,323). It could be reasonably
argued that a clear understanding of what democracy is about is a necessity for
knowing precisely what principle of democracy military attitudes, behaviors and
policies are threatening.
One wonders if Kohri would sound an alarm (as he does in his article) if
democratic values virtually matched military ones? Would he likewise perceive a
threat to civilian control if the American people passed a constitutional provision
requiring the military to formulate strategic military policy concerning the conditions
for applying force? Only Kohn could answer that question, but he would do so at
grave risk to his analysis. By permitting the military to formulate policy, he would be

violating one principle of his democratic framework: no interference or influence by
the military in the formulation of military policy. By dismissing the popular will and
relying on the Constitution as the governing document of all democratic theory, he
would violate another: the democratic principle of self-rule, popular support and
egalitarian participation.
Kohn is, then, in a lose-lose situation from the first moment that he fails to
clearly define what his target polity is, and subsequently claiming that his analysis
applies to "all new democracies". As May contends: "Four conceptions of democracy
vie for popular favor, most people veer between conceptions, often at the cost of
coherence" (May, 323). What makes analysis like Kohn's risky is that his audience
may all too readily jump with him to his conclusions without examining his
underlying assumptions about democracy. To act politically and legislatively on
Kohn's alert could prove impetuous at best, disastrous at worst.

Few contemporary scholars have trouble relating the hallowed words of
"civilian control of the military" to any specific civilian or civilian institution. In spite
of this apparent precision in identifying the subject of research, fewer still engage in
analysis of alternative constructions of what constitutes the most legitimate civilian
to be analyzed. This chapter will closely examine various interpretations of the
concept of civilian control of the military. Because few authors explicitly define what
they mean by "civilian", many of my assertions as to which scholars mean what are
gleaned from contextual hints tucked away in various texts. For example, when one
author loosely discusses "civilian control of the military", the context of which
civilians he or she is referring to can be found in the central argument of the text.
You may see this term in conjunction with an analysis of congressional oversight of
national security policy, which typically means that the author is referring to civilians
as those legally and formerly charged with oversight or military administration. At
first this may seem obvious, though it is important to point out that it is almost
always the reader responsible for making these connections.
Much literature has been generated regarding military politicking with each
branch of government. Most of this work, however, is not grounded in the context of
civilian control of the military perse, but more detailed analysis of national security

effectiveness. I wish to make clear at the outset of this chapter that such literature
does not concern itself with judgments over the health of U.S. civil-military relations
but, as a rule, the effectiveness of the policy process visa vis national security and
military policy outcomes.
For example, Sharon Weiner's article, "The Changing of the Guard: The Role
of Congress in Defense Organization and Reorganization in the Cold War" (1997),
examines the organizational changes instituted by the passage of the GNA and
subsequent changes in congressional oversight procedures. In her article, Weiner
lambastes the GNA as legislation that has eroded the power of congressional
oversight due to the increase in military policy power because military advice
becomes more unanimous (Weiner, 3-5). Regardless of whether or not one agrees
with this argument, my point here is to illustrate that Weiner's article makes a claim
that military power in policymaking increased as a result of the passage of the GNA.
However, she does not make any claim whatsoever as to whether such power is
dangerous to democracy in principle, or specifically hazardous to the American
institutional arrangement. She argues that Congress may have short-sheeted itself in
the deal, but does not conclude that there has been a fundamental breach in the
tradition of military subordination to civilian authority.
It is important to make this point quickly and as dearly as possible,
particularly for readers who may have considerable exposure to vast tracts of

scholarly literature devoted to domestic military affairs, particularly research
concentrating on analyzing military policy influence. It is easy to confuse an article
like Weiner's as concerning civil-military relations, when in fact, it is an article that is
focusing only on the policy process and its outcomes in nearly value-neutral terms.
In short, the arguments I make in this chapter about the lack of clarity of the term
civilian that is often seen in genuine civil-military relations work are not generally
affected by policy literature focusing on military or policy efficiency. I write this
somewhat convoluted passage in the hopes of deterring the reader from examining
my arguments with such literature in mind as a counter-argument.
There are essentially two types of military policy influence writings that deal
with civilian masters and their respective relationships with the military. The first
type, in which I indude Weiner's article, is literature that discusses military political
power strictly in relation to how policy decisions are made and their practical
outcomes, i.e.; did the GIMA increase military political power and thus decrease the
competence of national security deasion-making? Such research often does have
some implication for scholars of civil-military relations without question, an overlap
that we will see more vividly in the concluding case study in this chapter. However, it
rarely makes the claim that military political influence, or power, is somehow
damaging to the tradition of civilian control or to any aspect of democracy. It is
simply evaluating the state of the national security decision-making apparatus.

The second type of literature, often appearing very similar in form and even
content, is true civil-military relations literature. As a counterexample with Weiner's
article, I would offer Christopher Bourne's work to demonstrate this second type.
Bourne makes many of the same arguments that Weiner makes, including her
ultimate claim. He argues that the GNA does diminish the oversight ability of
congress and has a negative effect on the competence of policymaking in national
security circles. What distinguishes his work from Weiner's, and thus makes his work
relevant to civil-military relations analysis for this thesis, is that he further claims that
this increase in power violates both the tradition of civilian control of the military and
represents an incremental encroachment on democratic principles.
This chapter will not address literature of the first type, such as Weiner's
article, though I often use such literature as examples to prove a point, and will do
so in my concluding illustration. I am taking direct aim at literature such as Bourne's
that seeks to make an empirical claim that democracy is threatened, or that the
tradition of civilian control is being eroded or strengthened. The remainder of this
chapter will be dedicated to discussing the problems I associate with a lack of
linguistic clarity over the term civilian in "civilian control of the military". It is my
intention to demonstrate that as a rule, the more expansive and vague the meaning
of civilian is in any analysis, the greater the likelihood for concluding that there is a
civil-military relations problem in the United States.

The Mysterious Fourth and Fifth Branches of
Government: Qvilian and Military
It seems clear that the work of Samuel Huntington in 1956-57 has
subsequently dictated the conceptual foundations, language used, and the rules of
analysis in civil-military relations scholarship. He provides an adequate starting point
for an analysis of how the term civilian has become so problematic. In The Soldier
and the State. Huntington begins by creating what is in effect a fourth branch of
governmentciviliansopposed to a fifth branch, the military. In his introductory
discussion of methodological premises, Huntington states that "Any system of civil-
military relations thus involves a complex equilibrium between the authority,
influence, and ideology of the military, on the one hand, and the authority,
influence, and ideology of nonmilitary groups on the other" (Huntington 1957, viii).
There are several noteworthy aspects to this methodological assumption. The first is
that civil-military relations should be analyzed with an eye towards identifying some
equilibrium. Equilibrium, by definition, implies tensions of equal strength that result
in balance. Is Huntington suggesting that the military is inherently in conflict with
civilians of any conceivable stripe? Or is he suggesting something more specific?
Huntington apparently has particularthough unnamedcivilians in mind when he
asserts that "Civil-military relations is one aspect of national security policy. The aim

of national security policy is to enhance the safety of the nation's social, economic,
and political institutions" (p. 1).
Thus begins the contemporary quandary of civil-military relations scholars
who have a hard time matching up just what civilian(s) are relevant as the subjects
of civilian control of the military. First, Huntington does, as we saw in the last
chapter, what Albright accused him of doing: creating a monolithic group of
"civilians" opposed to "military" power. Second, Huntingtonwriting his book during
the social scientific infatuation with structuralism and institutional analysisis
apparently limiting the choice of civilian subjects to those laboring on behalf of his
treasured national institutions. It appears that very little of Huntington's
methodological and philosophical assumptions concerning the relationships between
military and nonmilitary institutions has changed all that much over the previous
forty-three years (Albright, 556).
In the type of literature that I classify as genuine civil-military relations
scholarship, this fixation both on institutional democratic dynamism of the American
variety and civilians as a comprehensive, comparative variable is as pervasive as it is
inappropriate. Timothy Boylan sums it up best:
"Just what our forefathers did envision ((regarding civilian control of the military)), or
would have envisioned had they foreseen modem conditions, must be divined from
materials almost as enigmatic as the dreams Joseph was called upon to interpret for
Pharoah. A century and a half of partisan debate and scholarly speculation yields no
net result but only supplies more or less apt quotations from respected sources on
each side of any question" (Boylan, 235).

Boylan, of course, ultimately commits himself to the path he has just eloquently
skewered, by focusing on civilian control as a matter of institutional jurisdiction and
relevance (233). At no time does he consider the possibility of a legitimate
democratic relationship between the people and the military directly. As we saw with
the Kohn illustration in the previous chapter, such an assumption is ill advised, since
the introduction of separation of powers arguments tends to encourage more
alarming claims than any analytical or empirical evidence justifies. As Kohn twisted
the meaning of democracy into an incoherent, contradictory potpourri of popular
mythology and social commentary, so do scholars when it comes to the meaning of
civilian in civilian control. If Boylan is correct in suggesting that scholars are merely
regurgitating historical "evidence" in order to frame a preexisting policy preference in
more favorable light, perhaps it could be argued that the more broadly utilized
civilian permits a greater analytical flexibility when studying the issue of domestic
civil-military relations.
There may be one other explanation for the prevalence of ill-defined civilians
in contemporary literature. Reading Bill Kaplan's article, in which he defines civilian
masters as "managers", one gets the sense that many researchers are attempting to
avoid the quagmire of analyzing the theoretical problems of delegation mentioned in
the last chapter (Kaplan, 78). It becomes easier with a definition as broad, and
virtually useless, as "managers", to focus less on the precise degradation or

enhancement of the civil-military relationship and more on questions of institutional
power increases or decreases. Since Huntington apparently framed the context of
civil-military relations scholarship as an issue of national security, this could explain
why so many scholars are content to focus on what I believe is an easier empirical
problem: the study of military policy influence relative to national security policy
outcomes. It is, in other words, much simpler (relatively) to make a claim that the
GNA either increased or decreased policymaking competence then it is to relate the
act itself to more theoretical claims that some tenet or fundamental principle of
democracy has been attacked, however obliquely. The latter question seems less
relevant to the practical realm of international relations professionals, an observation
made by Daniel Nelson (137-39).
The domination of national security studies by groups of scholars generally
devoted to comparisons between democratic and non-democratic systems probably
also contributes to this phenomenon, since relational distinctions between types of
legitimate civilian masters operating in democratic and non-democratic states seems
much simpler and far more relevant to the scholarly goals of national securities
studies (Burk 1999,67-8). The most reasonable conclusion that I can draw for the
tendency of so many scholars to generalize the civilian object in their research,
almost always to the point of uselessness, is that two problem areas are avoided if
such an approach is taken. The first problem area is that of delegation, though such

debate flourishes in a different context, something I will address in depth shortly.
The second reason is that the natural tensions and power fluctuations of American
democracy, with its competing institutions, makes proclamations about legitimacy far
more difficult than in non-American democracies (Feaver, 16).
It is, then, the existence of the constitutional structure that perhaps creates
the incentive for scholars to poorly define the variable of civilian. Given the fact that
multiple civilian agencies and actors vie for power, and thus for legitimate control
over military organizations and policy, it becomes difficult for any scholar to know
precisely how a legitimate hierarchy of civilian authority should be constructed. Does
the President trump Congress when it comes to controlling the military? Does the
Supreme Court trump everyone? Without knowing the answers to these kinds of
theoretical questions with any certainty, most scholars would spend most of their
time attempting to sort out often confusing relationships between governmental
institutions in order to determine where a military act of inappropriate behavior has
caused its damage or where such behavior had no effect. This reflects the original
argument by Huntington when he asserted that much of the problem in American
civil-military relations analysis was a result of the often confusing intent behind the
various military and militia clauses of the Constitution (Huntington 1956,677).
One thing is certain: there appears to be a normative dilemma for many
scholars of civil-military relations. The very subject matter, particularly for

comparative civil-military analysts, requires that there exist an ability to discern what
military political behavior is or is not appropriate in any democratic regime. How else
to tell the South American generals that they are being naughty when they insist
that they are taking power not to prohibit democratic government, but to ensure
that conditions are stable enough so that someday, democracy might flourish in their
country?1 On the other hand, by emphasizing the national security aspect of civil-
military relations, it is tempting to focus on assessments of military policy influence
and power only inasmuch as such analysis enables a scholar to determine whether
or not the United States is successfully pursing a particular policy course.
The greatest complication from relating military behaviors to antidemocratic
results is that there remains the possibility of having to conclude that some obscure
General who has just taken over the reigns of power in a Third World country,
ostensibly as a result of public urging, may not be such a theoretical violation of
democratic principles after all. For national security aficionados obsessed with
replicating the American experience worldwide, as Lawrence Whitehead noted in the
previous chapter, such a confession would be unthinkable unless, of course, the
general happened to be a virulent anti-Communist.

Checks and Balances. Aaencv Theory and the Problem
of Delegation: Part II
Having posited the argument that many scholars embrace vague and overly
broad conceptions of civilian objects in their research in order to minimize the
theoretical complexity that greater detail would incur, I would like to now examine
the antithesis of this argument. This section will examine the consequences of
combining assumptions of what democracy is with unjustifiably narrow definitions of
what is meant by the term civilian in "civilian control of the military". My argument
here will be that the narrowing of the meaning of civilian to political actors in one
governmental branch or another causes the same problems outlined in the previous
section. Further, I believe that such a delimiting practice is a result of both the initial
assumptions of what democracy is or isn't, as well as an avoidance of the same
theoretical complexity of sorting out democratic components with which to claim a
In "Delegation, Monitoring, and Civilian Control of the Military: Agency
Theory and American Civil-Military Relations", Peter Feaver focuses on the creation
of a new model of measurement designed to more correctly assess the health of
civil-military relations. Much of his model will be examined in considerable detail in
the following chapter, but for now, I would like to briefly point out the reasoning

trajectory that determines how Feaver defines his civilians. By asking himself the
question "How do civilians control the military?" Feaver is almost obligated to look at
institutional mechanisms of oversight, which is precisely what his model is oriented
towards (p. 1). What he does not do, like most scholars, is precisely define what
civilians are responsible for the control functions of the military.
What Feaver means by civilians is first, that they are "masters" of the military
(p. 2). Second, like Kohn and others, he both expands and narrows his definition
nearly simultaneously, beginning the first descent into the quicksand of imprecision.
Feaver suggests that civilians must be "leaders", institutional officials, and then,
ultimately, embodied in a three-way dynamic that demonstrates Feaver's civil-
military orientation towards national security issues. His civilians become two
principals in relation to the military: Congress and members of the executive
branch's national security apparatus (Feaver, 3-4: 7). Like Kohn, Feaver does appear
to inadvertently introduce the idea of public participation in national security and
civil-military discourse when he identifies increasing "public vitriol directed at the
Commander-in-Chief..." as a manifestation that there has been a crisis in American
civil-military relations (p. 2). What he means by "public" is not entirely dear in the
context of the article.
In Feaver's defense, however, he is the only agency theorist who has
recently been willing to examine complex relationships between the military and

multiple principals, unlike Weiner and others. As Feaver himself puts it, there are
"multiple sources of friction between the military and various institutional agencies
charged with oversight and management" (p. 4). In turn, the various responsibilities
that each agency has in controlling the military^-defined by Feaver as ensuring that
the polity is not subvertedmust be analyzed independently of qualitatively different
responsibilities of other principals. For example, Feaver might argue that you cannot
analyze civilian control by lumping together congressional oversight and executive
managerial responsibilities and treating them identically. In spite of Feaver's laudable
sophistication concerning the variety of control permutations available for analysis,
he winds up where many other scholars do: ignoring the third prominent element in
the American democratic dynamic, popular will.
What precisely is the relationship between scholarly preferences of American-
style democracy and assumptions of the legitimate civilians utilized for study? Much
of the answer to this question came in the previous chapter, where we learned that
focusing on constitutional primacy and institutional dominance in the American polity
causes most people to focus on only one aspect of democratic legitimacy: the
elected government. Focusing on delegatory legitimacy probably reinforces this
tendency, since the average citizen is not involved in issuing orders to military
officers. The people, in short, do not delegate authority to the military (strictly
speaking). This paints a portrait of the democratic dynamic that looks something like

this: citizens elect their officials, who in turn manage the affairs of state on their
This attitude exemplifies classic representative theory. Once the election is
over, then, there is no formal political role for the people, at least in terms of
constitutional provisioning. As we have noted earlier, this is only one of several
interpretations of the American political tradition. Further, it is an excruciatingly
formal interpretation of the American political process, one that encourages a nearly
complete lack of civil-military relations research between public sentiment and
military political behavior. Ironically, Feaver notes that earlier interpretations of the
nature of institutions and their subsequent creation emanates from popular desire in
a democracy (p. 3). Why, then, does Feaver ignore the possibility that there is a
third principal worthy of examination and utilization in agency theory? Perhaps, like
Kohn, he is not ignoring it completely. More likely, the fact of public participation in
policy formation via polling and organized legislative pressure are understood to
exist, but not easily fitted into an analysis of the finer points of legalistic delegation
of powers. In other words, without understanding the formal legitimacy that the
people have in participating in policymaking, how do you incorporate them as a
serious principal?
I believe that what we see happening time and again is a demonstration of
an intuitive understanding that the popular will matters, but not a conscious

perspective of how to analyze institutions systematically while incorporating a third
principal not easily lent to such a constricted empirical model. It is much easier, in
short, to see the product of congressional legislation, understanding what happened
and why in a branch of 535 elected representatives, than it is to determine with any
accuracy what role a population of 260 million may have played in the passage of
such legislation. Further, there are so many vague references to self-rule, popular
will and egalitarian political participation by many of these authors as to suggest that
there is an implicit understanding that the people have a legitimate role to play in
the control process, but few seem to have any idea of how popular control of the
military should formally manifest itself. As I will argue in the following section
discussing the public role in permitting the U.S. military to operate domestically for
the purposes of drug interdiction, it seems clear that on several occasions, the real
controlling civilian authority over the military has been, in fact, the American people.
For the moment, let us be content with understanding two things clearly:
first, that broad definitions of civilian tend to be used to permit more practical
national security analysis without having to engage in disputes over democratic
legitimacy outside the constitutional framework. This is, in essence, a policy affinity
on the part of these scholars. The second important point is that scholars who too
narrowly define civilian for the purposes of analyzing the distribution of political

power between institutions tend to exemplify a legalistic affinity in order to stabilize
the variables they use for analysis.
Citizens and Civilian Control of the Military
In Charles Dunlap's fictional military coup of the future, he argues that the
primary reason for the American military to take over the government was a result
of popular sanction through a fictional referendum (Dunlap, 2-3). Considering the
popularity of this essay as a reference for those who study potential negative
impacts of military attitudes on democracy, an amazingly scant amount of attention
has been paid to Colonel Dunlap's legitimizing device for the coup: popular electoral
ratification of a military regime. Interestingly enough, while Dunlap takes vicious
shots at civilian leadership for permitting the conditions that prompt the coup to take
place, one could infer from his essay that his civilian control objects are elected
representatives. It never occurs to Dunlap, much as it rarely occurs to anyone else,
that the popular referendum was arguably legitimate. In this essay, the civilian
masters of the militaryCongress and the executive branchare the protectors of
democratic virtue while the people are its enemies. This value judgment is certainly
implicit in Kohn's work as well. But what about those who simply cannot account for
the reality of popular pressure as a mechanism for influencing the policymaking
process? Our question is to determine if including the people as a reasonable

component of civilian control of the military is warranted. I think the answer to this
question is clearly "yes"
Perhaps no other use of the United States military mobilizes public opinion
more effectively than proposals to employ soldiers as domestic police or law
enforcement auxiliaries. This section will examine recent and historical uses of the
military on domestic soil, the literature written about such events and the
relationship between domestic constabulary use of the military and civil-military
relations scholarship. Additionally, I will argue that it is in this area where our third
civilianthe popular willcan be seen as a powerful and legitimate controlling force
over military activity and attitudes. Further, I will argue that relegating public opinion
and pressure as a "side-show" democratic principal in analyzing contemporary civil-
military relations is completely unjustified and contributes to the confusion
surrounding platitudinous uses of the word civilian.
Much is made of contemporary use of the American military in domestic law
enforcement operations during the last two decades. It is important to note,
however, that this nation has a long history of deploying federal, professional
troopsnot just state militiasto regions of an expanding frontier where civilian
policing was minimal or entirely absent (Kenner 1995,46-7). From obvious events
such as the federal occupation of the defeated South during Reconstruction to
obscure policing duties in the frontier West prior and subsequent to the Civil War,

the military has often been called to duty in a law enforcement or constabulary role
in the United States. Kenner notes that one of the more obscure roles of military
policing was immediately after Texas became a state. Cavalry troops were used as
escorts and protectors for the critical cattle trade of the region. More importantly, for
our purposes, this role was foisted onto an already thinly stretched military presence
in the southwest by a combination of public opinion and political persuasion on the
part of the region's politicians of that time (Kenner, 48-50). In Big Springs, Nebraska
during 1877, federal components of the Army were utilized to assist in investigating
and apprehending the perpetrators of a major train robbery (Ball 1995, 34). In this
latter example, popular opinion was not responsible for the decision to utilize the
militarythis was initiated by the meager law enforcement resources in the area
but was instrumental in ensuring that some transgressions committed by military
investigators were halted and procedures changed to accommodate what were
perceived as democratic principles (Ball, 40).
Today is, in some fundamental ways, no different from the post-Civil War
frontier. The military is currently utilized in some function or another for the
purposes of drug interdiction, illegal immigration, riots and potential cases of
terrorism. Emergency powers clauses and legislation do give a president some
authority to employ federal troops domestically, but what is important for us is to
understand the role that popular opinion and pressure plays in such decisions. In

other words, are such uses of the militaryalong with the corresponding increase in
military power and influencesolely promulgated by elected and appointed officials?
According to a later article by Charles Dunlap, the answer to this question is that the
people have indeed played a tremendous role in instigating the domestic
employment of federal troops for purposes other than war-fighting (Dunlap 1994,
According to Illinois State political science professor George Kiser, Ronald
Reagan's decision to utilize the military as an adjunct force in drug and illegal
immigrant interdiction efforts functioned both as an ideological and popular response
to what was, and often still is, perceived as a national crisis (Kiser 1997,32).
Matthew Hammond agrees, arguing that it is a combination of public desire and
political ideological opportunity that tends to positively influence decisions to deploy
military force domestically (Hammond 1997, 953-54). This public pressure in favor of
domestic military employment tends to rub scholars of civil-military relations the
wrong way. Both Hammond and Kiser lament public sentiment because for them, it
is contradictory to more legitimate democratic principles embodied in the Posse
Comitatus Act of 1878, which prohibits the domestic employment of military forces
for purposes of law enforcement.
It seems once again that the public is an enemy of democratic values whose
opinion cannot be considered as democratically legitimate unless that opinion

reflects a more "correct" understanding of democratic values (Hammond, 963).
While Kiser repudiates Congress for passing legislation enabling Reagan and Bush to
deploy troops in law enforcement (or pseudo-enforcement) capacities, he does
identify them as the primary civilian controllers of the military and national security
policy (Kiser, 34). Just as public opinion in Big Springs, Nebraska was a key factor in
adjusting military investigative procedures of the train robbery, so too did public
opinion begin to chip away at the power that both the public and the government
gave the military to interdict drugs and illegal immigration in Texas. Public outcry
over the accidental shooting death of Esequiel Hernandez, a Marine
corporalprompted a suspension of interdiction operations in that area and a
subsequent reevaluation of the policy of having U.S. Marines augment law
enforcement agencies along the southwest border (Gwynne 1997,40). Patrice
McSheiry inadvertently complicates this relationship by suggesting that much of
what is considered undesirable policymaking regarding domestic military
employment is a result of decisions made by civilians in the national security
apparatus, an argument that I doubt few of the scholars mentioned in this thesis
would disagree with (McSherry, 1). However, this introduces the possibility that
perhaps the people are not only a legitimate check on government on election day,
but a persistent and real political actor en masse whose attitudes and behaviors are
equally relevant in controlling military power and influence.

Lest we come to the conclusion that only the American public has such an
effect on policy preferences expressed by government agencies, can we find any
confirmation of this hypothesis in literature about other democratic states? Hugh
Smith points out that Australia, as well as other western democracies, is frequently
subject to public pressure to increase or decrease the relative military policy
influence and power in national security matters. In Smith's analysis, the democratic
dynamic is not purely institutional, but a tangible dialogueperhaps even dialectic
between citizens and their elected government (Smitii 1998,219). Additionally,
Smith introduces the element of democratic consensus-building into his analysis by
describing the way the public and government officials interact and make policy
adjustments accordingly (Smith, 225). If this constant, synthetic process is at the
very least as accurate as most of our models of inter-agency relationships, then it
would be sheer hubris for any scholar to ignore the impact on military power and
influence by our third civilian, the people.
What is Meant bv Civilian Control?
Paralleling the illustration in the last chapter, this section will systematically
deconstruct the use of linguistic references to civilian masters of the military in
"When Should the Military Speak Up? A Qvil-Military Dilemma", an article published
by Sam Sarkesian (1998). The central argument suggests that the military should,

contrary to conventional wisdom, become more politically engaged in the policy
formulation process, not less. Sarkesian explicitly analyzes civil-military relations
principles of military non-politicization (Huntington's objective control) in order to
determine whether or not the national security decision-making process can be
improved if the military becomes more politically vocal. I have chosen this article for
three reasons. First, Sarkesian is a more conservative, pro-military scholar than Kohn
is. Sarkesian's viewpoint in this article represents an opposing analytical faction
rarely driven to make alarmist claims. If anything, Sarkesian represents a more
apologetic viewpoint towards military political participation. His ultimate claim in this
argument is that if there is a civil-military relations crisis, it is because the military is
too uninvolved in politics. Another reason that I have chosen this article, like Kohn's,
is that I believe it very representative of mainstream confusion over civilian
legitimacy in terms of defining which civilians should control the military and how.
My final reason for selecting this work is that it represents the odd synthesis of
institutional power analyses (like Weiner's) and genuine civil-military relations
research (like Bourne's and Kohn's).
Sarkesian at no point clearly defines what he means by civilian control of the
military, though there are several obvious hints in the text. In arguing for the
legitimacy of what he calls "constructive political engagement", Sarkesian's first
tangible statement concerning who his civilian masters are goes like this: "Nor are

constructive political engagement and loyalty to the country, civilian leadership, and
the Constitution in any way incongruous" (p. 11). Without actually saying so, one
could assume from this sentence that the civilians ostensibly responsible for
controlling the military are the "civilian leaders". Sarkesian's distinction between
loyalty to the country, the leadership and the Constitution in effect sets up the
classical dilemma of civil-military civilian legitimacy. What we know for the moment
is that there is the country, there are civilian leaders, and there is the Constitution.
He continues: "It is because the U.S. military is under such tight civilian control that
it needs to make its voice heard in civilian councils" (p. 11-12). This sentence seems
to further clarify what Sarkesian means by the term "civilian". Apparently, officials
affiliated with the civilian councils exercise this tight civilian control. This seems the
most internally consistent meaning given his previous identification of civilians as
leaders. Referring to H. R. McMasters' book Dereliction of Duty2. Sarkesian further
delimits his civilian masters to national security officials charged with development of
security and military policy, saying that "...the military brass in the 1960s was not
following an American tradition when they kept their silence and followed the
civilians lead" (p. 12). This appears to confirm that the object civilians are not only
"leaders", but leaders affiliated with the policymaking apparatus in the executive

It is at this point that Sarkesian takes a radical turn in discussing the
military's direct relationship with the population, not the government. Considering
that Sarkesian is attempting to argue that increased military politicization is a good
thing, his introduction of the populist democratic tradition as a counterweight to the
traditional institutional focus comes as no surprise. He is, in effect, redefining what
he means by "civilian control of the military" away from Huntingtonian terms in order
to enhance the credibility of his case. He first quotes General Ridgway: "Since
George Washington's time, no top soldier has forgotten that he is a citizen first and
a soldier second, and that the troops under his command are an instrument of the
people's will" (p. 12). This is the first concrete appearance of our third civilian in the
democratic dynamic. Sarkesian then completes the legitimacy split with the
executive branch in the following passage, quoting General Fred Weyand:
"The American Army is really a people's Army in the sense that it belongs to the
American people who take a jealous proprietary interest in its involvement...The
American Army is not so much an arm of the Executive Branch as it is an ami of the
American people" (p. 12)
What we are learning about Sarkesian's assumptions at this point is both
subtle and critical. The cumulative impact of all of the previous references to
civilianseither in the context of military controllers or as non-military members of
the publicis that civilian control of the military is unquestionably a responsibility of
civilian leadership. Sarkesian, in other words, was perfectly aware of his distinction
between "the country, civilian leaders, and the Constitution". He has taken the

liberty of defining the term civilian in civilian control of the military to mean elected
or appointed officials. Additionally, he has taken the time to introduce our more
obscure third democratic principle, the people, but not in the context as civilian
masters of the military, but as an object of loyalty for the military. What I mean
when I say this is that for Sarkesian, the people do not literally represent controlling
actors of military policy. As such, the principle of civilian control of the military does
not specifically relate to them as principals in the analysis of civil-military relations.
What they do represent, however, is a component of the polity which the military is
pledged to serve.
It should be noted that just because Sarkesian is developing an argument
that includes the people as a legitimate political force does not mean that he is
considering them as principals in civil-military relations. In fact, the opposite is most
likely true, as we can see from this statement:
It seems dear that the American military belongs to the American people, and
military professionals have the duty and obligation to insure that the people and its
political leaders are counseled and alerted to the needs and necessities of military
life. This cannot be done by adhering to a notion of the military profession as a silent
order of monks isolated from the political realm" (p. 12).
The military, then, does not take orders from the people. It is not even controlled by
the people, at least directly. The military, according to Sarkesian's rationale,
addresses the people in a guardian-like role in order to prevent any subversion of
the people's will by the civilian leaders who he authenticates as the legitimate

controllers of the military. This is a somewhat confusing approach, unless one takes
some extra time to infuse a bit of healthy cynicism into Sarkesian's real pitch.
I believe that Sarkesian's purpose in introducing the people as a legitimate
democratic element worthy of incorporation into his analysis is to permit him to
argue for a change in the conventional thinking about civil-military relations. Since
we know that the Huntingtonian principle of objective control is, for the most part,
still dominant, Sarkesian cannot make a credible argument for permitting a politically
vocal military leadership without separating military duty from purely institutional
and constitutional obligations. Further, given the already known fact that Sarkesian
is sympathetic with McMaster's criticism of civilian competence in military matters^
such a move helps him legitimate greater.military political participation, which is the
stated goal of this article. Another way of looking at my argument is this: Sarkesian
is unhappy with what he sees as an increasing civilian incompetence in military
matters at the national level and thinks that perhaps the best chance to correct
these deficiencies is by finding a way to torpedo the idea of objective control and
giving the military greater legitimacy to conduct public discourse and attempt to
correct civilian policy mistakes. Let us see if there is any further evidence in
Sarkesian's article to confirm my theory.
It appears that Sarkesian, like General Powell, was opposed to the idea of
intervention in Bosnia.

"Military views about the involvement in Bosnia is a case in point At the national
level, the Joint Chiefs of Staff can and should clarify military concerns about the U.S.
peacekeeping mission there and about so-called "non-traditional" missions in
general. At some point, the wise, professional, and patriotic course may be for high-
ranking officers to insist that such missions simply cannot be done with the present
resources and structure" (pp. 12-13).
With the prevailing notion that the military is supposed to be immune to policy
preferences in international relations, or at least obligated not to air them publicly
without the explicit approval of the "civilian leadership", such negative sentiments
about a policy course would otherwise have to be stifled as "just another day in a
democracy". Sarkesian, in my view, has introduced the legitimacy of the American
people not as an additional comparative, analytical variable, but as a way of
circumventing objective control of the military. This brings us back to the original
question of this section, and, indeed, of this chapter. What do we mean by "civilian
control of the military", and what is the impact of the lack of specificity of this
meaning in Sarkesian's tract?
It seems clear that for many scholars, including Sarkesian, the tradition of
civilian control of the military remains a legalistic, institutional concept embedded in
the fabric of American governmental structure. Most civilians are generally
assumedat least in the context of controlling the militaryto be elected and
appointed national government officials of some variety or another. Though scholars
such as Sarkesian come closest to introducing the third civilian as a legitimate
mechanism of control conforming to the tradition of civilian control of the military,

they rarely take the additional step of explicitly making the connection and
discussing the ramifications of such a direct relationship. By using the term civilian
loosely, leaving the reader to interpret what he means, Sarkesian allows himself to
rush headlong into an analysis of civil-military relations that permits him to push a
specific policy agenda: organized opposition to peacekeeping missions, preferably
spearheaded by military advice. By subsequently separating, again implicitly, the
American people from legitimate mechanisms of control, Sarkesian's implied
meaning of the term "civilian" conspires to ensure that democratic assumptions
continue to be skewed in favor of uniquely Madisonian interpretations. What I mean
is that once the American people are no longer considered to have as valid a claim
to the role of civilian masters as elected and appointed officials have, they are
relegated to supportingthough still importantplayers in a drama played out
between governmental agencies and their respective representatives. As discussed
in detail in the previous chapter, such a view of democracy is far too fragile to use as
a reliable comparative variable with which to analyze the health of civil-military
A second impact of the imprecise and ill-defined civilian, particularly in
Sarkesian's article, is that if he chose to approach his argument directly linking
civilian control of the military to the people, he would have had a much easier time
justifying military political participation on the grounds that the people, as equal

masters of their own military, have a right to know military opinions in order to
participate as political actors. Instead, Sarkesian attempts to make what looks
instead like a cynical end-around a contemporary political gripe he has with the
current chief executive. If Sarkesian analyzed the democratic dynamic as Dahl had
done, and concluded that the people have a right to be involved in the policymaking
process more directly, Sarkesian's argument would be a cakewalk. I believe it is very
reasonable to argue that most civil-military relations analysis that utilizes vague
definitions of what the term civilian means suffers the same fate. The lack of precise
identification of the relationship between the military and its legitimate masters, and
why such identification is justified, frequently ruins the credibility of many claims of
harm or beneficence. If Sarkesian had simply been clear about who his civilian
principals were, and then justified why the people should be included among them
for purposes of defining the principle of civilian control of the military, his argument
could have looked very different. He could have argued that the people, as the
source of civilian control of the military, are being misled by government officials.
Since the military has an obligation to serve its rightful masters according to the
principle of civilian control of the military, it is the duty of the military to speak out
publicly in order to ensure that their legitimate mastersthe peoplehave all the
information they need. Instead, Sarkesian winds up arguing that he needs the
intervention of the American people at this time because the legitimate civilian

controllers of the militaryelected and appointed officialsare making policy
decisions that he does not necessarily agree with. Whereas the former argument at
least appears theoretically consistent, the latter appears to be a manifestation of
resentment There is a world of qualitative difference between the two, in spite of
the similarity of the appeals. In my view this chasm between possible analytical
outcomes is caused primarily by the lack of specificity of the meaning of civilian
control of the military and alternative permutations of democracy.
1 An overwhelming number of scholars acknowledge that most military coups occur with
pronouncements of how such action is "temporary" and in the target nations best long term
interest My rationale for making this point is to demonstrate that criticizing such a move
requires that some normative democratic principle has been violated, particularly when
military regimes most often superficially claim a long-term commitment to democracy. In
order to criticize a coup under these conditions, democratically affiliated officials and
advocates must be able to undercut the argument that democracy requires social stability
prior to its successful adoption. For an excellent review of this phenomenon, read Nelson's
"Civil Armies, Civil Societies, and NATO's Enlargement" and Geoffrey Demarest's (1995) 'The

Overlap of Military and Police in Latin America", Foreign Military Studies Office Internal Paper,
Fort Leavenworth, Kansas, United States Army (April): 1-13.
2 Dereliction of Duty details the conflict between the Joint Chiefs of Staff and McNamara's
"whiz kids" during the Vietnam War. Sarkesian is specifically referring to civil-military
relations tensions between policymakers in the executive branch and the military.

We have seen in the literature that the prevailing model type in civil-military
relations is generally scalar. By this I mean that most models try to measureusing
some set of variableswhether or not there is more or less civilian control of the
military. The variables used for measurement most often depend on what kinds of
events move towards the precipitation of either a military coup or policy dominance
in national security decision-making by the military (Feaver, 7). Variables most
frequently used in civil-military relations analysis are measurements of policy
influence as determined by policy outcomes, social attitudes and divergence
between military and so-called civilian value sets or, as Daniel Nelson suggests, any
particular political goal that an analyst wishes to plug into the referent poles at the
ends of a linear scale.
"None of the prevailing models of the military in politics, or of paths toward civilian
control of the defense establishment, can provide more than an imprease goal. Little
or nothing in the theories of academic literature provides a guidebook for the
remaking of military-civilian relations. Rather like trying to create a market economy,
one finds that trial and error, coupled with a sense of what ought to be (italics mine),
guide daily improvisation" (Nelson, 139).
Much of this thesis can be interpreted as a more detailed dissection of where
various normative preferences exist in the civil-military relations discipline. This
chapter will focus on more formal modeling approaches, arguing that linearity in
measurement of control factors is inherently unstable since such normative

preferencesas we have seen in depthare not very well clarified. It is the lack of
conceptual precision seen in our previous examples that I believe has contributed to
the possibly erroneously constructed linear models currently in vogue. I will attempt
to describe and discuss six distinct model types in an effort not only to cast doubt
upon the utility of such methods in empirical civil-military relations research, but also
to demonstrate why the previous two types of conceptual unclarity increase the
improbability that such models will ever be of any assistance in determining a
justifiable state of civil-military relations health.
Six Prevailing Analytical Models in Civil-Military Relations
Let me begin this section by saying that many of the characteristics of these
models overlap other models. There is also something approaching a hierarchy in
these methods of research, meaning that some models are directly inspired by the
same methodological assumptions of previous analytical frameworks. It is important
to note, however, that the similarities between these models do not necessarily
render them non-unique. Further, since my purpose is in demonstrating simply what
other scholars are using to make empirical claims, grand distinctions between each
one are not necessary, since even subtle permutations of the model have
considerably different effects on research. My final preliminary observation is that all
of these models, in spite of claims to the contrary, are unequivocally linear in nature.
Regardless of the referent points they utilize as measurement standards, all

ultimately are constructed to answer the following question: "Do we have more or
less civilian control of the military?"
Model 1: The Equilibrium Principle
It was Samuel Huntington who popularized the acceptance of a general
equilibrium, or balance, model in analyzing civil-military relations (Huntington 1957,
viii). This first model is less an analytical method then a guiding principle to which
most of the other five models adhere. I include it separately for two reasons: first,
there was little subsequent clarification of Huntington's concept of civil-military
relations balance for nearly 20 years (Albright, 557), and second, there are several
analyses that attempt to determine the existence of an equilibrium between
"civilians" and the military without any substantive, systematic method of empirical
Huntington assumed that there would be a natural antagonism between
military and civilian officials throughout the institutions of American government As
we discussed in the previous chapter, this arbitrary division of government into
competing groups labeled "civilian and "military" permeates contemporary civil-
military relations scholarship. Albright utilizes a comparative linear model himself. He
analyzes the health of U.S., and other democratic types of, civil-military relations by
comparing democratic systems with communist ones (Albright, 553-4). Bourne's
article is an example of how there are still occasions where analysis is conducted

using a murky, heuristic concept of equilibrium without any systematic modeling. He
defines civilian control and then uses certain historical and logical arguments to
suggest that there is movement away from an unspecified balance in civil-military
relations (Bourne, 16).
As a general rule, equilibrium models can be seen as internationally
comparative, such as the Albright and Sutter models, or they can be variable
comparative models. The distinction between the two types is that the first analyzes
differences between political systems, and the latter investigates discrepancies
between norms and practices in the same political system. It is this comparative
quality of the models that is used to determine movement toward or away from
There are three dominant characteristics of equilibrium models. The first
characteristic is that they are based on an assumed antagonism between our
mythical civilian and military branches of government. If it were true that such a
"natural tension" was na'ther universal nor constant, then an equilibrium model
would have to be rejected as inappropriate, since there is no balance to strike
between like-minded institutions. This rationale closely mirrors Huntington's idea of
subjective control of the military, where there is no competition between military and
civilian values and hence, no need to control the military. The converse, of course,
could also be true. I would suggest that the distinct possibility exists that adopting
an equilibrium model at the outset of one's analysis could very well eliminate any

chances of uncovering whether or not the assumption of "natural antagonism" is
justifiable. The second characteristic of these kinds of models is that they assume
we can reliably identify with any degree of certainty what a balance between the
military and civilians looks like. The third, and perhaps the most critical,
characteristic, is that such models discourage movement towards the positive end of
the scale. In other words, while most scholars are busy trying to find the equilibrium
point, movement even further towards the positive referent in the scale might be
subtly discouraged. It is fairly rare to uncover scholarly work inside the civil-military
relations community that emphasizes attainment of a perfect state of civilian control
of the military.
Model 2: Input versus Output Models
This is an equilibrium model that attempts to ascertain the policy preferences
of both the military and respective civilian agencies and institutions. Gibson and
Snyder (1998) note that such a model attempts " evaluate the changing power
of the military by analyzing the character of decisions taken at higher civil-military
levels. This often has been done by analyzing preferences and matching them with
outcomes" (p. 194). If a policy outcome contains the preferences of the military, it
must be a case where too much military influence was exercised. This would seem
to suggest that perfect civilian control of the military would mean that any piece of
legislation had better not contain anything considered valuable or preferable to the

military. Since few scholars would concede such a point, such models typically try to
distinguish between preferences that are symbiotic between the two groups, or to at
least separate outcomes as primarily political or military. In short, if the outcome is
seen as political, but preferred by the military, then it is a violation of civilian control
of the military. The procurement of a new weapons system that the military was
lobbying for, for example, would not in and of itself be seen as threatening.
If one were to broaden this explanation, it could be said that input versus
output models would typically examine patterns of outcomes over time in order to
make a determination of whether or not the civil-military equilibrium is in jeopardy.
Such a methodological approach is argued for by both Richard Kohn and Gregory
Foster (Kohn, 2; Foster, 42). However, there are also examples of where even this
longer-term analytical approach is ignored in favor of what I call a "snapshot"
assessment. Such is the case of Previdi's work in which he criticizes the GNA simply
because military advice was consolidated (p. 4). Previdi, in other words, does not
wait to actually see if military influence increases in real terms, he just assumes that
the equilibrium has been tinkered with by the passage of the act. This differs slightly
from Bourne's approach, which also criticizes the GNA for many of the same reasons
as Previdi. Bourne, however, bases his conclusion that the equilibrium is affected
negatively because there is an empirical likelihood of inappropriate influence being
increased (Bourne, 28). This is, to be sure, a subtle distinction. In either case, both
authors' approaches fall into this input versus output method of measurement

One of the obvious difficulties with this model is the lack of precision with
which preferences are measured and legitimized as an accurate reflection of the
institutions as a whole. If two service chiefs disagree on a political mattersuch as
what conditions are appropriate for military deploymentwhich opinion gets the nod
as reflective of "the military'? Certainly one could not expect Congress to have more
or less uniform political agendas. In spite of these shortcomings, this model is still
used widely in some variation or another, particularly by agency theorists, whom we
will address specifically in following sections.
Model 3: Attitudinal and Social Cleavage Models
Attitudinal and cleavage models claim to measure, in some form or other,
social attitudes amongst civilian leaders (often categorized as a reflection of popular
values) and values clung to by professional military officers. These are the most
difficult models to identify as linear, although they are often the most confusing
when it comes to what democracy means as well as how coups originate. These
models rarely focus on policy influence, but on remote possibilities of a coup or the
chances of orders getting disobeyed by socially isolated, recalcitrant, military
Thomas Ricks (1996) provides an example of this method in his article "On
American Soil: The Widening Gap between the U.S. military and U.S. Society". This
particular analysis relies on survey research conducted among military professionals.

Ricks claims that "...a striking indication of alienation from the larger society, an
overwhelming portion of the Basic School lieutenantssome 81%said that the
military's values are closer to the values of the founding fathers than are the values
of society" (p. 7). Ricks is attempting to demonstrate that the self-identification of
military personnel with the value system of the founding fathers is cause for alarm.
Ricks believes that the increasing "guardian" attitude of junior military officers can
lead to a direct increase in the possibility of military intervention in government. In a
nutshell, these kinds of models are attempting to demonstrate that the military is
both inordinately politically active, and that their attitudes are increasingly divergent
from the public they are sworn to protect. The primary problem with such a "social
gap" is that it could lead the military to believe that they are the true guardians of
the nation, thus increasing the need to defend, by force if necessary, the "real"
values of the Republic.
Ricks acknowledges that both this type of research and the model itself is
often problematic. To begin with, he admits that "...the evidence in this sensitive
area is admittedly hazy, partly because data are skimpy, and partly because the
definitions of "conservative" attached to the data are unstated but almost certainly
shifting" (p. 6). Further, identifying the place of optimal equilibrium is tricky as well,
though Ricks claims that if all trends are somehow negative, that this is an indicator
that civilian control may be in long term jeopardy (p. 8). For the moment, let us be
content with the fact that attitudinal models are scaled in that they attempt to

determine "trends" toward or away from a desirable social equilibrium between
civilian and military values.
Model 4: Single Principal-Single Agent Models
This is the first of the agency theory model types that I will discuss. Sharon
Weiner's article focusing on alternative congressional oversight mechanisms of the
military is a good example of how many scholars work with one civilian principalin
her case, Congressand the military as a unitary agent. Almost universally, these
models are utilized in influence and power distribution analyses that attempt to
ascertain any shift in influence and power between the civilian and military groups
under study. As a general rule, these models either explicitly exclude or ignore the
existence of multiple principals. In Weiner's case, it is likely that this model was
utilized because her research question was principal specific. She was, in her
defense, only analyzing congressional oversight mechanisms, not making general
claims about the overall health of civil-military relations in the U.S. Other scholars
are not as satisfied with proving modest claims.
Input versus output models, such as Gibson and Snyder's "issue networks"
analysis, often converges with single principal-single agent models. In spite of the
contention made by Gibson and Snyder to be utilizing a non-agency theory model,
they are splitting semantic hairs more than proposing anything unique. They study
preference inputs and policy outcomes between two institutions, which is essentially

agency theory in shorthand, though they do not explicitly assign characteristics of
agent to the military and principal to Congress. However, in their conclusion, they
state that the desire to maintain an equilibrium between the military and civilian
leadership indicates that the military is generally an object player to the subject
player of Congress, the precise "big brother, little brother" relationship that agency
theory assumes exists between branches of government and their subordinate
agencies (Feaver, 7; Gibson and Snyder, 206).
There seem to be few problems with single principal-single agent models
providing they are used to demonstrate the veracity of modest institutional claims.
The central problem with this kind of linear modelgenerally measuring increases
and decreases of relative power and influence between agencies and institutionsis
the central premise that subordinate agencies of governmental branches have no
right to be considered equally legitimate political players. In other words, agency
theory in general assumes that the agents exist solely for the purpose of faithfully
executing the political will of their parent institutions. On the other hand, as a model
used to analyze a single relationship between agencies, claims about the overall
health of U.S. civil-military relations are dubious, since the role of other principals
and, in some cases, other agents, are not adequately, if ever, incorporated into any
given analysis (Feaver, 4).
Model 5: Multiple Principal-Single Agent Models