Defining populism has been plagued by many difficulties. Looking at the dynamics of these debates, at least three patterns can be identified. First, the discussions tend to initially focus on what specific theme should determine the definition of populism. Thus, one major issue has been whether populism should be defined as an ideology or as a strategy (Mudde, 2017; Weyland, 2017). Focusing on one theme, however, is only one option within a specific family of analysis of which those who struggle for the appropriate theme to define populism are not necessarily aware. Second, scholars often believe they are defining populism, when in fact they are defining either an aggravated version of the definiendum, i.e., an authoritarian, charismatic leader who mobilises masses to achieve his or her selfish political goals; or what counts as populism is an object that is merely adjectivised as populist. And third, the definition of populism usually culminates in an initial sentence, which provides sufficient groundwork for research, but is inevitably incomplete. Though scholars understandably want to keep their definition simple, it seems inevitable to come to terms with a follow-up sentence that includes further definitional aspects until one arrives at a more than minimal, yet compact definition of populism.

The purpose of this article is to highlight several formal definitional techniques that can help address these three shortcomings in the definitional work on populism and, on this basis, clarify the role of ideology in defining populism.Formal techniques do not care about substantive aspects of definitions, nor do they care about normative expectations associated with the definiendum. Moreover, the evidence supporting the methodological formal techniques presented here lies not in the actions of populists in reality, but in the impact of mental maps on our way of grasping the world. The formal requirements in definitional work can range from the most basic to the most complex. As for the former, the definition of populism is already in formal disarray when scholars jump from one topic to another in one and the same text (critical Mudde, 2007, p.12). So Peruzzotti (2013, pp. 62, 65, 72), who refers to populism in the same article linking it interchangeably to concepts such as ‘regimes’, ‘movements’, and ‘strategy’, or ‘form of politics’. This article will, however, focus on more sophisticated formal challenges in the definitional work.

The first formal technique is to trace the generic family of analyses that tacitly underlies the typical dispute between scholars about the theme that should dominate the definition of populism. One advantage of this technique is that one becomes aware of alternative formal families of analysis. The second technique tackles the risks of consideringan either extremely strong or extremely weak manifestation of populism as populism. This technique does not propose to exclude these two extremes, but to anchor them as part of a more complex field of definition, provided that oneestablishes the conceptual centre that lies between the two extremes. Given the formal structure of the two extremesaccompanying the centre, this formal technique logically leads to a division of the definitional field into three segments.The last and third technique helps to organise the increasing elements that appear from the introductory sentence to the subsequent sentences that define populism into a chain of staggered levels. Obviously, this technique could also be used to tackle stages before the initial and after the follow-up sentences of the definition. However, to illustrate this technique, I present only three levels here, since this is sufficient to see the difference between adding content to an original definition and stabilising a definition.

In the following sections, these techniques of backtracking, building a segmented definitional field, and multilayered organisation of the definitional content will be explained. Since the issue of (thin) ideology has become a widely used mode of defining populism, the final section applies the presented techniques to assess what role ideology might play in defining populism. The point is not to deny the role of ideology in populism, but to better embed this factor in the multitude of other, perhaps more or less weighty, factors that make up populism as a whole.

The formal technique of backtracking the generic family behind the struggle for the theme defining populism

Let us imagine a typical thematic-centred discussion setting. In such a setting, some may question whether defining populism as a thin ideology is persuasive (Schroeder, 2020). A fellow contender might challenge ideology itself as the adequate theme. He might argue that populism may challenge grievances that may have already been perceived as ‘unjust’ before any leader or ‘political entrepreneur’ creates an ideology for the social movement (Fitzi, 2019, p. 53). In fact, for Snow (2013, p. 474), social movements are not ‘carriers of pre-configured…ideologies’.

Moreover, focusing on ideology blurs the fact that it is only part of an ideational dimension. Those who define populism, privileging themes such as discourse or language, actually adhere to an ‘ideational’ approach, as do those who stick to ideology. Dean and Maiguashca (2020, p. 14) even count attitudes or styles as expressions of the ‘ideational’ domain. If this is so, why choose ideology and not, say, ‘discourse’ as the fulcrum for defining populism(De la Torre and Mazzoleni, 2019, p. 83)? But again, this debate goes round in circles and remains trapped in thematic pro and con arguments. This article is about becoming aware of this revolving door by backtracking to which generic family of analysis the practice of isolating a theme like ideology, or style, or strategy belongs. Only in this way can scholars open themselves to learning from other possible analytical families. The need for complementarity with other families of analysis becomes apparent at the latest when analysts overcome a mono–thematic approach (i.e. choosing one theme such as ideology) and endorse a multi-thematic one (De la Torre and Mazzoleni, 2019, p. 83). It turns out that even a multi-thematic approach may need a methodology to know how to formally articulate the various themes in a consistent way. But let us first take a closer look at the dominating thematic family.

The thematic family of analysis

There is a tendency to define populism by pinning the concept of populism to a single theme, even though the theme itself may vary. Scholars note, for instance, that populism is a ‘means’ —of assuring political power— (Barr, 2017, p. 44); a ‘discursive phenomenon’ (Müller, 2017, p. 591); a ‘logic’ (Ochoa, 2017, p. 617); a ‘signifier’ (De Clean et al., 2018, p. 651); a ‘particular moralistic imagination of politics’ (Müller, 2017, p. 592); a ‘political strategy’ (Weyland, 2017, p. 55) or a ‘political practice’ (Jansen, 2011, p. 82). Admittedly, many of these authors begin with a list of conceivable topics. Mudde (2013, p. 2) presents ‘three meanings of populism’ as the ‘most dominant’, namely ‘redistributive politics’, ‘leadership style’, and ‘communication style’, before settling on ideology as the main theme. Priester (2011, pp. 190, 196) also offers a tripartite list, claiming that populism has been defined as an ideology, a strategy, or a discourse, and then concludes that populism is actually a mentality. These authors either select one of the listed elements as the correct one or point out to another not priorly mentioned theme. In either case, they all end up isolating one theme for defining populism.

If there is a beneficial effect of dealing with a theme, it is the pressure to organise thematic clusters. Mudde and Rovira (2012, p. 3) distinguish four ‘approaches’, namely, defining populism as movement, style, discourse, or ideology. Moffitt (2016, p. 5) concurs with the number of elements, mentioning ideology, strategy, discourse, and (political) logic; but suggests to incorporate style as a fifth element, which he concludes represents the true nature of populism.

After reviewing the relevant literature, one can outline the following thematic clusters. Works that define populism as a (thin) ideology (Mudde, 2013, p. 3), or a ‘discursive approach’, (Moffit and Tormey, 2014, p. 385) are clustered under the thematic heading of the ideational. Under this cluster, the notion of ideational ranges from (antistatus quo) discourse to the textual content (of ‘populist’ texts), and would also include what Brubaker (2019, p. 29) calls ‘interpretative frameworks’. The classical ideational content for defining populism is ideology.

Another thematic cluster under the heading logic would involve a ‘way of doing politics’ in the sense of opening up the political as proposed by Laclau (Mouffe, 2018, p. 11; Straßenberger, 2016, p. 45; Mudde and Rovira, 2012, p. 6; Aslanidis, 2017, p. 309). The ‘logic’ approach insists on dichotomising social space itself, and refers to a (logic of) articulation of demands through common symbols and a leadership able to embody this process (Laclau, 2006). Logic differs from discourse understood in ideational terms. Thus, some authors separate logic from ideology or discourse.1To comprehend populism as a logic, one must stick to Laclau’s views. While this view has been located within the ‘realm of ideas’ (Hawkins and Rovira, 2017, p. 516) or seen as overlapping with the ideological conception due to its rhetorical thrust (Urbinati, 2019, p. 32), I argue that Laclau’s understanding of logic is not ideational. As De Cleen et al., (2018, p. 652) posit, this logic is not about ideas or ideologies, but rather about how populism leads to a sociopolitical articulation.

            A further thematic cluster relates to style (Moffitt, 2016, p. 3; Ostiguy, 2017, p. 74), covering ‘visual, performative, and aesthetic elements’ and the ‘form of communication’ (Moffitt and Tormey, 2014, p. 386; Patapan, 2019, p. 744).  Here I would also include the simplification of reality as a rhetorical tool (Moffitt and Tormey, 2014, p. 387; Brubaker, 2019, p. 41), in the sense of valorising ‘common sense and first-hand experience’ (Brubaker, 2019, p. 33). When placed up front in defining populism, style does not represent a variation of an ideational approach. Thus, contrary to Mudde (2017, p. 41), I would not count style as part of the ideational cluster, nor see the ideational momentas part of style (Moffitt and Tormey, 2014, p. 384).

An additional cluster can be summarised under the theme strategy, which conceives populism as a means for achieving personalistic gains in power (Weyland, 2017; Barr, 2017). Organisational patterns (Hawkins and Rovira,2017, p. 523; Urbinati, 2014, p. 129) as well as the task of constructing goals and tactics within a populist action, could also be included into this strategic approach. Finally, there is a cluster that strikes me as the most relevant and whichhas so far been largely neglected. In this approach populism is understood as a social movement. Dean and Maiguashca(2020, p. 20), after criticising ideational approaches, rescue the idea of a ‘social-movement oriented approach’ for defining populism. To be sure, some authors repeatedly mention ‘movement’ as one possible theme for defining populism. Mudde (2013, p. 2, my italics) asks whether populism is ‘a political ideology, movement, strategy, or style’,and Urbinati (2014, p. 129) speaks of a ‘populist movement’. Yet, it remains open whether the term ‘movement’ is meant in a strong sense (even if rejected afterwards); or is simply a façon de parler. Mudde and Rovira (2012, pp. 3, 5) have pointed to a ‘movement’ variant for defining populism. ‘Seen in this light’ —Mudde and Rovira (2012, pp. 3, 5)conclude— the essential feature of populism is ‘mainly the formation of a movement appealing to very heterogeneous social groups’. I assume that the reference to movement here is more than rhetorical. But ultimately, they discard defining populism as a movement because they feel that this does not allow for a distinction between populism andother expressions of mass politics, such as Volksparteien. Furthermore, they consider that the connection to social groups, or better the search for ‘multi class alliances’, is not something specifically related to populism, but rather ‘a central element of mass politics’ (Mudde and Rovira, 2012, pp. 3, 5).

Using the term ‘movement’ may, nevertheless, be a façon de parler. Urbinati (2019, p. 113, my italics)contends, for instance, that ‘contemporary scholarship can be divided into two groups according to whether it treats populism as an ideology and a style, or whether it treats it as a strategic movement to remake political authority’. A hasty reading could lead one to believe that Urbinati understands the term movement in the sense of populism as a (social) movement. However, this is not the case. Her actual point is to understand populism as a ‘populist strategy for conquering power’ (Urbinati, 2014, p. 113). Furthermore, it may also be that the specificity of the idea of ‘movement’in the context of populism is not clear. For instance, one might think that Diehl (2011, p. 276, my italics) understandspopulism as a ‘social movement’ when she associates it with the figure of ‘collective political actors’. However, her understanding of collective actors also includes parties, whereas populism’s collective character is different from a party’s collective quality (see Rucht, 1987, p. 300; critical Hutter et al., 2019, p. 169). Because populism as a movement can pretend to govern (Rucht, 1987, p. 299), it may at some point confront its own informality as an obstacle to accessing power under the conditions of a democratic regime (Hutter et al., 2019, p. 172). If this pretention is to be realised, the movement may opt to form a party only to dissolve afterwards, or to give political resources to a party without disappearing (Rucht, 1987, p. 301; Hutter et al., 2019, p. 169). Some authors see the deterioration of links between parties and voters (Kenny, 2017, p. 48) as the main reason for the transition from movement to party, while others see this transition as the result of the crisis of representation (Hutter et al., 2019, p. 169). A populist party mediates between populism proper and the goal of officially governing a country. However, the party as such has left the domain of the social movement, which is characterised by underdeveloped institutionality (Rucht, 1987, p. 300), regardless of its varying degrees.

Up to this point, we have seen definitions that end up deciding on one theme as the guiding topic for defining populism. Populism, however, does not seem to be a one-dimensional phenomena (De la Torre and Mazzoleni, 2019, p. 94; Dean and Maiguashca, 2020, p. 20; Diehl, 2011, p. 275). Good news is, the monothematic approach is only oneanalytical strategy within the thematic family. Some authors combine at least two themes. De la Torre (2019, p. 253, my italics), for instance, defines populism as ‘political discourses and strategies’, and Brubaker (2019, p. 29, my italics) claims that populism is ‘a discursive and stylistic repertoire’. Others may conceive of populism as oscillating between themes, as Wiles (1969, p. 166, my italics) does when saying that populism is ‘any creed or movement’. Sometimes themes may appear in a historical sequence, with populism being a ‘movement’ that later morphs into ‘regime’. Others go further and understand populism as a combination of discourse, political strategy, thin ideology, political logic, and style, arguing that isolated aspects ‘taken by themselves’ are not sufficient to provide a comprehensive definition of populism (Cohen 2019, p. 393; Dean and Maiguashca, 2020, p. 20). Thus, understanding populism results from the combination of more than two issues, even if one of these issues is ultimately declared to be the most important one.

Alternative families to the thematic approach

There are two additional families beside the thematic family: the systemic and the perspectival family. Since my emphasis will lie on the systemic family, the following remarks will suffice to contour the perspectival family.

The perspectival family of analysis means taking a particular point of view, as in the so–called ‘structuralist approach’ that associates populism to a ‘type of regime’ (Hawkins and Rovira, 2017, p. 515), whereby here regime is understood as the country’s current ruling coalition (Hawkins and Rovira, 2017, p. 522). Another significance of populism as a regime has appeared when talking of a ‘full fledged populist regime’ in contrast to a ‘populist government’ (Cohen, 2019, p. 402, my italics). Likewise, Canovan (1982, p. 550) treats populism as a regime when she mentions ‘populist dictatorships’, as does Urbinati (2019, p. 135) when speaking of populism’s transformation from an ideology to a ‘power regime’. Priester (2011, p. 190; see also Mouffe, 2018, p. 11) criticises linking populism to the concept of regime. So–called ‘populist regimes’, Priester argues, should be called ‘plebiscitary leader democracy’ or competitive authoritarianism. The perspective could also be moral, sociological, or institutional. An example of a socio-historical perspective is what Rovira (2014, p. 496) calls a ‘structuralist approach’, this time in the sense of explaining populism’s emergence as ‘the product of certain transformations at the socio-structural level’. Oddly, after the structuralist and economic approaches, Rovira also mentions a ‘political-institutional’ and ‘ideological or ideational approach’. It becomes clear that he equates the perspectival family (institutional and economic) with the thematic family (strategic or ideological). Viewed up close, Rovira’s structuralist and economic approaches barely refer to defining populism, but relate more to explaining why or how populism emerged or what populism may produce in terms of policy.

To clarify the difference between perspectival and thematic perspectives, one can say that it is one thing to define populism from an ideological perspective, another to declare populism as an ideology. An ideological perspective could lead to defining populism as something that challenges capitalist oligarchies. Thus, it is possible to take an ideological perspective and still not declare ideology as the defining theme of populism.

The holistic approach as a subtype of the systemic family

More informative for the definitional work is to compare the thematic family with the systemic family, in particular with a subtype that I will call the holistic approach. It is not the same thing to combine multiple themes that togetherconstitute populism as it is to first consider populism as a whole, independent of its elements that constitute it and, in particular, independent of the specific themes those elements may represent. The holistic approach, thus, goes beyond multithematicness, asking the observer to pause for a moment until an initially contentless whole is mentally internalised. This whole is not just any whole, but approximates what Althusser (1968) called a ‘complex whole’(Gordy, 1983, p. 13). With the ‘complex whole’, Althusser sought to go beyond the commonplace that a whole is more than the sum of its parts. A complex whole means that there is one determinant element within the whole, and this element need not be the dominant one. This mindset is different from what multithematicans like De la Torre andMazzoleni (2019) achieve by calling attention to the many themes relating to populism. It is even more complex than Cohen’s (2019, p. 393, my italics) characterisation of populism as a combination of several themes, identifying strategyas the ‘most basic conceptual level’. Althusser’s complex whole is more sensitive because, based on the determinant/dominant principle detached from any specific theme or themes, the observer must not only carve out an elaborate hierarchy of the elements that make up the whole, but must decipher the rules governing the relationshipsbetween them. How this may help in the definitional work on populism will be discussed in the last section.

The formal technique of a segmented definitional field

The second formal technique is to construct a segmented definitional field to obtain a robust definition of populism. One may observe that in addition to a core meaning of populism, two complementary defining moments are regularly formulated in the works of populism experts. I call this core populism proper, and the two adjacent moments proto2-populism and perfect populism. The point here is that defining populism may remain unreliable without activating a definitional field with three segments. To flesh out this second technique, I will first explain the adjacent segmentsbefore addressing the core segment itself in section four.

Proto-populism: the weak extreme of the definitional field

Proto-populism is given when worldviews, isolated objects, individuals, or small social groupings —detached from a social movement— convey ideas or expose a behaviour that can be genuinely described as populist. Accordingly, Mudde (2017, p. 39) praises the ideational approach because it allows to see populism (also) as ‘an attitude of the individual’. Genuinely here means that the adjective ‘populist’ represents more than a rhetorical manipulation of the masses. Viewed from this perspective, Canovan’s (1982, p. 550) populism embodied in ‘revolutionary intellectuals’ isnot populism but proto-populism. This also holds true for the so–called Russian populism of the nineteenth century, which Hermet (2013, p. 85) describes as formed by ‘only small and isolated circles of intellectuals’. Conversely, proto-populism would be too narrow a label for the American populism at the end of the nineteenth century, which has beendescribed as ‘grass-root’ and ‘popular’ (Hermet, 2013, p. 85), and which Mudde and Rovira (2012, p. 3) call a ‘mass movement’.

Proto-populism is especially linked to the adjectival world, where almost anything can qualify as populist: parties, individuals, actions, speeches, or even paintings without a necessary connection to a major social subject. In the words of Dean and Maiguashca (2020, p. 14), we have here ‘incidents of populist rhetoric’ but not ‘fully fledged populist politics’. Latin America shows some examples of populist rhetoric without active ‘popular mobilisation’ (Jansen, 2011, p. 85).

The adjectival world around populism

One can ascribe a populist attitude to a person or a populist tone to a book, but this does not automatically convert them into populism proper. Nor does a politician using populist rhetoric automatically generate populism. Although populism and populist are qualitatively different terms, scholars often shift ‘imperceptibly from populist leaders to populism’(Patapan, 2019, p. 743). In her taxonomy of populism, Canovan (1982, pp. 550, 551) regrettably jumped from things being populism to things being populist (Mudde, 2017, p. 42). Hence, as Urbinati  (Urbinati, 2014, p. 129) asserts,while the Occupy Wall Street movement used a ‘populist discourse’, it did not, because of that, constitute populism.

It is important to understand that proto-populism does not coincide with the adjectival world; they merelyoverlap as shown in Figure 1. While proto-populism falls short of being populism proper, it is more than a mere demagogic reference to the people or simply an opportunistic use of populist rhetoric that belongs to what can be called the ‘pure’ adjectival world. To be part of proto-populism, the adjective populist has to have gained so much weight that it becomes part of the definitional field of populism; it must be qualitatively different from mere ingratiation to the people (Priester, 2017, p. 54). Brubaker’s (2019, p. 30) description of a politician displaying populist traits ‘only occasionally’ is insufficient. Proto-populism must imply —using Brubaker’s words— a ‘chronicle’ use of populists tools by political actors. In such a case, the adjective populist reflects something authentic concerning the people targeted by populism. Of course, a social movement can also be one of these single dispersed discrete objects adjectivised as ‘populist’. For instance, Aslanidis (2017, p. 305) speaks of a ‘populist social movement’. However, adjectivising a social movement as populist could simply mean that it behaves opportunistically in a populist mannerand thus remains in the pure adjectival world outside of the definitional field of populism. Or, if a ‘populist social movement’ pursues goals other than those specific to populism, regardless of its use of a populist ideology, it may actually represent a xenophobic right-wing social movement using some populist rhetoric. Mény and Surel (2000, p. 208) locate Le Pen, for example, within nationalism while mentioning populism only as ‘modus operandi’.

One important cognitive risk of adjectivisation, alongside the confusion between populism and populist, is its coincidence with a degreeist understanding of populism (Brubaker, 2019, p. 30; De la Torre and Mazzoleni, 2019, p. 88). Diehl (2011, p. 278) dilutes a qualitative distinction between ‘populist’ and ‘populism’ by attributing a different ‘intensity of populism’ to ‘populist’ movements, parties, or politicians. Considering that a person becomes populist, she believes that this person can exhibit a ‘high degree’ of populism (Diehl, 2011, p. 276). Similarly Aslanidis (2016, p.95), who writes that ‘discursive elements are scattered across the ideological spectrum’ leading, thus, to a ‘continuous nature of populism…where political parties are not easily classified as either populist or non-populist’. Contrary to this degreeist view, I share Mudde’s (2017, p. 35) confidence in classical definitions and separate populism as a ‘classifier’ from the term populist as a ‘qualifier’.

Figure 1: Definitional Field of Populism

Perfect Populism: the strong extreme of the definitional field

Characteristics of populism may obviously vary. In fact, my idea of a perfect populism, in which the parameters of populism proper attain a ‘high’ value (‘authoritarian leadership’, ‘constituency composed of the poor’), reflects this variability. Perfect populism corresponds to what many observers commonly think is populism. Perfect populism is not an ideal-type that never becomes reality, but embodies an exacerbated populism proper.

Perfection as exacerbation translates into the role of the leadership. Perfect populism implies a strata of society whose exclusion, driving them to scorn the establishment, necessarily comes along with the emergence of a ‘strongleader’ (Urbinati, 2014, p. 129), or a ‘political entrepreneur’ (Fitzi, 2019, p. 53). In perfect populism populist leaders behave following a top-down logic, prefer direct contact with their constituents, and are reluctant towards informed opinions. Roberts (2015, p. 683) goes even further and believes that the top-down logic between leader and masses defines populism in contrast to social movements. Again, by linking populism to the top-down issue, he actually approximates perfect populism rather than populism proper. One important fact of perfect populism is that ‘perfection’simultaneously and paradoxically signals a degeneration of populism. The degeneration that perfect populism entails stems from the fact that when a populist leader dominates his social following, he becomes the new elite (now able to manipulate the masses), which is a paradox since populism is about fighting an elite.

To conclude the presentation of the second technique, the notion of populism proper, which is at the centre of the definitional field of populism, comes next. I would like to do this by introducing the third formal definitional technique at the same time.

The formal technique of a multistoried definition

Different Layers of an Object’s definition

Of all three segments of the definitional field, populism proper represents what populism is.3 Let me start by taking up Wiles’ (1969, p. 167) thoughts on populism. He wrote: ‘Populism is strongly opposed to the establishment… It arises when a large group, becoming self-conscious, feels alienated from the centres of power’. This definition conveys worthy elements: It links populism to a social base (‘large group’), it posits populism’s contentious nature (‘opposed to the establishment’), mentions its ideational dimension (‘self-conscious’), and addresses the gap between rulers and ruled (‘feels alienated’). While Wiles’ definition is the prototype of a useful, concise definition, it may still be too minimal. Therefore, I propose to improve this definitional work by using a ‘three-layered’ scheme. The first layer represents an initial approximation to the object, characterised by definitional conciseness; in the second layer, the information about the object becomes richer, but at the obvious price of being heavier, involving additional attributesthat round out the object to be defined. The third layer, which will not be discussed in detail here, breaks down the implications and meanings of the concepts laid down in the first two layers. This third technique is inspired by Weber’s definitions of his basic concepts of sociology, which usually begin with a stipulation, immediately followed by a long comment that stabilises that stipulation. Looking at Weber’s definition of social relationships, we find it condensed into two sentences, immediately followed by dozens of sentences in seven paragraphs ‘explaining’ them (Weber, 1984, p.46). In these follow-up sentences, Weber does not extend attributes, but uses a third storey to qualify and contextualise the content of his definition, to prevent possible misunderstandings and to point out grey areas, thus, corresponding to my idea of the third storey. In the end, one has the feeling that the definition is completed only after the third storey has been taken into account. However, Weber seemingly works only with a first and a third storey, and there is no evidence that Weber would have considered his third storey ‘explications’ (Swedberg, 2020, p. 434) as part of the definition itself. In this article I want to point to the possible benefits of a second storey. While the first and second stories contain the core of the definition, I do not consider the definition complete until its stabilisation is achieved in the third storey.

 To illustrate how this works, let us resume the issue of populism proper, the core of the definitional fieldintroduced in the second technique. In my understanding, in the first storey, populism (proper) signifies first and foremost an enduring, non-elitist, non-regular, massive social movement,4 contesting5 the political gap between rulersand ruled. This initial definition can be expanded in the second storey with a follow-up sentence in this way:6 Populism is a mass social movement (see the first definitional layer) whose contestation of the particular political gap occurs after it combines with a specific code of action composed of stylistic, strategic, organisational and, above all, ideational mechanisms; not by assuming an exogenously preconceived ideational view, but by emotionally rescuing fears and resentments present in this social group and transforming these feelings into an embarrassing discourse against the political ruling elites, so as to legitimise the siege or entry into state power in order to reincorporate and redeem the people in politics, whereby the idea of ‘the people’ is not to be equated with the procedurally constructed demos that underlies democracy as a regime.

 In light of this multistoried model, the initial statements ‘populism is a social movement…etc.’ is just as insufficient as ‘populism is a thin ideology…etc.’, because both statements only capture the content of the first storey.Hence, defining populism as a thin ideology is not ‘wrong’, but has the problem that, independently of whether ideology is appropriate as a fulcrum, it conveys only reduced ‘first-layer’ information. This is why a the definition of the core of populism can only then be regarded as achieved when the contents in the first and second stories are combined. Admittedly, a multistoried definition may have all the shortcomings that a minimalist definition avoids. Nevertheless, searching for a ‘complex’ aggregation of variables (Dean and Maiguashca 2020, p. 22; Jansen, 2011, p.82; Weyland, 2017, p. 54) yields definitional benefits that ‘first storey’ definitions are unable to offer.

Aggregating the formal techniques and placing ideology

It has become clear that one may define populism so that ideology is not only one among the many aspects that constitute the phenomena known as populism, but it is also not as determinant as the fact that populism is primarily a mass social movement (MSM). No doubt, ideational aspects are decisive for populism. Ideology is part of the ideational and, in this quality, is an element of populism, in that the social movement that has taken the form of populism needs a ‘cognitive map articulating the problem’ this movement has identified as its main concern (Zurcher and Snow, 1981, p. 457).  And yet, populism is better understood as a ‘product of a certain type of social situation…’ rather than as a ‘system of ideas’ (Stewart, 1980, p. 180).

Indeed, to challenge the gap between the people and the ruling elites —the main concern of populism— requiresmore than an ideology. Quoting McCormick (2017, p. 42), one needs ‘Massenbewegungen’. Dean and Maiguashca (2020, p. 13) rightly assert in this context that populism would not attract as much interest if it were reduced to language. Relatedly, Jansen (2001, p. 85) asserts that ‘a set of ideas can float about in the ether of political discourse without ever being instantiated in an actual mobilisation project’. Turner and Killian (1987, p. 270, my italics) put it this way: ‘Only… a body of like-minded persons whose continuing interaction provides mutual support …translates fantasy into serious business’. To conclude that a MSM and not ideology is the root of populism is still far from obvious. Although Hawkins and Rovira (2017, p. 523, my italics) acknowledge that populism ‘derives from two movements of the …IXX century’, they see no need to conclude that populism is ‘primarily’ a MSM. Rather, they insiston the notion that populism is ‘primarily a moral body of thought’ (Hawkins and Rovira, 2017, p. 523).

Aslanidis (2016, p. 97) complains that those who adhere to the ideological approach ultimately measure the discourse of political actors, not their ideology. He believes to solve the problem by rejecting to consider populism as a (‘thin’) ideology, and urges to see it as a discourse. I believe that substituting ideology for discourse does not solve the definitional problem at its root. What Aslanidis calls discourse is, again, something that belongs to the ideational domain, whereas I argue that populism is a MSM that combines with a particular code, in which ideology is only a moment needed, amongst other things, to give grievances a meaningful framework for political action.

Some authors have criticised the prominent role of ideology in the mainstream definition of populism from the perspective of the social movement’s role. Roberts (2015, p. 682), for instance, argues that in conceptualising populism in ‘discurse[ive] or ideological terms’, one errs in assuming that a social movement that ‘discursively construct[s] the political order as [a] binary realm…between authentic popular will and an unaccountable…political power’ can readilybe taken as an expression of populism. He argues for emphasising the ‘different mode of popular subjectivity’ that both populism and social movements employ to distinguish themselves from one another. In my view, Roberts is right to caution us against automatically labelling a social movement that shares to some extent with populism an unease with political elites as populist. However, I think that once a social movement adopts a populist code it can turn intopopulism no matter how ‘bottom-up’ the organisational logic of the movement happens to be. That being so, contrasting Roberts (2015, p. 682), populism is not an alternative mode of representing a ‘mass constituency’ alongside social movements and political parties. On the contrary, populism means that a mass constituency acting as a social movement has chosen a populist code of action.

            I conceptualise the code of populism as composed of four dimensions: a) ideational; b) style; c) organisational; and d) strategy/tactics.7 For populism to emerge, these dimensions, or at least some of them, must take on a specific populist content. Let me briefly outline these four dimensions. The style dimension relates more specifically to the world of gestures, staging and aesthetic attitudes (Ostiguy, 2017; Moffitt and Tormey, 2014; Moffitt, 2016), rather than to the kind of hierarchies between leaders and followers or the more specific tactics for accessing the heights of state power. Under the organisational dimension, I include aspects that have already been discussed for social movement organisations in general. For instance, Della Porta and Diani (1999, p. 140) mention that social movements display a segmented, polycephalous, and reticular organisation. In this sense, the aforementioned topic of a bottom-up as opposed to a top-down logic in the relationship between a leader and the social base is an organisational question. The dimensions of the code concerning strategy and tactics deal with the concrete means the MSM uses to achieve its objectives. The implementation of mobilisations refers to tactics, while the decision not to enter the state administration would belong to the strategic dimension.

The suggested sequence of ‘combining social movement with a populist code results in populism’ (see Figure 2) may seem overly simplistic, but it helps to put the many aspects considered in competing approaches to defining populism in a meaningful order. Diehl’s important appeal to not conform to one-dimensionality carries, on the contrary,an inverted order by considering that populism can be expressed in different dimensions of political acting (Diehl, 2011, p. 276). Populism, however, does not ‘express itself’ in many things such a MSM, rather the latter becomespopulism if it adopts a specific code. Populism proper means that a MSM chooses a clear, minimally consistent, though not harmonious, populist code and acts accordingly. No populism is given as long as populism is expressed only, say, inideology.

Admittedly, one and the same theme can be present in two dimensions. For example, Diehl (2011, p. 288) locates populism’s resistance towards institutional mediation as part of what she calls the ‘socio-political organisational’ dimension of populism. It could be argued that this resistance is rather part of ideology, whereas the demand of plebiscites or the act of mobilisation are more related to what I call the tactics of populism. Likewise, opting for direct democracy can be part of both populism’s ideational framework and part of its tactics (in the form of popular initiated legislation or referenda), alongside protest or lobbying (Della Porta and Diani, 1999, p. 238). Similarly, the ‘unmediated and hierarchical’ connection between leader and people belongs to the organisational dimension of the code, yet one could also argue that it also belongs to style. These boundaries are not always clear cut.

Finally, there is the issue of coherence. Diehl (2011, p. 276) has drawn attention to the fact that the multiple dimensions of populism need not form a harmonious whole, as one dimension (say ideology) may fit populism, but another (say organisation) might not. Diehl is right to reject any kind of harmonious integration of populism’s dimensions, yet I think there must be some degree of coherence. Following the holistic approach, I argue that part of the coherence of the conceptual structure of populism is given not only by the fact that within the whole one element is determinant (i.e. the MSM), but that simultaneously one element of the code plays a dominant role within the same whole. In populism, the ideational component, and especially ideology, seems to play this dominant role. For populism proper, the decisive point is that the ideational component remains stable, evident and effective, even if the other components are weak. Put simply, it brings little impact if a group displays a populist style without seriously challenging the political divide ideologically. Questioning the political divide needs first and foremost an ideological anchoring and then, at the same time, decisive stylistic, organisational, and strategical implementations. In other words, it is not enough to recall that many issues feed into populism, as De la Torre and Mazzoleni (2019, p. 94) understandably plead, overlooking how the balance between a defining element (i.e., the MSM) and a dominant element (i.e., ideology) assure systemic coherence to the interplay of all the elements constituting the whole. It goes without saying that the holistic approach is not necessarily associated with making the MSM the determining factor. Another element, such as ideology, could take over this role.

Figure 2: Populism resulting from combining a populist code with a social movementConclusion

This article has introduced three formal techniques to fill a gap in the search for a consistent definition of populism. Introducing the backtracking of analytical families, constructing a three-segmented definitional field, and articulating a multistoried definitional procedure may contribute to scholarship’s efforts to understand populism in more nuanced ways. Rather than invalidating the role of ideology in defining populism, the new approach embeds this element in a complex structure of definition building. In terms of a holistic approach, populism is defined first and foremost as a massive social movement (the determinant element) that becomes populism once it combines with a code consisting of ideational (with ideology as the dominant element), strategic, stylistic, and organisational elements that contain a populist thrust.

Notes

1. In the context of defining populism, Mudde (2013, p. 3) considers ideology and discourse as synonymous.

2. The meaning of ‘proto’ differs from Jansen’s (2011, p. 90), who uses it for a mobilisation that is not of a large scale. It also differs from ‘proto-populism’ used by Wiles (1969, p. 168), which is more diachronic (referring to medieval peasant revolts).

3. Skeptical about trying to coin a definition of populism, see Colliot-Thélène (2019). See also Collier (2001, p. 11816). Urbinati (2019, p. 7) urges to abandon the question of what populism is, and encourages to ask what populism will do. Critical of those who shy away from defining populism, see Mudde (2017, p. 36).

4. Wiles (1969) distinguishes between ‘proto-populism’ and populism ‘proper’, linking proper to ‘mass-movement’. Regrettably, Wiles also intimated that populism proper actually existed prior to its link to a mass movement.

5. Corresponds to the idea of ‘resistance politics’ of Dean and Maiguashca (2020, p. 20).

6. Terms and notions used in this expanded definition are based on the following sources. For the ‘ideational view’ see Dean and Maiguashca (2020, p. 21). For the term ‘emotional’ and sentiments of love and populism, see further Ostiguy (2017, p. 92). For emotive or moral (‘irrational’) attraction in populism as opposed to interest or programatic reasons for collective action, see Collier (2001, p. 11814). ‘Fears’ are, of course, not unique to populism. For Lofland (1981, p.427) fears are, besides hostilities (leading to antagonism, oppositions, and resistance), a basic emotion behind collective behaviour. For the phenomena of ‘resentment’ see Jörke and Selk (2018, p. 4). For the notions of ‘embarrassment’ see Ostiguy (2017, p. 76). This notion also holds for Schedler’s (1996, p. 301) anti-political-establishment actor. For the idea of ‘reincorporation’ that is similar to the idea ‘to restore the good people to their rightful (sovereign) place…’, see Arato and Cohen (2017, p. 286). For ‘redeem’ see Ostiguy (2017, p. 92).

7. I would place the populist ‘logic’, as in De Cleen, et al., (2018, p. 654), for articulating demands leading to a confrontation between the people and the elite as part of the tactics of populism.