Frontiers of Teaching

I have been fortunate to encounter different and talented students from undergraduate classes to doctoral candidates on 3+ continents and in more than a dozen universities.

I have taught survey and elective courses on the history of the Russian imperial formations in Northern Eurasia, Soviet and post-Soviet History, Intellectual history, Nationalism, Comparative and Global History, Global History of Empires, and Historiography and Methodology of Historical Research.

I am particularly proud of my contribution to the field of global history in the form of co-organizing a doctoral program “Global Histories of Empire” (est. 2017) as a joint venture between the University of Turin (Università degli Studi di Torino) and Higher School of Economics (Национальный исследовательский университет Высшая школа экономики) in St. Petersburg. The program lives on and my former graduate students who finished the program are publishing their books in English and Russian.

For more than twenty years I have taught the class on Nationalism, including history of nationalism in Europe and the world. This arc comprised the period of intellectual dominance of classical theories of nationalism, the decline of interest in nationalism under the challenge of globalization and various forms of new global and imperial histories, and the return of interest in nationalism prompted by populist politics, wars, and the end of the end of history... As I prepare the syllabus for the next semester, I am going back to my interview with Benedict Anderson "We Study Empires as We Do Dinosaurs" published in Ab Imperio in 2003. Were empires really long dead in 2003 (what a consequential year for students of empire!)? And if empire's death was greatly exaggerated in 2003, was it not due to some essential blindness of otherwise incisive theories of nationalism? As I continue teaching nationalism, I pose these questions to my students and seek to contextualize the history of nationalism and nation-form in global and imperial history without which it risks remaining blind.

The course on the History of the Russian Empire in Eurasia is based on a new historical narrative developed together with my colleagues from Ab Imperio Journal. This narrative avoids provincial nationalisms of all sorts and addresses the bigger picture of historical change in the context of diversity in the span of one and a half millennium in the region of critical import for the world history – Northern Eurasia as part of Greater Eurasia. Threads of comparative and global and entangled history are naturally woven into this new narrative for the historical exploration of this space is not whole without the nomadic empire of Chingiz Khan, the Pacific (Alaska) and Arctic regions and, of course, the shifting boundary of Western and Eastern Europe. Today this class comes with experience of analysis of primary historical sources from the unique collection published by Bloomsbury which privileges voices from across the space of Northern Eurasia, not just from the political capitals.

In 2022 I declared to my class that the discipline of history is uniquely equipped to interpret the impeding change. In the fall of that year a drastic change was upon all of us with the release of first versions of generative artificial intelligence for lay use. While many colleagues rushed to switch to in-class exams, I thought of this challenge as part of pedagogy in a new research and educational environment. So, I invited students to experiment with available generative AI models in three broad areas relevant to humanities education and to the historical craft.

First, we tested the summarizing function and explored different strategies of reading. It quickly became apparent that while the generative AI was sufficiently adequate in piercing keywords and “factual” summary (as in basic bibliographic search), it lacked and continues to lack in exploration of semiotics of the text. This further leads to discussion of what is the nature of fact in social and cultural studies. The obvious status of fact becomes problematized when considered from the vantagepoint of meaning that can only be interpellated contextually and with full equipment of semiotics as in the famous example of difference between a blink and a wink. These reflections have been especially rewarding in the teaching of intellectual history and meaning of classical texts in social and political thought.

The second area was focused on authorship and writing/thinking. These exercises often involved students writing reviews of AI-generated essays. As with any writing, the process is both lonely and social. The semiotics of the text always include the author, the text itself, and the reception. These exercises allowed students to think critically about “authorship” and ethical dimensions of social sciences, pursuing Neo-Kantian exploration of questions of objectivity and value judgment. Above all, these exercises served as a mirror to students’ own writing, many of them reflecting on how to produce own texts and not to repeat the drawbacks that they found in AI-made texts. These classes also highlighted the importance of a question and, specifically, a research question in generating inquiry and producing non-circular research findings. Students compared AI summaries of academic texts with the actual chain of thinking and attending environment that produced them and saw with their own eyes how the spoiler of a detective story differs from enjoying the story from cover to cover. As such this area of reflection proved to be very valuable as an elemental training for a research seminar with more tangible for the current generation examples of research questions and argument development.

The third area has been particularly darling to me. I have always considered myself a student of the Russian school of historical source-critique studies. While source-critique was the foundational element in the modern discipline of history, it was somewhat sidelined in the US historical profession. Students’ encounter with the AI ranged between a techno-optimist admiration and almost religious abstinence. The source-critique approach takes away the anxieties associated with these extremes. In the eye of a historian any human or AI-produced text (including visual) is either a primary or secondary source, including the dialectics between the two. The generative AI relies on the cosmos of texts in re-production of responses to prompts. This raises the question of the chain of primary-secondary sources and brings forth the task of identifying the primary sources and their critique. I argue that a rather boring – in fact excitingly detective like – qualities of the historical craft, i.e. source-critique, become especially relevant in the present moment of the AI revolution. The generative AI tends to produce/spit out in a matter of nanoseconds flat information or data without delving into the context of its production or chain of primary-secondary source flow. Even when asked to reveal the sources behind the produced text it grudgingly gives a snippet. It is highly relevant to employ the source-critique to unpack this image of flat information/data and reflect on the mechanism of AI production of a text which in turn allows to understand the inherent limits of the presented narrative. All the more relevant is another original side of the historical source-critique. I mean the detection of fakes and forgeries which laid at the beginning of the historical craft. In the AI world these fakes and forgeries are called hallucinations. The bigger the avalanche of AI generated texts becomes with every day, the more prescient the methods of the historical discipline get. As more people use the generative AI, the more essential becomes the basic historical education and its source-critique function not just for historians, but for everyone.

The pre-class assignments were a type of assignment I had never done for a class before. It helped my learning because it forced me to not only do my readings critically but more importantly it made me pay extra attention to the things I believed AI should have been able to pick up on. The discussions on AI also helped me understand what an ethical way of using AI can look like and how I can try to include some of these ethics in my daily life and work.

The AI homework is pretty innovative, and I think more classes should try that.

Through the writing assignments, especially the AI assignments, I was able to improve my understanding of AI and to argue about what it is capable of. These assignments were also helpful to reflect on my own writing and see what the AI fails to do and what I need to improve on.

I really liked the PCA assignments. Evaluating AI-written essays provided me with a deeper understanding of the texts on which the essays were based. Moreover, the class discussions, where everyone offered different perspectives, challenged my own views on the material and significantly enhanced my learning in this course.

The assignments were particularly interesting – I had never been asked to review an AI generated essay and it was a wonderful exercise in pushing my own writing and understanding the abilities and limitations of AI as well.