Professional Documents
Culture Documents
1 I would like to thank the British Academy for supporting a larger project, of which this article
is a part, by electing me to a two-year Research Readership.
A good discussion of these approaches may be found in J. B. Thompson, Studies in the Theory
of Ideology (Oxford: Polity Press, 1984).
3 By thought-behaviour it is not implied that thought is distinct from language but that it is often
poorly articulated and needs to be inferred from-because it cannot be reduced t m f t - i n ad eq u at e
speech and writing acts, which may contain additional meanings that are unintentionally commu-
nicated. See also P. Geach, Mental Acts (London: Routledge and Kegan Paul, 1 9 5 7 ) ~pp. 78, I S .
0 1994Basil Blackwell Inc., 238 Main Street, Cambridge, MA 02142,USA, and 108 Cowley Road,
Oxford, OX4 IJF, UK.
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 141
meanings, are imparted to the wide range of political concepts they inevitably
employ. Political concepts acquire meaning not only through temporally accu-
mulative traditions of discourse, and not only through spatially diverse cultural
contexts, but also by means of their particular structural position within a
configuration of other political concepts. That meaning is crucially imparted
through the morphological attributes of ideologies for, whatever else they are,
ideologies are particular patterned clusters and configurations of political con-
cepts. An ideology is hence none other than the macroscopic structural arrange-
ment that attributes meaning to a range of mutually defining political concepts.
But this is no simple structuralist assertion. For the history of an ideological
tradition, the conventions through which it is understood and perceived, and its
geographical variations play central roles in attributing meaning to the ideology
in question, superimposing diachronic on synchronic a n a l y ~ i sThe
. ~ specific struc-
tures of political thinking embedded within an ideology are themselves formed
by permissible and legitimated meanings at the disposal of a particular society.
An ideology is thus located at the meeting point between meaning and form: it
constitutes a significant sampling from the rich, but unmanageable and incom-
patible, variety of human thinking on politics, contained within and presented
through a communicable and action-inspiring pattern.
4 For a discussion of the dual importance of these perspectives see R. Koselleck, Futures Past
(Cambridge, Mass.: MIT Press, 1985), esp. pp. 73-91.
142 MICHAEL FREEDEN
including political ones, over whose uses insoluble disputes exist. On Gallie’s
own assessment the following are the most important characteristics of essentially
contested concepts. They have to be ( I ) appraisive, signifying or accrediting
‘some kind of valued achievement’; (2)internally complex; (3) containing various
rival descriptions of their component parts; and (4) open to modification in the
light of changing circumstances.s If Gallie were merely saying that, because
political concepts have normative elements within them, and because there are
no universally agreed schemes of values, it is impossible to agree on the worth
of a political concept-on whether, say, it is good or bad to pursue equality-
essential contestability would boil down to a matter of moral choices and human
tastes. It is however more than that, even on Gallie’s understanding, though he
does not supply us with sufficient grounds for appreciating why his notion is a
richer one.
To begin with, Gallie adopts a confusing characterization of an essentially
contested concept as appraisive, signifying a valued achievement. True, when we
identify a political concept such as liberty, we may wish to attach value to it.
Me may, for example, include the notion of self-determination within our concept
of liberty, and regard the former as a good, as valuable. But two major qualifi-
cations need stating. First, not all essentially contested concepts signify valued
achievements; they may equally signify disapproved and denigrated phenomena,
the ambiguity attached to the concept of power being a suitable example. ‘Ap-
praisive’ cannot solely mean ‘valued’ and certainly not, as Connolly suggests,
‘moral’;6 it must include, as Gallie does not, the notions ‘unvalued’ and ‘deval-
ued’. Second, political concepts are not merely appraisive, and this applies to
essentially contested concepts as well. To suggest that self-determination is a
value is not to deny that it also has descriptive aspects, that something must
happen in or with a person for that person to be designated as self-determining.
Concepts may have empirically describable and observable components that may
in addition be conceived of as desirable and thus become values. In similar
fashion, the speed of a car is a fact that can be described, but its speed may also
be a value if ‘time is of the essence’. Those components of concepts must be
distinguished from words and concepts that are always positively appraisive-
beautiful, wise, just. For some people liberty is always positively appraisive, but
for others it may signify an ascertainable human condition irrespective of any
desirability or undesirability that may additionally accrue to it. It is only in the
first instance that essential contestability will pertain to the simple issue of the
positive evaluative appraisal; in the second instance, as I shall presently argue,
Gallie, p. 184.
For a discussion of intension, see Social Science Concepts: A Systematic Analysis, ed. G . Sartori
(Beverly Hills: Sage, 1984),esp. Sartori, ‘Guidelines for Concept Analysis’, pp. I 5-85. Intension
refers to the features and properties included in a concept. Extension pertains to the referents, in the
external (including social) world, to which a concept applies.
Connolly, pp. 22-3.
lo As Feinberg has rightly pointed out in another context, a decision on a matter such as whether
foetuses can be included in the category of rights-bearers ‘is a conceptual, not a moral, question’
(J. Feinberg, Rights, Justice and the Bounds of Liberty [Princeton: Princeton U.P., 19801, p. 180).
I44 MICHAEL FREEDEN
See W. N. Hohfeld, Fundamental Legal Conceptions (New Haven: Yale University Press, 1919).
l2 For a discussion of these issues, see M . Freeden, Rights (Milton Keynes: Open University Press,
‘99I), PP. 34-42.
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 14s
may be culturally rather than logically essential in the human worlds as we know
them. For example, for a complete set of positive legal rights to be identical to
that of moral human rights, we would have to postulate a counter-factual world
where the law is perfect, where everyone’s understanding of fundamental human
needs is in total accord, where there is an agreed morality that generates no new
ethical thinking. This logically-conceivable philosopher’s paradise falls short of
the possibilities incumbent in socially-embedded practices. Inevitably, certain
legal rights will be preferred to others.
Contestability, in sum, will apply to both levels: the obvious contestability of
normative judgments will co-exist with the equally important contestability of
the range of components deemed to contain the empirically ascertainable units
of the concept. Nor is it necessary, for a concept to be essentially contestable,
that all its components be incompatible, or capable of being ranked in order of
importance, or culturally unrealizable. Essential contestability would obtain, even
if its units displayed internal ambiguity or multiplicity in one area alone.
Finally, Gallie’s use of ‘contested’ rather than ‘contestable’ suggests actual
conflicts on a level of ideational awareness. However, concepts can be essentially
contestable even when they are not in fact contested in a given usage, or when
not all aspects of the concept are contested. In ideological practice it is quite
possible for a concept to be contested with limited or no awareness on the part
of the contesters. Nor is it plausible to suggest, as Gallie does, that a contested
concept is derived from an original exemplar whose authority is acknowledged
by all the contestant users.13 The postulation of an original exemplar is inimical
to the very notion of essential contestability, as it presumes an agreed or correct
position from which deviations have o ~ c u r r e d . ’ ~
Gallie, p. 180.
l4 That fallacy has been maintained by some analysts of political thought influenced by Quentin
Skinner. See ‘Introduction’, Political Innovation and Conceptual Change, eds T. Ball, J. Farr and
R. L. Hanson (Cambridge: Cambridge University Press, 1989),p. 4.
Is Connolly, p. 14.
146 MICHAEL FREEDEN
within the same cultural context? And if liberty is a capsule that incorporates a
variety of properties, some overlapping with others but some possibly quite
distinct, are we talking about the same concept or about the same polysemic
word that may disguise the existence of different concepts?
The guidelines adopted to tackle this issue derive from the assertion that
words have no intrinsic meanings, but are social constructs whose meaning is
determined by their usage. Meaning has consequently to be ascertained empiri-
cally, and on that basis it will be contended that there exist a number of main
political concepts. By the main political concepts I refer to the recurrence, in
general as well as academic discourse, of terms (in no particular order of pref-
erence) such as liberty, rights, equality, justice, power, democracy, etc. To that
extent, the tradition of usage acts as a rough indicator of the presence or absence
of a political concept, though not of its agreed content. Nor does this rule out
the emergence of new political concepts as social and cultural conditions change,
or as knowledge increases. The merit of this approach is to relate empirically to
those concepts that exist as linguistic and cultural artifacts and that are, further,
in reasonable general use; otherwise the minimum permanence and spread nec-
essary to locate the very socio-cultural existence of a concept would be lacking.
Put differently, the interpretation of the meaning of political concepts must rely
on diachronic perspectives, even if-as will presently be suggested-their mor-
phology incorporates important synchronic aspects.
I now proffer the following proposition concerning the morphology of polit-
ical concepts: The main political concepts-those found both in political theo-
rizing and in ideological discourseronsist of both ineliminable features and
quasi-contingent ones.
The ineliminable features of political concepts will display two properties:
I. They are not intrinsic or logically necessary to the meaning of the word to
which they attach, but result from actual linguistic usage. In that sense, they
come close to Saussure’s notion of the ‘arbitrary’, except that we append arbi-
trariness to whatever substantive meaning is incorporated in the ineliminable
feature.I6 The feature is ineliminable simply in the sense that all known usages
of the concept employ it, so that its absence would deprive the concept of
intelligibility. Thus equality as a political concept appears always to have some-
thing to do with differences among human beings and the alleviation of those
differences, rather than, say, with mathematical identity. 2 . They cannot carry
the concept on their own, that is, the concept cannot be reduced to its inelimin-
‘ 6 Whereas Saussure appends arbitrariness to the employment of linguistic signs and their internal
relationships in a linguistic system (langue). See F. de Saussure, Course in General Linguistics
(London: Duckworth, 1983), pp. 71-8, and R. Harris, Language, Saussure and Wittgenstein (Lon-
don: Routledge, 1988), pp. 47-50.
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY I47
able component.’’ At this stage we shall merely observe that the concept will
therefore contain more than its ineliminable component, though its precise con-
tents are impossible to establish.
The choice of the word ‘ineliminable’ rather than ‘core’ is deliberate. Many
political theorists suggest that concepts have a clear core or centre and a hazy
circumference where they merge into other concepts.18 Linguists have argued
persuasively that the signified needs no essential core of meaning in order to be
the proper signified for the ~ignifier.’~ I wish to advance beyond that assertion
of relationship to examine the question of conceptual cores itself. There is reason
to doubt the clarity of the core, even to query in which sense an ineliminable
component constitutes a core or part of one. Later I will attempt to set out more
precisely the relationship between the integral components of a concept and its
circumference. If a core implies a pivotal and specific element, lucidly spelled out
and able to stand on its own, to which more marginal components are added in
order to enrich it, the main political concepts do not possess cores. Rather, they
have components that are inelirninable not in a logical sense, but simply in the
sense that an empirically-ascertainable cultural consensus ascribes to them some
minimal element or elements. Democracy may mean the rule of the people, as
its etymological meaning indicates (taking into account the burdens of translation
from the Greek), but there is no logically entailed definition of either ‘rule’ or
‘people’, or any logical reason why this arbitrary word-combination, which
signals the apparently ineliminable component of democracy, should not be
altered or made unrecognizable to earlier users. Indeed, the deft implicit insertion
of ‘liberal’ into many current accepted uses of democracy suggests a struggle-
unsuccessful to date-over granting ineliminable-component status to a new,
tacitly implied, notion.
Inasmuch as the actual linguistic usage of a concept displays a generally shared
and therefore de facto conventionally ‘constant’ feature, that feature may be
termed an ineliminable aspect of it: to eliminate it means to fly against all known
usages of the concept (though this does not rule out its removal in future). To
deny that political concepts have or can have such an ineliminable element is to
concede that the word used to represent the concept in question refers to more
than one concept. But to do that, to break down major political concepts into
hosts of minor ones, is heuristically unmanageable as well as ontologically un-
necessary. The prevalence of a particular political term that covers ostensibly
As we shall see below, the quasi-contingent features cannot carry the concept on their own
either, though their further relationship with the given ineliminable component is a complicated one.
18 B. Parekh, ‘Social and Political Thought and the Problem of Ideology’, Knowledge and Belief
in Politics: The Problem of Ideology, eds R. Benewick, R. N. Berki and B. Parekh (London: Allen
and Unwin, 1973), pp. 57-87 at p. 77.
I9 J. Culler, Saussure (London: Fontana, x976), p. 23.
148 MICHAEL FREEDEN
20 See W. L. Weinstein, ‘The Concept of Liberty in Nineteenth Century English Political Thought’,
Political Studies, 13 (1965), 145-62; M. Freeden, Liberalism Divided: A Study in British Poiiticaf
Thought 1914-1939 (Oxford: Clarendon Press, 1986), pp. 280-5.
21 For that reason, MacCallum’s triadic analysis offers a necessary, but insufficient, characterization
of the concept of liberty (G. MacCallum, ‘Negative and Positive Freedom’, Philosophy, Politics and
SocieQ, 4th Ser., ed. P. Laslett, W. G . Runciman and Q. Skinner [Oxford: Basil Blackwell, 19723,
PP. 174-93).
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY ‘49
or what would not be constrained: individuals, groups, states, any living or social
entity? And by what would they not be constrained: physical actions, internal
hindrances, impersonal forces, other objects? And how would that non-constraint
manifest itself: through not impeding the actions of the free entity, through not
interfering with its development (or non-development), through permitting the
exercise of its abilities? And would non-constraint be a valuable or deplorable
condition to be in? We can immediately see that an isolated and abstract discus-
sion of non-constraint leaves us with far too many imponderables to be able to
make sense of it. Political concepts are always richer and more sophisticated
constructs than could be inferred from their ineliminable components alone.
Why do political concepts have ineliminable components? Mainly because
concepts are idea-artifacts that serve human convenience as ways of coming to
terms with the world, and because, arguably, this is how our mind makes sense
of that world. Certain political ideas, however vague, are deemed to be both
fundamental and different from each other. They each become centrally inelim-
inable anchors for different concepts that are designed, or are formed, around
them. If liberty were to be defined entirely by a randomly fluid collection of
components, and would therefore be bereft of a stable meaning, it would have
no specific defining attributes and could be dispensed with as a specific political
concept. But this is not the case. The specificity of political concepts derives from
the combination of two factors: the presence of an ineliminable component,
albeit an undifferentiated form rather than hard and substantive; and a non-
random, even if widely variable, collection of additional components that are
locked in to that vacuous ‘de facto’ core in a limited number of recognizable
patterns.
An analogy with the overworked table might be useful here.22All tables have
one ineliminable feature: they constitute raised surfaces on which objects can be
placed; they therefore have the property of raised levelness (or approximate
levelness: the fact that tables can be tilted and caused to lose that property
temporarily does not annihilate it, just as blindfolding people will not destroy,
but simply interfere with, their capacity of sight). The further fact that not all
objects with raised level surfaces are tables merely serves to illustrate that the
ineliminable component of one item (or concept) is not necessarily exclusive to
it, but may be found within the clustered features of other items (or concepts).
By analogy, non-constraint-the ineliminable feature of liberty-may be an op-
tional feature in the cluster of the concept of rights, in particular those rights
termed choice-rights.
The attribute of a level raised surface, while indispensable to the idea of a
table, is insufficient. We have not spelled out any of its additional properties and
we have no notion of what it will look like. The surface may be brown, it may
22 Connolly draws this analogy but his discussion lacks the notion of quasi-contingency.
150 MICHAEL FREEDEN
stand on four legs, and it may be made out of wood. None of these attributes
is essential; we can dispense with any of them and yet may still have a table.
Likewise, the additional elements that attach to the ineliminable core of a political
concept are not likely to be geographically or historically universal features of
that concept.
Here, however, we encounter the important notion of quasi-contingency. The
additional components of a table, while individually dispensable, occupy cate-
gories that may not be. A table will have a colour, it will incorporate some device
to raise it from the ground, it will be made of a hardish material. Those categories
are necessary, while their particular instances are contingent. We are consequently
able to choose within each category, to offer a variety of different combinations.
Moreover, there is no inherent reason why one instance of a category should
have preference over another, why a table should be brown or white, have four
rather than three or six legs, or be made of wood rather than plastic. Because
this is structurally the case, the choice we exercise is essentially contestable, and
the feature selected to fill each necessary category is contingent to the general
idea of a table, though explicable in particular contexts and circumstances.
Why is such a category necessary? First, because without some additional
categories the core will remain vacuous, devoid of content and meaning. Second,
because particular concepts will need a specific range of categories: liberty re-
quires the notion of a subject, it requires the idea of obstruction, and it may
require an evaluation as to its desirability if it is in competition with other
concepts. Specific, contingent occupiers of those necessary categories accord sense
to the core of non-constraint. However, third, there may exist entirely contingent
categories that attach to political concepts as they may attach to tables, but have
no bearing on the fully-fledged concept or object. For example, some theorists
will regard the question of the utility of liberty as superfluous. It is similarly
beside the point whether a table is a work of art, or whether it supports a bowl
of fruit.
We have now come one step further. The above argument implies that there
are constraints on the indefinite range of meanings the idea of essential contest-
ability may be thought to invoke. The initial limitation suggests that the main
political concepts will each have to include certain categories, such as a unit of
political analysis, a view of human nature, a notion of social structure, a con-
ception of moral ends, and so forth. Beyond that there will be categories asso-
ciated specifically with the concept in question: all approaches to the concept of
authority, for example, will have to call up the concept of power. It may be
further surmised that, although the concrete instances of each category will be
contingent, their broad permutations may be limited. After all, the unit of
political analysis will have to be the individual, or society, or some other group,
or more than one such entity, but nothing else. And social structure is usually
viewed on a spectrum that, roughly speaking, takes in positions from atomism
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 151
23 This notion of constraint differs from the familiar one used by P. E. Converse, ‘The Nature of
Belief Systems in Mass Publics’, Ideology and Discontent, ed. D. E. Apter (New York: Free Press,
1964), pp. 207-12. Converse provides no theory concerning the constraints operating on the internal
structure of concepts, sees such constraints as more prevalent among elites than mass publics, rather
than a property of political concepts irrespective of their articulators, and does not allow for the
overriding of logical by cultural constraints.
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 153
tems, influential theories and beliefs (to include reactions to external events and
to unintentional or non-human occurrences).
The first type of cultural adjacency acts as a brake operating within the
framework of logical adjacency. For whereas logical adjacency must draw into
the meaning of liberty all the additional components that are entailed by the
ineliminable component, thus overloading the concept with more parts than it
can simultaneously hold (owing to possible incompatibilities among them), cul-
tural adjacency will intervene against making all those logical connections and
will thus preserve the viability of the concept. For example, it may be that in
one observed version of the concept of liberty, the answer to the logically adjacent
question ‘what is the effect of non-constraint on the subject?’ will be self-
development rather than self-determining stagnation. The choice of self-devel-
opment follows the introduction of a cultural or historical preference for an
intervening factor, in this case a specific conception of human nature, subject to
processes of evolution and maturation and capable of self-activating or enhancing
those processes. This particular conception of human nature cements the quasi-
contingent relationship between non-constraint and self-development and pre-
vents both developmental and non-developmental conduct from simultaneously
being expressions of liberty. Similarly, the cultural translation of the concept of
rights into civil and political rights is only one logical possibility, but in the
Western political tradition it would be inconceivable not to include them in any
reasonable discussion of the term.
The second type of cultural adjacency refers to elements that do not follow
logically from the ineliminable components of a concept, but are regarded in
ordinary usage as legitimate, if not indispensable. Wittgenstein observed that
‘when a sentence is called senseless, it is not as it were its sense that is senseless.
But a combination of words is being excluded from the language, withdrawn
from c i r ~ u l a t i o n . Here
’ ~ ~ the opposite happens. A combination of words, which
logicians and some philosophers would disallow, is accorded social and ideational
legitimacy. For example, it is not logical to espouse the concept of political
equality, to regard women as human beings, and yet to deny them the vote.
Nevertheless, this variant has frequently been employed even within liberal-
democratic systems, most notably Switzerland. Alternatively, ostensibly logically
paradoxical features may be culturally pressed into one concept. Some versions
of liberty will allow for force to be used in order to attain the behaviour
considered consonant with non-constraint. If non-constraint is construed as
applying to the removal of rational hindrances to individual choice, and those
hindrances are deemed internal to the subject rather than caused by others, then
the formula ‘forced to be free’ may be culturally acceptable despite what appears
to be its defective logic.
Cultural adjacency, then, refers to specific historical and socio-geographical
circumstances that encourage the association of different political concepts or
smaller idea-units within a political concept, and which either operate within
broader categories that are logically interlinked, or override such logical linkage.
It will relate to a historical and geographical usage of ideas and of language that
may either be customary or innovative. This analysis need not lead to a radically
relativist position, for it may well be the case that universal, or nigh-universal,
usages of political ideas can be demonstrated to exist. We may term this phe-
nomenon contingent universalism, for although such universalism is not an
invariable attribute of the structure and meaning of the political concept in
question, it can pertain to the concept at a particular point in time for socio-
cultural reasons.
Concepts have occasionally been analysed as composed of parts that are
combined in a particular way. These attempt to correspond, on the whole un-
successfully-as Wittgenstein argued-to the empirically observable component
parts of something composite from which we construct our picture of reality.2s
But an important element has usually been missing in such philosophical expla-
nations. When Wittgenstein claimed that a triangle can be seen as a mountain,
a wedge, a geometrical drawing, and so forth, or when he gave his famous duck-
rabbit example, using a drawing that could signify either, he was attacking the
notion of ‘what is really seen’ as a poor guide to defining a material concept.
But the failure to interpret the object known as a duck-rabbit results from the
isolated contemplation of the object. Were the duck-rabbit to be depicted as
floating on a pond, the likelihood that it was a duck rather than a rabbit would
be extremely high. Were the duck-rabbit to be located in a warren, or were a
picture of it to appear on the cover of a book called ‘The Tale of Peter Rabbit’,
the likelihood that it was a rabbit would be similarly high.
The point is that the empirical context of the material object accords its
related concept a particular interpretation. In parallel, political concepts will
gather meaning from their empirically-ascertainable ideational context, from the
idea-environment in which they are located. To that extent the synchronic rela-
tional perspective of Saussure is a useful one.Z6 Wittgenstein does indeed raise
the problem of looking at objects and concepts from multiple viewpoints, and
he usefully regards concepts as having blurred edges.27 It is precisely because
concepts are not hermetically sealed and allow for fluid, if controlled, movement
across their vague boundaries that the areas beyond those boundaries-the idea-
environments-acquire importance. However, Wittgenstein does not take the
extra step of exploring the possibility that the multivalence of concepts is revealed
only when detaching them from a given environment.28
In addition, as we have seen, political concepts regularly exhibit a feature that
material objects display infrequently and only with difficulty, namely, some of
their constituent parts are also part of their idea-environment, because they
appear simultaneously in other, external concepts. Ideas, after all, are not limited
to a particular time or space; they can be both here and there. We can hence
conceive of a concept as turned inside out, by externalizing its so-called internal
parts and treating them as free-floating units that intersect with the concept’s
ineliminable component. In other words, the analysis of a (political) concept is
inadequate in so far as all its components or properties are treated as internal
to it, as independent, self-supporting, and sharply demarcated from other con-
cepts. Conversely it will benefit from a viewpoint that sees most of those com-
ponents as externalized, and available to be drawn, in different patterns, into its
skeletal structure. Obviously, in the absence of that skeleton-the ineliminable
component-the concept would be annihilated. Its presence, however, is insuf-
ficient for the concept to emerge. Were we not to attach that skeleton, or vacuous
structure, to any one of a number of logically and culturally adjacent environ-
ments, the interpretation of a phenomenon, and the transformation of words
into meaning-endowing concepts, would become impossible tasks.
28 Wittgenstein appreciated that concepts are formed through ‘aspects of organization’, namely,
through associating different parts (Philosophical Investigations, p. Z O ~ ) ,but he did not go on to
develop the notion that locating a vacuous concept in a context or idea-environment is a prime
method of such association.
29 Even then, many soi-disant philosophers do not escape ideological argument.
156 MICHAEL FREEDEN
logical adjacencies will allow, by relying heavily on the notion of cultural adja-
cency, a socially-situated and partisan value-arbitrated choice among adjacent
components, and the result will display various mixes of rational criteria, emo-
tional inclinations and cultural value-preferences. A similar process of choice
will also be at work in deciding how to organize the wide range of political
concepts encompassed by any ideology and how to interlink them.
In parallel to logicians, most linguists would challenge the attribution of a
one-to-one relationship between signifier and signified. A word may be related
to many meanings and to changing meanings. Ideologies, however, behave in
precisely the converse manner. They aim at cementing the word-concept rela-
tionship. By determining the meaning of a concept they can then attach a single
meaning to a political term. Ultimately, ideologies are configurations of decon-
tested meanings of political concepts, when such meanings are ascribed by meth-
ods at least partly foreign to those employed by scientists, philosophers, linguists
or political theorists. Political philosophers, on the other hand, may choose not
to decontest meanings at all when they are engaged in the clarification of con-
c e p t ~ and
; ~ ~when they do engage in decontesting, they may claim to do so by
means which preserve accepted philosophical or moral standards of analysis.
In concrete terms, an ideology will link together a particular conception of
human nature, a particular conception of social structure, of justice, of liberty,
of authority, etc. ‘This is what liberty means, and thut is what justice means’, it
asserts. Ideologies need, after all, to straddle the worlds of political thought and
political action, for one of their central functions is to link the two. The political
sphere is primarily characterized by decision-making, and decision-making is an
important form of decontesting a range of potential alternatives. Thus, while the
very nature of political concepts lies in their essential contestability, the very
nature of the political process is to arrive at binding decisions that determine the
priority of one course of action over another. Ideologies serve as the bridging
mechanism between the two, converting the inevitable variety of options into
the monolithic certainty that is the unavoidable feature of a political decision. It
is plain to see why so many theorists of ideology connect that term with power,
for the act of decontesting, of deciding, of closing options, is an instance of
power-wielding. Competing ideologies are struggles over the legitimate meanings
of political concepts and the sustaining arrangements they form. But it is equally
plain that the nature of action-orientated thinking necessitates such decisions, so
that to attribute a pejorative connotation to the power aspect of ideology, in the
absence of an alternative, is a meaningless qualification. Similarly, social psy-
chologists draw attention to the inevitable cognitive ordering of the perceived
(political) world required by individuals in order to make it intelligible and
30 See D. D. Raphael, Problems of Potiticd Philosophy (London: Macmillan, 1970), pp. 11-17,
although the possibility of clarification that does not assign meaning is questionable.
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 157
31 See S. W. Rosenberg, Reason, Ideology and Politics (Oxford: Polity Press, 1988),esp. pp. 18-
‘9.
158 MICHAEL FREEDEN
32 For a useful discussion, which incidentally reinforces Saussurian insights, see B. Ollrnan, Alien-
ation, 2nd edn (Cambridge: Cambridge University Press, 1976), pp. 256-62.
33 See Sartori, op. cit., p. 5 2 .
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 159
34 R. Dworkin, ‘Liberalism’, Public and Private Morality, ed. S . Hampshire (Cambridge: Cambridge
room will be repeatedly cloned; far more probable that identical units of furniture
may be used to create very different rooms. That is, after all, what modular
furniture is designed to achieve. Ideologies, as we have argued, will display most,
if not all, of the major political concepts within their system. The key lies in the
relation of the units to one another, in their positioning vis-5-vis the centre and
in the way units are made to interlock and support each other. A room with a
table at its centre may be a billiard room, a dining room or a study. It is unlikely
to be a bedroom. If a table is surrounded by four chairs and a tablecloth, rather
than by strong arc lights, surgical equipment and an anaesthetised person, it is
most probably a dining-table, not an operating one. Now this is exactly the case
with an ideology. If we find liberty, rationality and individualism at its centre,
while equality-though in evidenceaecorates the wall, we are looking at an
exemplar of liberalism. If order, authority and tradition catch our eye upon
opening the door, while equality is shoved under the bed or, at best, one of its
weaker specimens is displayed only when the guests arrive, we are looking at a
version of conservatism. Core, adjacent and peripheral units ‘fingerprint’ the
room so that we can categorise it. It must be emphasized that only closed
ideologies will succeed in forcing out a concept altogether. Open ideologies will
have to engage, by default at least, all the major political concepts.
Matters can, albeit, become more complicated. What if my study also happens
to contain a sink and a fridge in a salient position? What if-taking this further-
two ideologies, one calling itself liberalism and the other conservatism, happen
to share the decontested concepts of limited power, liberty from intervention by
other people, constitutionalism, and private property? This is perhaps the most
important facet of ideological morphology: the absence of absolute boundaries
which separate the features of ideological systems. Multiple instances exist of
ideological hybrids that could only be described as conservative liberalism, liberal
socialism, and the like. This is however the crux of any analysis. To compart-
mentalize ideologies into prefabricated categories called socialism or liberalism
flies in the face of the evidence. Ideologies are modular structures, frequently
exhibiting a highly fluid morphology. The myriad variants they manifest can
only be broadly reduced to the few main categories, or families, we are conversant
with. It is empirically useless to entertain the notions of precise ideological
boundaries, or of features exclusive to one ideology or the other. These are
merely popular as well as scholarly conventions for simplicity’s sake. They are,
unfortunately, conventions that have caused great misunderstandings both in the
world of politics and of academe, and have frequently vitiated the subtlety
requisite for the serious investigation of ideologies.
An appreciation of ideologies as multi-conceptual constructs, and as loose
combinations of decontested concepts with a variety of internal combinations,
provides the clue to understanding the linkage between political concepts and
the many forms ideologies adopt. The comparative study of ideologies-syn-
POLITICAL CONCEPTS AND IDEOLOGICAL MORPHOLOGY 163
36 Thus while J. S. Mill and L. T. Hobhouse display strong morphological similarities, the liber-
tarianism of Hayek, by excluding the notion of rationally-guidedprogress, shares insufficient elements
with them, while it shares with some conservatives the notion of a self-generating spontaneous order.
I 64 MICHAEL FREEDEN