The success of Ab Imperio’s New Imperial History is beyond doubt. Ricarda Vulpius praises the first Ab Imperio anthology (Novaia imperskaia istoriia postsovetskogo prostranstva, Kazan’, 2004) as a “manifesto”^[1] that not only assembles conclusions from previous research engagements but claims to constitute a new research paradigm – the New Imperial History of Russia and Eurasia. Aleksandr Kamenskii calls the editors of Ab Imperio “masters of intellectual provocation”^[2] and concedes that they have opened up an entirely new field of research.
The anthology and the journal have attracted not only condign attention. The editors’ claim to have established a new way of thinking about empire is timely. Indeed, the need to overcome older, structuralist approaches to imperial histories may be seen in one of the last issues of the flagship of theory-oriented historiography, the journal History and Theory. Although a recent call for papers asked for contributions to the complex of “theorizing empire,” all contributions found worthy of publication are either expressions of speculative philosophy of history or mere replications of older structuralist theories.^[3]
This is especially true of Richard Hellie’s article, “The Structure of Russian Imperial History.”^[4] His stature in the field of the social, economic, and legal history of Muscovy and early modern Russia notwithstanding, his conclusion – that the history of Russia has been governed by “path dependency” for over 500 years and characterized by three service-class revolutions that help to explain almost the whole of Russian history – is surprising. His use of path dependency is useless absent the concept’s theoretical background: path dependency is rooted in game theory, and without this framework it does not explain very much. Even worse is the attempt to find three “clue events” that explain the developments of more than 500 years of history.
One can therefore understand why the editors of Ab Imperio have tried to constitute a new research paradigm – or, as they say, new lenses and optics. Nevertheless, even when Roman emperors returned from a victorious campaign and entered the city in a triumphal procession, they had a slave standing behind them who murmured: memento quod es homo! My task is more ambitious: I want to elucidate some of the weaknesses of the programmatic statements of the editors of Ab Imperio from an analytic point of view. This is especially problematic, because one of the editors is not only a friend of mine but also invited me to write this critical review. It is therefore not an easy task, but appreciating this unusual openness to critical remarks, I will try to do my best.
My starting point is analytical philosophy of science – a field not very popular in the historical sciences. This is a pity. Analytical philosophy of science not only intends to clarify and thereby to improve scientific inquiry, it also has been trying (for a long time) to define rules for scientific inquiry that are valid for historiographical writing. This is true because analytical philosophy of science is concerned first of all with the logic and linguistic structure of scientific statements. Logic does not distinguish between chemistry and history.^[5] One of the most important rules of scientific logic is to make the theoretical assumptions underlying scientific statements explicit, and I will argue that the editors of Ab Imperio have not yet made this grade. This appraisement depends, as I will show, on a very strict understanding of theory.
However, this review cannot, of course, provide an alternative, coherent theory of empire. This is especially unfortunate, because I understand scientific progress in the terms of Imre Lakatos. In response to Thomas S. Kuhn’s attacks on Karl Popper’s philosophy of science, Imre Lakatos tried to show that the notion of scientific progress must not be discarded only because the development of science does not follow the rules of deductive reasoning. In fact, as Imre Lakatos was able to show, paradigms rest on nomological cores and can be identified by them. Scientific progress could now be measured by the development of a paradigm (or, as Imre Lakatos called it, “scientific research programs”^[6]) over a certain period of time. The comparison of theories would, however, demand that there are at least two alternative theories. As I cannot provide a second, alternative theory to the propositions of the editors of Ab Imperio, this is at least a serious restriction of my article.^[7] I can only hint at some aspects that might help the editors of Ab Imperio to develop a well-founded theory themselves.

WHAT IS AN EMPIRE? THE QUEST FOR AN APPROPRIATE KIND OF DEFINITION

One of the main problems the editors of Ab Imperio must confront is the impossibility of defining exactly what an empire is. The most obvious obstacle to such a definition is, in their view, the fact that “the semantics of the term ‘empire’ is overloaded with superlatives and loud epithets.”^[8] The concept of empire seems to have acquired so many meanings that reducing them to some clear or stable meaning for the purposes of historical writing has become impossible. An alternative seems to be a history of the semantics of the term “empire.”
However, the Begriffsgeschichte of empire cannot replace the definition of empire for research purposes. As Alexander Motyl put it: “the etymology of empire can tell us only how the term has been used and not what the concept means – until we first make a conceptual leap toward it.”^[9] The Begriffsgeschichte of empire is not a method but rather an object of research. To write the history of the changing meanings of the word “empire” assumes at least a stable object: the word “empire” itself. What changes is the meaning of this word. This requires a theory of language that respects the insight of the pragmatic philosophy of language, namely, that language is determined first of all by its social use.^[10] To identify empire with the changing meanings of the word would mean to write about an object of research that is not the same in the ancient world, in the Middle Ages, or in the 20th century. The result would be confusion and a lack of understanding between historians of various epochs. In sum, the history of the word “empire” is not the history of empire.^[11]
What is needed is a kind of definition that makes clear what we are talking about: to what phenomenon in the world do we refer when we use the word “empire?”^[12] Renate Mayntz, a German sociologist, has described the task of definitions as follows: “If terms shall perform their organizing and communicative tasks, they must have a consistently and precisely defined empirical referent.”^[13]
Historians have long sought Realdefinitionen that are intended to describe the substantial features, essence, or nature of an entity. A Realdefinition^[14] usually follows the classical understanding of a definition: to define a concept as a certain species you have to give the genus proximum and then to add the differentia specifica. An empire could thus be some kind of polity^[15] with specific characteristics, such as non-symmetric structures of power. The main problem with Realdefinitionen is that they are mostly based on the premise that entities exist in the world and that we are able to discover their true nature or essence. I guess this is why the editors of Ab Imperio try to avoid a stable definition of empire: Realdefinitionen tend to reify or to essentialize their Definiendum. In this respect, maybe, they claim “to focus on problems of method for the analysis of empire and not on classifications and definitions.”^[16] It is, however, difficult to “focus on methods for the analysis of empire” if they do not even clarify what they mean by empire.^[17] In short: what do they want to analyze?
An alternative understanding of definitions is more common in the natural sciences: the Nominaldefinition that can be understood as a linguistic convention. Nominaldefinitionen are usually introduced to replace already known and intersubjectively shared concepts by a shorter term. They are arbitrary fixations, and therefore they might be an alternative to Realdefinitionen from the point of view of constructivist approaches. However, Nominaldefinitionen are restricted to scientific theories in which all terms are already known and fixed. They are intended first of all to shorten scientific texts. Historical writing rarely uses fixed terms and mathematical reasoning. To conclude, Nominaldefinitionen seem to be as poor a solution as Realdefinitionen.
Anyhow, we have already gathered all the conditions for developing a valid definition that might be fruitful for historical work: it will necessarily be arbitrary (if we do not want a reifying term), and it will have to take into account the everyday use of the term “empire.” Analytical philosophy of science offers an alternative to the last two types of definitions and thus fulfills these requirements: the explication, which is intended to give an imprecise or an ambiguous term a more precise meaning for the purposes of a specific research interest. In my view, this approach is the solution for some problems of historical concept formation. An explication would comprise both a logical analysis of the meanings of the word “empire” and an empirical analysis of the conditions under which the term refers to phenomena in the world.
The logical analysis would therefore have to analyze the components of the term and to specify necessary and sufficient conditions for its use, while linguistic analysis would determine those cases in which the term could be used legitimately. The empirical analysis would have to integrate the term into a precise system of scientific terms and statements. This means that in many cases not one but several terms would have to be made explicit, and that the explication of a term is only part of a bigger undertaking. The goal is to arrive at an empirical-analytical theory – “empirical” because its task would be to explain real-world phenomena, “analytical” because it should have a deductive-nomological structure and allow covering-law explanations.^[18]
As Carl G. Hempel said, explications must fulfill at least one further requirement: they should not go too far from the existing meanings of a word. To give an imprecise term a precise meaning must not lead to an entirely new term that has nothing in common with the older term. Michel Foucault’s use of the word “archive” is a counter-example: his “archive” does not have any connection with the usual meaning of that word. It would thus be a bad explication.^[19] This does not mean, however, that all meanings of a word that are “currently available” should be embraced. Explications are not intended to merely reconstruct the meanings of a word. They have to end in a new and more precise meaning of the term.
This has consequences. As the editors of Ab Imperio rightly say, the word “empire” is often used as a pejorative in contemporary political rhetoric. It serves to discredit unwanted regimes or policies and is most often associated with illegitimacy or unjustified use of force. In this perspective, empires are something that belong to the pre-modern era and have to disappear from the map of world history. To an explication of the term “empire,” however, these associations and connotations will not contribute anything. They are all charged with normative implications and cannot account for a useful concept of empire. It is, however, no problem in principle to distinguish between the scientific and the political use of a concept.
A second consequence is that explications as such cannot be right or wrong. Explications are arbitrary fixations that at the same time have to conserve much of the older meanings. But a good explication not only preserves much of the meaning of the word. Its performance should be measured by the theory building that follows. As Carl G. Hempel says, it should be possible to develop a comprehensive, strict, and coherent theoretical system that makes use of the explicated terms.^[20] The evaluation of the explication thus follows the fruitfulness of the new term for theory-building and appraises first of all its explanatory power as part of a theory. And one of the problems in the theoretical approaches that the editors of Ab Imperio have embraced is the fact that they have not yet stated clearly what they want to explain.

“WHAT, THEN, IS THE PURPOSE OF USING THE TERM EMPIRE?”^[21]

“Theory” has become a rather diffuse concept, at least in postmodern discussions. Theories are confused with emplotments (Hayden White), mixed with descriptive languages, or merely understood as something that structures our perspectives on the world. From the point of view of analytical philosophy of science, however, theories are sets of universal statements that are logically bound together. Their task is to explain or to prescribe something. In the first case, we would call them empirical-analytical, and in the second – normative.^[22] In any case, the statements that are related to each other must not contradict each other. The problem is that one can deduce anything one wants from contradicting hypotheses. Following this definition of a theory, it once again becomes clear that definitions are an important part of any theoretical framework. Clear definitions help to avoid contradictions that might be overlooked if the terms are characterized only vaguely. It should be no surprise that for the analytical philosophy of science, defining central terms and concepts and building a theory always go hand in hand:

“If we recognize that concepts, referents, and theories are incestuously related, it follows that how concepts are bounded – how they are defined – is not just a matter of casual, or abstruse, concern. Rather, it is part and parcel – indeed, the central part – of theorizing.”^[23]

The question is: are there any more or less successful definitions of empire that are integrated into a theoretical framework and are useful within this context? If this is not the case, my demands might be too excessive. I think, however, that there are some approaches that meet these requirements. Alexander Motyl, for one, defines empire:

“as a hierarchically organized political system with a hublike structure – a rimless wheel – within which a core elite and state dominate peripheral elites and societies by serving as intermediaries for their significant interactions and by channeling resource flows from the empire to the core and back to the periphery.”^[24]

This definition is integrated into a wider theoretical framework and helps to explain something. The explanandum is not empire itself, but the fact that almost all empires in world history have broken down at one point or another: “this book aims not to provide the last word on all aspects of empires but only to make sense of the downward slope of their trajectories.”^[25]
The editors of Ab Imperio might object that Motyl’s undertaking is part of the project of modernity that interprets empires as premodern entities doomed to decay. Motyl himself, however, does not speak of historical necessities or laws that have any relation to modernity as such. He rather concentrates on determining if there are any features in empires (as he defines them) that make them vulnerable to specific historic contexts: “I argue that the very structure of empires promotes decay and that decay in turn facilitates the progressive loss of territory. At any point of this trajectory, shocks can intervene and lead to collapse.”^[26] He does not restrict this question to modernity or the nation-state. Quite the contrary, he even asks under which conditions empires might reemerge.^[27]
Another objection might be that approaches like this have normative implications, that they are driven by the conviction that it is good for empires to decay because they do not fit into the modern world, in other words: that they have to decay. Mark Beissinger concluded that those who talk about empires today often assume the legitimacy of the nation-state and the illegitimacy of empire: “But to speak about empire today is not only to make a claim about the illegitimate nature of authority; it is also to make a claim about the nature of the community subject to illegitimate control: that it is a nation deserving of self-determination and sovereignty.”^[28] In fact, these normative implications are not necessary.^[29] Motyl (as an example) does not fit into this classical criticism. His starting point is the empirical fact that most empires have broken down and that this fact demands an explanation. Accordingly, the teleology that Mark Beissinger has recognized in analyses of imperial collapses is not the appropriate point of attack. As Beissinger put it, “the so-called inevitability of imperial decline is teleological, for we recognize polities today as empires largely because they have collapsed.”^[30] First, this would not be teleological but tautological, and second, not every researcher who tries to identify the causes of imperial decline and decay would use the word “inevitability.”
A last objection stems from the editors’ programmatic article itself. In their endeavor to show that empires have to be understood as a problem rather than a structure, they say: “the key to the paradox is the fact that the analytical apparatus of modernity is entirely ‘national’ and thus empire cannot be described within any single model or metanarrative.”^[31] I do not know if I have understood it right. As far as I see, my demands for clear and stable definitions (or explications) and for strict theory building are part of the “analytical apparatus of modernity.” I do not see, however, why this apparatus should be “national.” In fact, modern analytical philosophy of science is in no way chained to “nationness.” When political scientists talk about polities, policies, or politics, most of them will not defend any norms of modernity but try to restrict their inquiry to empirical analyses of real-world phenomena. These analyses would ideally follow strict theoretical reasoning. This reasoning could be measured by the rules of scientific inquiry, and the editors of Ab Imperio are obliged to observe the same rules.
In fact, this understanding of precise definitions and strict theory-building demands more scrupulous work than any archeology or paleontology. It demands that our theories do not comprise logical incoherencies. It demands, as well, clear concepts: “Because the concepts used by a theory must be coherent and fit one another, fuzzy concepts, like weak foundations, cannot sustain even the most richly empirical and theoretically flamboyant edifices.”^[32] I do not know any archeology that has already proven this kind of consistency or rigor in concept building.^[33] Alexander Motyl is thus right to claim: “defining empires may be difficult, but it cannot be impossible. Theorizing about empires may be a challenge, but it is not insurmountable.”^[34]
Nevertheless even Alexander Motyl with his wholehearted endeavor to formulate a theory that predicts, for example, the revival of empires, does not stay the course.^[35] His argumentation is that there are four promotive factors and one necessary condition that in sum make the revival of an empire after a collapse more or less likely. The four factors are regional instability, frustrated revolutionaries, an imperial ideology, and “abandoned brethren” in the formerly imperial space. The necessary condition for imperial revival is sufficient state capacity.^[36] The main problem with this kind of theory is that it can only combine variables as is usually done in comparisons in the social sciences. The relations between these variables demand, however, a micro-theoretical foundation, as Hartmut Esser has shown.^[37] Otherwise we can only state a correlation between two or more variables, and no “explanatory” statistics can explain anything without such a foundation. Alexander Motyl’s alternative is to describe some historical situations and contexts that make his argumentation plausible on a descriptive level. It would need, however, not a descriptive but a theoretical foundation. Hartmut Esser’s solution (and I agree with him) is the integration of a theory of action into this variable sociology.

THE CASE FOR “IMPERIAL COMPARATIVISM”

The example of Alexander Motyl’s “imperiology” shows that serious theoretical work can never exclude comparisons. The editors of Ab Imperio, however, do not seem to be open-minded about “imperial comparativism,” as they call it. Ricarda Vulpius calls their distancing from comparative approaches “unimpressive” and hopes that it will soon be forgotten.^[38] The editors’ anti-comparative turn deserves, however, closer attention.
The main problem is that the editors’ averseness to comparative approaches does not correspond with their claim to offer new theoretical impulses. Theories are not just languages to express thoughts in a stylish way. They are “coherent sets of logically related statements that purport to explain something”^[39] – at least from the perspective of analytical philosophy of science. Working with theories therefore demands problem solving (Explanandum) and a certain amount of statements that purport to be universally true (although we may never be able to decide definitively the truth value of these assumptions).^[40]
The minor problem is that the editors of Ab Imperio have not yet determined the problems that the older or comparative approaches were not able to solve or the questions that they hold to be more important. Much more problematic is the fact that the editors have not yet made their law-like assumptions – which are needed for explanation – explicit.^[41] This is the task of theories: theoretical statements are law-like assumptions that claim to be valuable in any given context that corresponds to the circumstances described in the theory. In other words, theories make comparisons possible if they are well constructed because they work in any context in which they pretend to work – and not only in the one case for which they may have been developed.
The editors of Ab Imperio seem to propose another type of research. They are obviously interested in an ideographic and precise description of the Russian Empire as a single case. This ideographic interest is legitimated by theoretical considerations: “The unique internal world of empire, though, remains a thing in itself, for it was never formulated rationally and therefore cannot be built into the modern system of universal scientific knowledge.”^[42]At this point of their argumentation, one question arises immediately: why does the fact that something has never been formulated rationally mean that it cannot be part of universal scientific knowledge? Is there any logical connection between these two statements? I do not know whether Mediterranean trade behavior in the Middle Ages was formulated rationally, but my guess is that it was not. Nevertheless we are able to describe and explain mediaeval trade behavior in modern scientific language, and we are even able to integrate our knowledge about this production in broader theoretical concepts.^[43]
Again, the main point seems to be that any comparative approach might narrow our understanding of the specifics of Russian history. What the editors of Ab Imperio have in mind is the “reconstruction of the modalities of representation and self-description of each historical empire.”^[44] I agree that the modalities of representation and self-description are an interesting research object. I agree with the claim that these modalities cannot be restricted to dynastic myths or religious messianism, i.e., the discourses that we are used to calling “imperial.” But what rules should this reconstruction follow? In any case, we would have to apply rules that transcend the Russian Empire – otherwise our argumentation would necessarily become tautological and unfalsifiable.
In fact, the editors of Ab Imperio propose an unconvincing method. They conceive imperiology as paleontology, a science that “reconstructs various forms of social and political organizations and their specific contexts not immediately or easily comprehensible in the modern world.”^[45] This approach contains some presumptions that I would not like to follow. The editors of Ab Imperio say that they regard the “imperial text” as “a dead language, the grammar and semantics of which remained locked in the past.”^[46] Where can I find this imperial text? An imperial text as such does not exist, but the text metaphor is a very popular one in postmodern discussions. What is a dead language? A language cannot be dead. Who has locked grammar and semantics in the past? Instead of strict theoretical work, the editors offer metaphors that do not explain anything but fit well into postmodern stylistics. Metaphorical writing of this kind raises more questions than it answers. Raising questions referring to one’s research subject is helpful; raising questions referring to what I actually meant when I wrote down my questions is not.
Of course, the problem of “translating” this historical language is a serious one for any historiographical approach. I wonder why the editors of Ab Imperio did not take into account the long discussion about Willard van Orman Quine’s thought experiments concerning the possibility of radical translation or Donald Davidson’s widening of these experiments into the question of the possibility of radical interpretation. In fact, Quine and Davidson offer the only discussions I know about a situation in which people are confronted with a language that they do not know at all.^[47]
And finally, a research agenda that focuses “on the original semantics of empire as different from the later palimpsest of projections and appropriations of the concept of empire in modern rational discourses”^[48] is not a new idea but merely a variant of nineteenth-century German historical thought: Historisierung (historicization, to historicize), which respects the historical context of a given phenomenon and does not judge it from later developments. But above all, it is not “different from interpreting the past with the epistemological concerns of the present.”^[49] How do the editors of Ab Imperio imagine this method? Do they really believe that one can reconstruct the semantics of a past language without the epistemological concepts of the present? That would be pure positivism. I do not believe that they really propose to avoid contemporary epistemological concepts. On the contrary, they need such concepts to comprehend better, as they call it, what kind of worldview or rationality was nourished in empire.
I have, however, the impression that the editors of Ab Imperio have not distanced themselves enough from this danger. In their presentation of three research questions for the year 2005, they say that “it is possible to differentiate between categories of analysis and categories of political practice and to approach empires as a bygone practice.”^[50] In fact, categories of analysis and categories of political practice operate on different levels. Categories of analysis, as I understand them, are theoretical concepts that structure a concrete research project. Maybe the dichotomy (as formulated by the editors) is wrong. Looking for an appropriate category of analysis, they decided to formulate this category in the terms of political practices. That would come closer to Mark Beissinger’s definition of empire as sets of claims, practices, and reputations (see below) or William G. Rosenberg’s definition of empire as “a set of solutions, in practices as well as in perceptions and ideologies, intended or otherwise, to a series of central and interrelated historical problems challenging both tsarist and Soviet Russia.”^[51] The main problem with this category of analysis is how to differentiate a set of practices that constitutes an empire from another set of practices that constitutes, for example, a football team. I do not want to show the absurdness of this concept. I want to show what is still missing: almost anything can be conceptualized as a set of practices – the question will remain which practices can be considered as constituting empireness. Only after this specification would we be able to decide whether this approach is fruitful, whether it is suited to pose interesting problems and research questions, and whether it is able to provide answers for them.
A last question I would like to pose: does the fact that it is hard to define empire necessarily mean that we should rather turn to the languages of self-description to find out how historical actors thought and spoke about empire? This is first of all a misapprehension of the category “research object” and the method to analyze this object. It is not at all clear whether the editors of Ab Imperio regard the languages of self-description as a research object or as a method. But that is not the only problem. As Jьrgen Habermas warned long ago, any approach that relies on the self-interpretation of actors and groups in a given cultural context is endangered to commit itself to this self-description and therefore to lose any explanatory power. The resulting “hermeneutical idealism”^[52] can do little more than reformulate rather trivial common sense if it does not use theoretical and methodical concepts that allow these languages to be interpreted in a certain way. Historical actors cannot decide for us how to conceptualize empire (Whom would we trust? A British actor? A Russian actor?), and they cannot give us the methodological tools to analyze their utterances.
Any approach to empire would have to say what it regards as being imperial or constituting empireness. Otherwise such an approach cannot claim to explain anything. To make this point clearer: the historiography of empire has opened up new fields of research. In his impressions from the Berlin conference, Kimitaka Matsuzato singled out mental geography, studies of border territories, and influences of the outer world on imperial management.^[53] One might add other fruitful objects of research. The question is why they are part of “imperiology.” An answer to this question would imply that there is something specific and characteristic about imperial borderlands or imperial mental geographies. By looking only at the Russian Empire, you can never decide whether some specific trait of the Ukraine or Siberia is really “imperial” or rather a Russian specificity. Only comparative studies would have the potential to show the “empireness” of our Russian findings. What is so bad about comparing directly “those phenomena that were characteristic of all social and political structures of a given era?”^[54]
Jürgen Osterhammel was correct when he criticized Kathleen Wilson for flawed theorizing. He warned:

“Where you might expect the core of the new concept, you find rather superficial rhetorics. What would you think of a definition of culture as “the networks of people, practices, values and ideas spanning continents and oceans”? What of a circumscription of “British expansion” as “disparate bonds of experience, identity and practice”? With its radical adverseness against structures the New Imperial History falls short of their claim to the status of a new paradigm.”^[55]

As long as the editors of Ab Imperio do not give any clear definition of the central concepts of their new approach, the same holds true for them.

THE REIFYING POTENTIALITIES OF NON-REIFYING DEFINITIONS

There are, of course, other possibilities for defining and conceptualizing empire. One could think about Michael Doyle’s “relationship, formal or informal, in which one state controls the effective political sovereignty of another political society.”^[56] This definition might turn out to be useful for the analysis of international relations.
Some of the newer definitions seem to be more welcome by the Ab Imperio editorial staff. One could, for example, look at Mark Beissinger’s description of empires “as relationships – as sets of claims, practices, and reputations – rather than as essences of things.” This is at first sight convincing. It seems to be a fresh approach that concentrates on things common to historians: claims, practices, and reputations. But can empires really be divided from non-imperial polities based on these criteria? Do nation-states (just to take the often mentioned counterpart of empires) not enjoy some reputation? Do they not claim to be something? Are they not constituted by certain practices? Where is the difference between an empire and a non-empire?
In my view, the main problem with definitions like these is that they necessarily assume empires to be “out there” – rather than to give a precise description of one’s own concepts and then to indicate which phenomena in the world correspond to these concepts. In this respect, Beissinger is closer to regarding empires as “a timeless, substantive reality” than he thinks he is.^[57] This is, however, not an argument against the focus on practices as such. No one would disagree that “a hierarchically organized political system with a hublike structure” (an empire in Motyl’s definition) is constituted by practices and that there are claims and reputations that are part of these practices. But the practices as such do not define the empire – at least as long no one clearly defines which practices have to be regarded as “imperial.” They may be part of theory-building and help to explain how empires evolve. But still, one would have to indicate more precisely what these practices have to look like to bring about an empire. And the question of a useful definition would not have been solved.^[58]

RESEARCH CONTEXTS AND RESEARCH PROGRAMS

How can one judge, against this background, the ambitions of the editors of Ab Imperio? What are, in their perspective, the deficiencies of existing imperiological theories, and which alternatives do they offer? The bar is raised high: Ab Imperio itself speaks of an “imperial turn” in Russian and Slavic studies^[59] and of a specific “research program.”^[60] Unfortunately, I must admit that I do not understand the alternative the editors of Ab Imperio put forward in their programmatic article: “In our view, empire is a research context rather than a structure, a problem rather than a diagnosis.”^[61]
One is tempted to add that this definition is a problem rather than a diagnosis. In their endeavor to prove the shortcomings of comparative approaches, they fail to formulate an alternative theory. Their proposition is yet another “thick description”: obviously their empire has to be understood as a specifically imperial context, a specifically imperial culture, “within which [social events, behaviors, institutions, or processes] can be intelligibly – that is, thickly – described.”^[62] But then again, this is not a theory. Geertz himself pointed at the weakest point of his argumentation: his culture “is not a power, something to which social events, behaviors, institutions, or processes can causally be attributed.”^[63] It follows that culture as such cannot explain anything.^[64]
But maybe I missed the editors’ point. I have already admitted that I do not understand exactly what a “research context” might be.^[65] As far as I understand, all historians who conduct research find themselves in a research context. If the editors of Ab Imperio understand empires as contexts rather than as structures and intend to show thereby their decision to follow Geertz, I think they do not cope with Geertz’s concept of culture. Geertz understands culture, which he calls a “context,” of course as a structure. At least, his semiotic understanding of cultures defines culture as “interworked systems of construable signs”^[66] – and systems of signs are structures just like any other structure.
In the editorial remarks to the last issue in 2005, the editors of Ab Imperio finally introduce the term “imperial situation”:

“The “imperial situation” is a working hypothesis whose insights and limits are probed in a continuing discussion on the pages of Ab Imperio. This discussion testifies to the fact that the imperial situation always involves the presence of different historical subjects and the relations between them never lead to the homogenization of political, social, and cultural space.”^[67]

Primarily, I do not see how a term (“imperial situation”) can be characterized as a hypothesis. Hypotheses are proposed explanations, in most cases are formulated as causal assumptions, and must be testable. How can the term “imperial situation” be probed in ongoing discussions? Terms as such cannot be tested because they cannot be true or false – they are merely conventions. I would thus like to know more explicitly what the actual hypothesis looks like. Or should the editors mean that their working hypothesis is the assumption that by using the term “imperial situation” something that could not be seen otherwise is illuminated? In this case, I prefer the more explicit comments by Aleksei Miller, who also proposes to use a “situational approach” – but as a method and not as a hypothesis.^[68] As a method the concentration on situations rather than on regions (this is Miller’s appeal) is convincing. As a way to come to a better definition or a characterization of empire as such, the “imperial situation” concept is less persuasive. Not only in empires, but in fact in any state or society one can find different historical subjects, and the relations between them nowhere lead to homogeneous spaces.^[69]
In the end, one might suspect that the Ab Imperio empire is what the editors of Ab Imperio themselves admit: “a historical category, an analytical concept, a metaphor for heterogeneity, or simultaneously all of the above.”^[70] Such a concept makes the editors’ definition invulnerable, because it does not say anything about how they understand empire exactly. Analytical philosophy of science calls this an “immunization strategy,” because theories built on vague or ambiguous terms cannot be tested. From vague concepts and unclear causal relations you can in fact deduce any conclusion you like. Theory would thus no longer exclude anything, in other words: this kind of theory would no longer explain anything.
There is, however, a second alternative that the editors of Ab Imperio propose: “Thus, new imperial history appears in the form of an ‘archeology’ of knowledge about empire. We understand ‘archeology’ in the sense of a Foucauldian post-structuralist paradigm, which deconstructs basic and normative concepts of the social sciences and humanities.”^[71] Now the task of new imperial history becomes clearer. Empire has to be understood as a counter-model to undermine the hegemonic position of the national paradigm and thus the fact that the nation-state appears to be the one polity understood as “normal” or “right.” Archeology helps to evade “the teleological and mono-logical paradigm of the building of a nation, class, or confession.”^[72]
I would agree with the claim to overcome older teleological philosophies of history. Causality, however, has nothing in common with teleology: Foucault’s archaeology explicitly rejected teleology and at the same time renounced causality. An archaeology that wants to do Foucault’s methodology justice would furthermore have to confine itself to describe the surface of the palimpsest. It would have to picture the palimpsest as such but would not be able to say how this palimpsest came about because this would already breach Foucault’s laws of discontinuity and non-causality.

CONCLUSION

However, even a critical review should not end with confusion or wholesale destruction. The editors of Ab Imperio are absolutely right to reconceptualize empire. If it is true that the nation-state was the paradigmatic polity of modernity and that empires were not able to develop their own discourse of self-description that could equal nationalism, this is an interesting problem for further research. Languages spoken by different subjects within empire are of course part of a modern social history of the empire.
It is, however, an open question whether these languages should be analyzed within the framework of the semiotic triangle of the Moscow-Tartu school or within another linguistic framework. I would tend to prefer linguistic pragmatics over semiotics. This requires speaking not about references of words as such but about the meanings of words, understood as the reference that a speaker makes. Not the word as such refers to something, but a speaker refers to something by using a word or speaking a sentence. The point is: Meaning is first of all what a speaker or an author intends to say, in other words, it is to what he refers. This is an action, and the explanation of his utterance would therefore require a theory of action. At this point a reader or a listener is not even necessary. This methodologically individualistic theory of language^[73] must, however, be integrated into a wider social theory and a theory of action.^[74] As far as there exist social norms (Who has the right to say something in a certain context? How do you have to say something within a given context? What is permissible for discussion?), and discourses in its Foucauldian meaning are nothing more than social norms, any speech actor would be better off in most situations to regard these rules. He would, on the other side, be free to break them. An analysis of imperial languages would accordingly turn into an analysis of utterances within empire. It would ask why certain actors expressed themselves in certain situations in certain ways, to what they referred and – in a larger sense – how they interpreted this situation. At this point the proposed action theory of language fits well into Aleksei Miller’s situational approach, and maybe enhances it by including his term “situation” into a broader theory of action.
My last point is that all this can be done without any postmodern conceptual apparatus. All we need is the scientific language of modernity. In a similar manner, Alexander Motyl warned us not to rely exclusively on the exotic language of postmodernism, only because new words seem to offer better solutions for old problems. New words can even have the contrary effect: they may damage scientific progress by making scientific language vague and ambiguous. In fact, new languages may even fall into the same traps as the accused words of the older conceptual apparatus of modernity before:

“In any case, we have little reason to take the postmodernists too seriously on this score, because, their useful admonitions notwithstanding, their concepts often resemble crudely essentialist contrivances. “The other” may as well be a Platonic form, always and everywhere present for ostensibly nonsubstantialist entities to “negotiate” and “interrogate” as they define and redefine themselves relationally. “Discourse” is no less of a Thing. […] As a result, Others and Discourses are not just ontologically real in the language and logic of postmodernism. As they are virtually all that is real, they almost necessarily become portrayed as reified, fetishized, and anthropomorphized world-historical personalities of Napoleonic dimensions.”^[75]

If the editors of Ab Imperio do not want empire to become yet another reified ontological entity, if they want to preserve empire as a concept for research purposes, they will have to define it. And if they want to come to a better definition than the existing ones, they will have to integrate it into a theoretical framework, a set of logically related statements (some of them universal), and prove the fruitfulness of the new term within this theory.